90416 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

DEPARTMENT OF TRANSPORTATION For non-legal issues, Mr. Mike Pyne, Regulation Identifier Number (RIN) Office of Crash Avoidance Standards I. Executive Summary National Highway Traffic Safety (telephone: 202–366–4171) (fax: 202– Administration 493–2990). Mr. Pyne’s mailing address The PSEA requires NHTSA to is National Highway Traffic Safety establish performance requirements for 49 CFR Parts 571 and 585 Administration, NVS–123, 1200 New an alert sound that is recognizable as a [Docket No. NHTSA–2016–0125] Jersey Avenue SE., Washington, DC motor vehicle in operation that allows 20590. blind and other pedestrians to detect RIN 2127–AK93 For legal issues, Mr. Thomas Healy, nearby electric vehicles or hybrid Office of the Chief Counsel (telephone: vehicles operating at lower speeds. This Federal Motor Vehicle Safety 202–366–2992) (fax: 202–366–3820). Standards; Minimum Sound final rule establishes FMVSS No.141, Mr. Healy’s mailing address is National Minimum Sound Requirements for Requirements for Hybrid and Electric Highway Traffic Safety Administration, Vehicles Hybrid and Electric Vehicles, which NCC–112, 1200 New Jersey Avenue SE., requires hybrid and electric passenger AGENCY: National Highway Traffic Washington, DC 20590. cars and LTVs with a gross vehicle Safety Administration (NHTSA), SUPPLEMENTARY INFORMATION: weight rating (GVWR) of 4,536 kg Department of Transportation (DOT). Table of Contents (10,000 lbs.) or less and LSVs, to ACTION: Final rule. produce sounds meeting the I. Executive Summary requirements of this standard so both SUMMARY: To reduce the risk of A. Summary of Requirements of the Final blind and sighted pedestrians can more pedestrian crashes, especially for the Rule B. Costs and Benefits easily detect and recognize by hearing blind and visually-impaired, and to these vehicles. Both blind and sighted satisfy the mandate in the Pedestrian II. Background and Summary of Notice of Proposed Rulemaking pedestrians have greater difficulty Safety Enhancement Act (PSEA) of 2010 A. Pedestrian Safety Enhancement Act and detecting hybrid and electric vehicles at this final rule establishes a new Federal National Traffic and Motor Vehicle low speeds than vehicles with ICE motor vehicle safety standard (FMVSS) Safety Act engines because hybrid and electric setting minimum sound requirements B. Safety Problem vehicles produce measurably less sound for hybrid and electric vehicles. This C. Research on Vehicle Emitted Sounds at those speeds.1 At higher speeds, in new standard requires hybrid and and Detectability D. Notice of Proposed Rulemaking contrast, tire and wind noise are the electric passenger cars, light trucks and primary contributors to a vehicle’s noise vans (LTVs), and low speed vehicles E. Summary of Comments to the NPRM III. Final Rule and Response to Comments output, so the sounds produced by (LSVs) to produce sounds meeting the A. Summary of the Final Rule hybrid and electric vehicles and ICE requirements of this standard. This final B. Applicability of the Standard vehicles are similar. rule applies to electric vehicles (EVs) C. Critical Operating Scenarios and to those hybrid vehicles (HVs) that D. Crossover Speed Hybrid vehicles with gross vehicle are capable of propulsion in any E. Acoustic Parameters for Detection of weight rating (GVWR) of 4,536 kg forward or reverse gear without the Motor Vehicles (10,000 lbs.) or less are 1.18 times more vehicle’s internal combustion engine F. Acoustic Parameters for Recognition of likely than an ICE vehicle to be involved (ICE) operating. This standard will help Motor Vehicles in a collision with a pedestrian and 1.51 G. Frequency (Pitch) Shifting and Volume times more likely to be involved in a to ensure that blind, visually impaired, Change and other pedestrians are able to detect collision with a pedalcyclist. NHTSA H. Sameness assumes that this difference in accident and recognize nearby hybrid and I. Customer Acceptance electric vehicles, as required by the J. Test Conditions rates is mostly attributable to the PSEA. K. Test Procedure pedestrians’ inability to detect the L. Phase-in of Requirements presence of these vehicles through DATES: Effective date: This rule is IV. International Harmonization and hearing. effective February 13, 2017. Stakeholder Consultation Compliance date: Initial compliance To further evaluate the assumption V. Analysis of Costs, Benefits, and that the difference in crash rates is is required, in accordance with the Environmental Effects phase-in schedule, on September 1, A. Benefits mostly attributable to differences in 2018. Full compliance is required on B. Costs vehicle emitted sound, the agency September 1, 2019. C. Comparison of Costs and Benefits conducted research to see if there was Petitions for reconsideration: Petitions D. Retrospective Review a difference in the ability of pedestrians for reconsideration of this final rule E. Environmental Assessment to detect approaching hybrid and must be received not later than January VI. Regulatory Notices and Analyses electric vehicles versus ICE vehicles. 30, 2017. Executive Order (E.O.) 12866 (Regulatory The agency also conducted research to Incorporation by Reference: The Planning and Review), E.O. 13563, and examine how the frequency incorporation by reference of certain DOT Regulatory Policies and Procedures composition of a sound influenced the Executive Order 13609: Promoting publications listed in the standard is International Regulatory Cooperation ability of pedestrians to detect that approved by the Director of the Federal National Environmental Policy Act sound in the presence of ambient noise. Register as of February 13, 2017. Regulatory Flexibility Act Section II.C provides much more ADDRESSES: Petitions for reconsideration Executive Order 13132 (Federalism) information on this research and how of this final rule must refer to the docket Executive Order 12988 (Civil Justice the agency used it in the context of this and notice number set forth above and Reform) rulemaking. be submitted to the Administrator, Unfunded Mandates Reform Act Paperwork Reduction Act 1 Garay-Vega, L; Hastings, A.; Pollard, J.K.; National Highway Traffic Safety Executive Order 13045 Administration, 1200 New Jersey Zuschlag, M. & Stearns, M. (2010, April). Quieter National Technology Transfer and Cars and the Safety of Blind. Pedestrians: Phase 1. Avenue SE., Washington, DC 20590. Advancement Act DOT HS 811 304. Washington, DC: National FOR FURTHER INFORMATION CONTACT: Executive Order 13211 Highway Traffic Safety Administration.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00002 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90417

A. Summary of Requirements of the 30 km/h. The final rule also adopts the agency acknowledges that the concept Final Rule agency’s proposal to conduct of increasing sound pressure level with On January 14, 2013, NHTSA compliance testing outdoors. increased speed is not a direct published a notice of proposed With regard to the scope of the final replacement for pitch shifting, but we rulemaking (NPRM) specifying rule and what level of sound to emit and believe it is a reasonable alternative that minimum sound requirements for when, however, the agency is adopting will provide useful audible information hybrid and electric vehicles.2 The numerous changes to the proposal in to pedestrians about the operating state response to additional analysis of nearby vehicles. NPRM discussed three alternative • means for the agency to establish conducted by the agency and in The NPRM proposed that sound requirements for, and measure response to comments, including the emitted by hybrid and electric vehicles compliance with, minimum levels of following: must contain one tone no higher than • The final rule will only apply to vehicle emitted sound. In the NPRM, 400 Hz and emit broadband content four-wheeled hybrid and electric the agency proposed its preferred including each one-third octave band vehicles with a gross vehicle weight alternative which was to establish from 160 Hz to 5000 Hz so that sounds rating (GVWR) of 4,536 kg (10,000) minimum requirements for vehicle emitted by these vehicles would be pounds or less. The NPRM proposed emitted sound using a psychoacoustic recognizable as motor vehicles. The that this rule would also apply to hybrid model. Sounds meeting the proposed final rule does not adopt these proposed and electric vehicles with a GVWR over requirements would contain acoustic requirements. We believe that 4,536 kg (10,000) pounds and to electric elements designed to enhance detection pedestrians will use other cues to motorcycles. We believe that we do not and to aid pedestrians in recognizing recognize EVs and HVs such as the have enough information at this time to the sound as coming from a motor location of the sound source and the apply the minimum acoustic vehicle. We believed that the preferred frequency and level changes caused by requirements of this final rule to these alternative placed the greatest emphasis the motion of the sound. vehicles. • In order to ensure that hybrid and on ensuring the vehicle emitted sounds • In this final rule, the agency is electric vehicles of the same make, were detectable to pedestrians. In reducing the number of one-third octave model, and model year emit the same addition to the preferred alternative, the bands for which there are minimum sound, as required by the PSEA, the NPRM also discussed minimum sound requirements. The NPRM proposed that NPRM proposed that vehicles of the requirements for HVs and EVs designed vehicles would have to emit sound same make, model, and model year to resemble sounds produced by ICE meeting minimum requirements in eight must emit the same level of sound, vehicles. This alternative would place a one-third octave bands. To comply with within 3 dB(A), in each one-third octave greater emphasis on recognizability than this final rule, hybrid and electric band from 160 Hz to 5000 Hz. We have the preferred alternative. Compliance vehicles will instead have to meet a instead decided to ensure that EVs and with both of these alternatives would be requirement specifying either two or HVs of the same make, model, and determined using a compliance test that four one-third octave bands. Vehicles model year emit the same sound by measured the sound produced by the complying with the four-band requiring that all vehicles of the same vehicle. requirement must meet minimum sound In order to provide an alternative that make, model, and model year use the pressure levels in any four non-adjacent would allow the most flexibility in the same alert system hardware and one-third octave bands between 315 Hz types of sounds that manufacturers software, including specific items such and 5000 Hz, including the one-third could choose to add to vehicles to alert as the same digital sound file where octave bands between 630 Hz and 1600 pedestrians, we also discussed using applicable, to produce sound used to Hz (these bands were excluded in the human factors testing to determine meet the minimum sound requirements NPRM). Vehicles complying with the whether a sound used to alert in today’s final rule. two-band requirement must meet • The NPRM proposed that each pedestrians was recognizable as a motor minimum sound pressure levels in two hybrid and electric vehicle must meet vehicle. After careful consideration of all non-adjacent one-third octave bands minimum sound requirements anytime available information, including the between 315 Hz and 3150 Hz. For the the vehicle’s propulsion system is public comments submitted in response two-band requirement, one band must activated, including when the vehicle is to the NPRM,3 the agency has decided be below 1000 Hz and the second band stationary. The final rule requires each must be at or above 1000 Hz, and the hybrid and electric vehicle to meet to adopt the preferred alternative in the two bands used to meet the two-band minimum sound requirements any time NPRM and many of the elements of the requirement also must meet a minimum the vehicle’s propulsion system is proposed rule. In the final rule, as band sum requirement. activated, including when the vehicle is proposed, the agency requires hybrid • The NPRM proposed that the stationary, unless the vehicle’s gear and electric vehicles to emit sound fundamental frequency of the sound selector is in the ‘‘park’’ position or the while the vehicle is stationary with the emitted by a hybrid or electric vehicle parking brake is applied (the latter for vehicle propulsion system activated. must vary as the vehicle changes speed HVs and EVs with manual (However, in the final rule this by one percent per km/h for speeds transmissions). requirement does not apply to vehicles between 0 and 30 km/h to allow • The NPRM proposed a phase-in that are parked with the propulsion pedestrians to detect vehicle schedule that required each system activated—see below.) Also as acceleration and deceleration. This manufacturer of hybrid and electric proposed, the agency requires hybrid requirement was referred to as ‘‘pitch vehicles to begin meeting the and electric vehicles to emit minimum shifting,’’ and it is not required in the requirements of the final rule with 30 sound levels while in reverse and while final rule. Instead, the final rule assists percent of the hybrid and electric the vehicle is in forward motion up to pedestrians in detecting increases in vehicles they produce three years before vehicle speed by requiring vehicle- the date for full compliance established 2 78 FR 2797. 3 ‘‘Federal Motor Vehicle Safety Standards; emitted sound to increase in sound in the PSEA. In the final rule, we have Minimum Sound Requirements for Hybrid and pressure level by a specified amount as modified the phase-in schedule to Electric Vehicles,’’ 78 FR 2798 (January 14, 2013). the vehicle’s speed increases. The provide additional time for compliance

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00003 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90418 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

for manufacturers of light vehicles; 50 for this rulemaking. If we assume that pedestrian and pedalcyclist injury rate percent of each manufacturer’s HV and HVs and EVs increase their presence in per registered vehicle for EVs HVs to EV production must comply with this the U.S. fleet to four percent of all ICE vehicles when four percent of the final rule one year before the date for vehicle registrations in model year 2020, fleet is HVs and EVs would be 2,390 full compliance established in the PSEA a total of 2,464 injuries to pedestrians fewer injured pedestrians and of September 1, 2019. and pedalcyclists would be expected pedalcyclists. We do not include any B. Costs and Benefits over the lifetime of the 2020 model year quantifiable benefits in pedestrian or fleet due to the pedestrians’ and pedalcyclist injury reduction for EVs As discussed in detail in Section V of this notice, the benefits of this final rule pedalcyclists’ inability to detect these because we believe it is reasonable to will accrue from injuries to pedestrians vehicles by their sense of hearing. assume that EV manufacturers would that will be avoided, based on the Taking into account the agency’s have installed alert sounds in their cars anticipated ability of this rule to reduce estimate of detectability of vehicle alert without passage of the PSEA and this the pedestrian injury rate for HVs and sounds complying with this final rule, proposed rule.4 We also estimate that EVs to that of ICE vehicles. As discussed which is discussed in the Final this rule will result in 11 fewer injured in Section II.B, a traditional analysis of Regulatory Impact Assessment, we pedestrians and pedalcyclists caused by pedestrian fatalities is not appropriate estimate that the benefit of reducing the LSVs.

TABLE 1—DISCOUNTED BENEFITS FOR PASSENGER CARS AND LTVS, MY2020, 2013$

Pedestrians Pedalcyclists Total PED + CYC 3% 3% Total 3% Total 3% Total Discount Discount monetized Total ELS Discount monetized Total ELS Discount monetized Total ELS factor benefits factor benefits factor benefits

(PC) ...... 0.8024 $132.3M 9.70 0.80243 $168.8M 14.55 0.8024 $301.1M 24.25 (LTV) ...... 0.7867 7.9M 0.58 0.78673 9.4M 0.80 0.7867 17.4M 1.39

Total .. 0 140.3M 10.29 0 178.3M 15.35 0 318.5M 25.64

7% 7% Total Total ELS 7% Total Total ELS 7% Total Total ELS Discount Discount monetized Discount monetized Discount monetized factor benefits factor benefits factor benefits

(PC) ...... 0.6268 $102.5M 7.50 0.62684 $130.5M 11.24 0.6268 $233.0M 18.74 (LTV) ...... 0.6077 6.1M 0.45 0.60775 7.2M 0.61 0.6077 13.3M 1.06

Total .. 0 108.6M 7.94 0 137.7M 11.85 0 246.3M 19.80

TABLE 2—TOTAL COSTS FOR PCS AND LTVS, MY2020, 2013$

Avg. Install Sales Sales Fuel Fuel costs install costs Total Total costs impacted costs/veh (total) costs/veh total cost/veh

3% discount: (PC) ...... 8,000,000 483,462 $4.70 $2,272,270 $74.36 $35,951,512 $79.06 $38,223,782 (LTV) ...... 8,000,000 46,428 5.30 246,067 71.97 3,341,333 77.27 3,587,400

Total ...... 16,000,000 529,889 $4.75 $2,518,337 $74.15 $39,292,845 $78.91 $41,811,182 7% discount: (PC) ...... 8,000,000 483,462 $3.80 $1,837,155 $74.36 $35,951,512 $78.16 $37,788,667 (LTV) ...... 8,000,000 46,428 4.20 194,996 71.97 3,341,333 76.17 3,536,329

Total ...... 16,000,000 529,889 3.84 2,032,151 74.15 39,292,845 77.99 41,324,996

TABLE 3—COSTS AND SCALED BENEFITS FOR LSVS, MY2020 5

Sales ratio Scaled Discount LSV to light Scaled Scaled benefits minus rate vehicle Sales Scaled costs injuries Scaled ELS benefits scaled (%) (%) (undisc.) costs

3 ...... 0.47 2,500 $197,264 11.28 0.1210 $1,502,807 $1,305,543 7 ...... 0.47 2,500 194,970 11.28 0.0934 1,161,989 967,019

4 As further discussed in the agency’s Final market and not the rule. This assumption makes our 5 Scaled benefits and costs for low-speed vehicles Regulatory Impact Analysis, due to foresight on the benefit figures conservative. On the other hand, we (LSVs) are estimated to be directly proportional to part of light electric vehicle manufacturers, paired did not assume that electric LSVs would be costs for light vehicles based on sales. Scaled costs with consumer expectations and style choices, light voluntarily equipped with speaker systems since include both installation costs for the system and vehicle EVs are all assumed to be equipped with none of these vehicles were known to have such fuel costs. speaker systems. NHTSA assumes the sound alert benefits for these vehicles are attributable to the systems currently.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00004 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90419

NHTSA estimates that the fuel and which they are installed, if LSVs will be 25.76 equivalent lives installation cost of adding a speaker manufacturers make no other changes to saved at the 3-percent discount rate and system in order to comply with the reduce vehicle weight, these vehicles 19.92 equivalent lives saved at the 7- requirements of this rule is $129.84 per would consume an additional 2.3 more percent discount rate (summing values vehicle for unequipped hybrid light gallons of fuel over the lifetime of a from Table 1 and Table 3). Converting vehicles (i.e., vehicles that did not passenger car and 2.5 more gallons of that to dollars, the benefits of this rule previously have any alert system fuel over the lifetime of a light truck for the HV portion of the MY 2020 light components installed), and $54.99 for which would result in an average fuel vehicle and LSV fleet are $320.0 million electric light vehicles. We estimate that cost of $4.75 per vehicle for over the at the 3-percent discount rate and lifetime of MY 2020 vehicles subject to for model year (MY) 2020, which is the $247.5 million at the 7-percent discount the rule at the 3-percent discount rate first model year to which the rate (Table 4).6 NHTSA estimates that and $3.84 per vehicle for over the requirements of this final rule will lifetime of MY 2020 vehicles subject to the cost per equivalent life saved for the apply to the entire light vehicle fleet, the rule at the 7-percent discount rate.). light EV, HV, and LSV fleet would range this final rule will apply to 529,889 To more easily compare the costs and from a cost of $1.67 million to a cost passenger cars and LTVs. The estimated benefits of this rulemaking, we have savings of $0.10 million across the 3- costs for manufacturers of complying converted pedestrian and pedalcyclist percent and 7-percent discount levels, with this rule is $39.29M in MY 2020, injuries avoided into equivalent lives respectively. When compared to our and we would expect that due to the saved. We estimate that the impact of comprehensive cost estimate of the additional weight that these this rule in pedestrian and pedalcyclist value of a statistical life of $9.2 million, components add to the vehicles in injury reduction in light vehicles and this final rule is cost effective.

TABLE 4—TOTAL BENEFITS AND COSTS SUMMARY FOR LIGHT VEHICLES AND LOW SPEED VEHICLES, MY2020, 2013$

3% Discount 7% Discount rate rate

Total Monetized Benefits ...... $320.0M $247.5M Total Costs (Install + Fuel) ...... 42.M 41.5M Total Net Impact (Benefit¥Costs) ...... 278.0M 205.9

II. Background and Summary of Notice statute, as a vehicle-emitted sound that ‘‘discern vehicle presence’’ along with of Proposed Rulemaking enables pedestrians to discern the the requirements that the sound allow presence, direction,12 location, and A. Pedestrian Safety Enhancement Act pedestrians to discern direction, operation of the vehicle.13 Thus, in and National Traffic and Motor Vehicle location, and operation. The term order for a vehicle to satisfy the Safety Act ‘‘presence’’ means something that is in requirement in the PSEA to provide an the immediate vicinity. The term On January 4, 2011, the Pedestrian ‘‘alert sound,’’ the sound emitted by the ‘‘operation’’ means a state of being Safety Enhancement Act of 2010 (Pub. vehicle must satisfy that definition. The functional or operative. Read together, L. 111–373) was signed into law. The alert sound must not require activation the definition of alert sound requires Pedestrian Safety Enhancement Act by the driver or the pedestrian, and that pedestrians be able to detect vehicle (PSEA) requires NHTSA to conduct a must allow pedestrians to reasonably presence when the vehicle is in rulemaking to establish a Federal Motor detect an EV or HV in critical operating operation. A vehicle with its gear Vehicle Safety Standard (FMVSS) 7 scenarios such as constant speed, selector not in ‘‘park’’ is in an requiring an ‘‘alert sound’’ 8 for accelerating, or decelerating. operational state even though it may not pedestrians to be emitted by all types of In addition to those operating motor vehicles 9 that are electric scenarios, the definition of alert sound be moving. It is therefore the agency’s vehicles 10 (EVs) or hybrid vehicles 11 in the PSEA requires the agency to position that the provision of the PSEA (HVs). Trailers are specifically excluded establish requirements for a sound that requires pedestrians to be able to from the requirements of the PSEA. while the vehicle is stationary but active detect the presence of a vehicle in The PSEA requires NHTSA to and when the vehicle is operating in operation requires that the vehicle emit establish performance requirements for reverse. PSEA states that the alert sound a minimum sound level when its gear an alert sound that allows blind and must allow pedestrians to ‘‘discern selector is in any position other than other pedestrians to reasonably detect a vehicle presence, direction, location, ‘‘park,’’ whether that be when the nearby EV or HV. The PSEA defines and operation.’’ 14 We read the vehicle is moving forward, stationary, or ‘‘alert sound,’’ as that term is used in the requirement that pedestrians be able to operating in reverse.

6 NHTSA’s benefits calculation does not include 9 Section 2(4) of the PSEA defines the term 11 Section 2(9) of the PSEA defines ‘‘hybrid light EVs because manufacturers of light EVs were ‘‘motor vehicle’’ as having the meaning given such vehicle’’ as a motor vehicle which has more than already adding sound to those vehicles prior to term in section 30102(a)(6) of title 49, United States one means of propulsion. As a practical matter, this NHTSA issuing the NPRM. However, this analysis Code, except that such term shall not include a term is currently essentially synonymous with trailer (as such term is defined in section 571.3 of includes LSVs because those vehicles currently do ‘‘hybrid electric vehicle.’’ title 49, Code of Federal Regulations). Section not have added sound. 12 The PSEA does not specify whether vehicle 7 30102(a)(6) defines ‘‘motor vehicle’’ as meaning a NHTSA is delegated authority by the Secretary vehicle driven or drawn by mechanical power and ‘‘direction’’ is to be defined with reference to the of Transportation to carry out Chapter 301 of Title manufactured primarily for use on public streets, vehicle itself (thus meaning forward or backward) 49 of the United States Code. See 49 CFR 501.2. roads, and highways, but does not include a vehicle or the pedestrian. This includes the authority to issue Federal motor operated only on a rail line. 13 PSEA Section 2(2). vehicle safety standards. See 49 U.S.C. 30111. 10 Section 2(10) of the PSEA defines ‘‘electric 14 Public Law 111–373, 2(2), 124 Stat. 4086 8 The definition of the term ‘‘alert sound’’ is vehicle’’ as a motor vehicle with an electric motor discussed below. as its sole means of propulsion. (2011).

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00005 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90420 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

The agency believes that it is vehicle’s alert sound were determined The PSEA defines ‘‘conventional motor reasonable to conclude that Congress to contain a safety related defect.19 vehicle’’ as ‘‘a motor vehicle powered intended the term ‘‘operation’’ in the Under the PSEA, the standard must by a gasoline, diesel, or alternative PSEA to be the condition in which a specify performance requirements for an fueled internal combustion engine as its driver is operating the vehicle, as alert sound that enables blind and other sole means of propulsion.’’ 24 We opposed to just the operation of the pedestrians to reasonably detect EVs believe that if Congress had intended vehicle’s propulsion system. It is the and HVs operating below their crossover the alert sound required by the PSEA to operation of the vehicle by a driver, not speed.20 The PSEA specifies several be recognizable as an ICE vehicle, the operation of the vehicle’s propulsion requirements regarding the performance Congress would have specified that the system, that creates the safety risk to of the alert sound to enable pedestrians sound must be recognizable as a pedestrians who fail to detect hybrid to discern the operation of vehicles ‘‘conventional motor vehicle’’ in and electric vehicles. Consequently, subject to the Act. First, the alert sound operation rather than a motor vehicle when the vehicle’s gear selector is in must be sufficient to allow a pedestrian because Congress acts purposefully in ‘‘park,’’ the propulsion system may or to reasonably detect a nearby EV or HV its choice of particular language in a may not be activated but, in such a operating at constant speed, statute.25 condition when the propulsion system accelerating, decelerating or operating While the mandate that NHTSA is activated, the vehicle is not operable in any other scenarios that the Secretary develop performance requirements for by the driver until the gear selector is deems appropriate.21 Second, it must an alert sound that is recognizable as a moved from ‘‘park’’ to some other gear reflect the agency’s determination of the motor vehicle does not mean that the selector position. Therefore, we have minimum sound level emitted by a sound must be based solely on sounds determined that the PSEA does not motor vehicle that is necessary to allow produced by ICE vehicles, the mandate require us to establish minimum sound blind and other pedestrians to does impose substantive requirements requirements for when a vehicle has its reasonably detect a nearby EV or HV that the agency must follow during the gear selector control in the ‘‘park’’ operating at or below the crossover rulemaking. The Vehicle Safety Act position. speed.22 Today’s final rule will ensure defines a motor vehicle as a ‘‘vehicle Because the PSEA directs NHTSA to that EVs and HVs are detectable to driven or drawn by mechanical power issue these requirements as an FMVSS pedestrians by specifying performance and manufactured primarily for use’’ on under the National Traffic and Motor requirements for sound emitted by these public roads.26 The requirement that the Vehicle Safety Act (Vehicle Safety vehicles so that they will be audible to agency develop performance Act),15 the requirements must comply pedestrians across a range of ambient requirements for recognizability means with that Act as well as the PSEA. The noise environments, including those that the pedestrian alert sound required Vehicle Safety Act requires each safety typical of urban areas. by this standard must include acoustic standard to be performance-oriented, Nothing in the PSEA specifically characteristics common to all sounds practicable 16 and objective 17 and meet requires the alert sound to be produced by vehicles driven by the need for safety. In addition, in electrically generated. Therefore, if mechanical power that make those developing and issuing a standard, manufacturers wish to meet the sounds recognizable as a motor vehicle NHTSA must consider whether the minimum sound level requirements based on the public’s experience and standard is reasonable, practicable, and specified by the agency through the use expectations of those sounds. appropriate for each type of motor of sound generated by the vehicle’s The PSEA mandates that the standard vehicle covered by the standard. power train or any other vehicle shall not require the alert sound to be As an FMVSS, the minimum sound component, there are no conflicts with dependent on either driver or pedestrian standard in today’s final rule will be the PSEA to limit their flexibility to do activation. It also requires that the safety enforced in the same fashion as other so. standard allow manufacturers to safety standards issued under the The alert sound must also reflect the provide each vehicle with one or more Vehicle Safety Act. Thus, violators of agency’s determination of the alert sounds that comply, at the time of the standard will be subject to civil performance requirements necessary to manufacture, with the safety standard. penalties.18 Vehicle manufacturers will ensure that each vehicle’s alert sound is Thus, a manufacturer may, if it so be required to conduct a recall and recognizable to pedestrians as that of a chooses, equip a vehicle with different provide remedy without charge if their motor vehicle in operation.23 We note sounds to denote different operating vehicles are determined to fail to that the requirement that the alert sound scenarios, such as stationary, forward or comply with the standard or if the be recognizable as a motor vehicle in reverse. Each vehicle of the same make operation does not mean that the alert and model must emit the same alert 15 49 U.S.C. Chapter 301. sound be recognizable as a vehicle with sound or set of sounds. The standard is 16 In a case involving passive occupant restraints, an internal combustion engine (ICE). required to prohibit manufacturers from the U.S. Circuit Court of Appeals for the District of providing anyone, other than the Columbia said that the agency must consider public reaction in assessing the practicability of required 19 49 U.S.C. 30118–30120. manufacturer or dealers, with a device safety equipment like an ignition interlock for seat 20 Section 2(3) of the PSEA defines ‘‘crossover designed to disable, alter, replace or belts. Pacific Legal Foundation v. Department of speed’’ as the speed at which tire noise, wind modify the alert sound or set of sounds Transportation, 593 F.2d 1338 (D.C. Cir. 1978). cert. resistance, or other factors make an EV or HV emitted from the vehicle. This language detectable by pedestrians without the aid of an alert denied, 444 U.S. 830 (1979). prohibits NHTSA from allowing 17 In a case involving passive occupant restraints, sound. The definition requires NHTSA to determine the U.S. Circuit Court of Appeals for the 6th Circuit the speed at which an alert sound is no longer said, quoting the House Report (H.R. 1776, 89th necessary. 24 PSEA Section 2(5). Cong. 2d Sess. 1966, p. 16) for the original Vehicle 21 PSEA Section 3(a). Under the PSEA, as with 25 See Keene Corp. v. United States, 508 U.S. 200, Safety Act, that ‘‘objective criteria are absolutely most legislation like it, the Secretary of 208 (1993) (stating the cannon of statutory necessary so that ‘the question of whether there is Transportation delegates responsibility for construction that ‘‘where Congress includes compliance with the standard can be answered by achieving the legislation’s objectives to the particular language in one section of a statute but objective measurement and without recourse to any appropriate Department of Transportation omits it in another . . ., it is generally presumed subjective determination.’ ’’ Chrysler v. Department Administration, in this case NHTSA. that Congress acts intentionally and purposely in of Transportation, 472 F.2d 659 (6th Cir. 1972). 22 PSEA Section 3(b). the disparate inclusion or exclusion.’’). 18 49 U.S.C. 30112 and 30165. 23 PSEA Section 3(b)(2). 26 49 U.S.C. 30102(a)(6).

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00006 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90421

manufacturers from installing an off B. Safety Problem crash factors of speed limit, vehicle switch or volume control switch that maneuver, and location were examined Comparing the Vehicle-to-Pedestrian allows the driver to turn off or turn Crash Experience of ICE Vehicles to HVs to determine the relative incidence rates down the alert sound used to meet the and EVs of HVs versus ICE vehicles and whether requirements of this standard. the odds ratio was different under Additionally, vehicle manufacturers, Crash Risk different circumstances. The analysis distributors, dealers, and motor vehicle Public safety advocacy groups have also indicated that the largest repair businesses would be prohibited raised pedestrian safety concerns differences between the involvement of from rendering the sound system regarding HVs because a vehicle using HVs and ICE vehicles in pedestrian inoperative under Section 30122 of the an electric motor may be quieter than an crashes occur with speed limits of 35 Vehicle Safety Act. A manufacturer or a ICE vehicle and may not emit the mph and lower and during certain dealer, however, is allowed to alter, sounds that non-motorists rely on for maneuvers typically executed at low replace, or modify the alert sound or set warning as vehicles approach them. speed such as making a turn, starting of sounds in order to remedy a defect or In 2009, NHTSA released the report up, and pulling into or backing out of non-compliance with the safety ‘‘Incidence of Pedestrian and Bicyclist a parking space. HVs were about 1.38 standard. Crashes by Hybrid Electric Passenger times more likely to be involved in a It is the agency’s intention that the Vehicles’’ which found that, when pedestrian crash than a vehicle with an requirements of this standard be comparing similar vehicles, 77 out of ICE during a low speed maneuver. The technology neutral. For this reason, we 8,387 total HVs reported to be in any results of the updated analysis show have chosen to establish minimum crash incident were involved in trends similar to those first reported in sound requirements for a vehicle-level pedestrian crashes, and 3,578 out of our 2009 analysis. The sample sizes of test, as opposed to a component-based 559,703 total ICE vehicles were pedestrian and bicycle crashes were re- bench test or some other type of test, to involved in similar pedestrian examined to verify that there was ensure any kind of technology used can crashes.27 The report used data sufficient statistical power in this be properly tested. collected from 12 individual states. The updated analysis. years for which data were available The agency interprets the requirement The state data set that NHTSA used to varied across different states. Generally, in the PSEA that each vehicle of the determine the pedestrian and same make and model emit the same the data used ranged from the years pedalcyclist crash rates for HVs did not sound as applying only to sound added 2000 to 2006. The ratio of pedestrian include any information about the to a vehicle for the purposes of crashes to overall crashes was 40- vision status of the pedestrians involved complying with this standard. We also percent higher for HVs than for other in the crashes, so we were unable to interpret the PSEA requirement that vehicles. In situations involving certain determine whether any of the NHTSA prohibit manufacturers from low-speed maneuvers, HVs were twice pedestrians involved in these crashes providing anyone with a means of as likely to be involved in a pedestrian were blind or visually-impaired. modifying or disabling the alert sound crash as ICE vehicles in similar and the prohibition on making required situations. While this updated analysis provides safety systems inoperative contained in In 2011 NHTSA released a second insightful comparisons of the incidence Section 30122 of the Vehicle Safety Act report ‘‘Incidence Rates of Pedestrian rates of HVs versus ICE vehicles as applying only to sound added to a And Bicyclist Crashes by Hybrid involved in pedestrian crashes, there are vehicle for the purposes of complying Electric Passenger Vehicles: An Update’’ some limitations to consider: The use of with this proposed standard. which verified these previous data from 16 states cannot be used to Many changes to a vehicle could findings 28 by adding additional years of directly estimate the national problem affect the sound produced by that state crash files as well as by increasing size; and there is still not enough data vehicle. In issuing this proposal the the number of states included in the to draw conclusions in all scenarios of agency does not wish to prevent analysis from 12 to 16, which increased interest such as for individual low- manufacturers, dealers, and repair the number of crashes included in the speed maneuvers such as making a turn, businesses from making modifications analysis. Overall, a statistical approach starting up, or in parking lots. to a vehicle such as adding a spoiler or referred to as odds ratios indicated that It has been an ongoing concern that changing the vehicle’s tires that may the odds of an HV being in either a HVs have a very small share among all have the effect of changing the sound pedestrian or bicycle crash is greater vehicles (approximately 0.5 percent). produced by the vehicle. than the odds of an ICE vehicle being in The conditional probability of HV The PSEA requires that the final rule a similar crash, 19-percent higher for pedestrian or pedalcyclist crashes is provide a phase-in period, as pedestrian crash odds and 38-percent very small if whole populations of both determined by the agency. In response higher for bicycle crash odds.29 The HV and ICE are included. Therefore, the to that requirement, full compliance sample size of HV may have an impact with the standard must be achieved for 27 R. Hanna (2009) Incidence of Pedestrian and Bicyclists Crashes by Hybrid Electric Passenger on the comparison of crash rates all vehicles manufactured on or after Vehicles, Report No. DOT HS 811 204. U.S. Dept. between HVs and ICE vehicles. For this September 1st of the calendar year of Transportation, Washington, DC. reason, NHTSA has further updated the beginning three years after the date of Available at http://www-nrd.nhtsa.dot.gov/Pubs/ comparison between HV and ICE crash publication of the final rule. This final 811204.PDf. data in order to include additional HV rule is establishing the requirement for 28 Wu, et al. (2011) Incidence Rates of Pedestrian And Bicyclist Crashes by Hybrid Electric Passenger crashes. 100-percent compliance for all light Vehicles: An Update, Report No. DOT HS 811 526. vehicles subject to the requirements of Dept. of Transportation, Washington, DC. Available for all ICE vehicles from that data set. Because this this rule produced for sale in the U.S. at http://www-nrd.nhtsa.dot.gov/Pubs/811526.pdf. proposal does not apply to HVs that always have by all manufacturers no later than 29 The incidence rates for pedestrian and their ICE turned on while moving, the agency pedalcyclist crashes involving HVs and EVs were removed the Honda Civic and the Honda Accord September 1, 2019. This requirement calculated from the State data by comparing the from the HV category and included those vehicles includes a one-year, 50-percent phase-in pedestrian and pedalcyclist crash rates for all HVs in the calculations as ICE vehicles in estimating the period beginning September 1, 2018. contained in the State data set with the crash rates incidence rate used in the benefit calculations.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00007 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90422 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

In our recent calculations 30 we used Nissan North America, Inc. (Nissan) maneuver data in our crash analysis the latest State data available up to 2011 stated that NHTSA should take into include crashes in which the driver was from the same 16 states, in which the account the fact that the ‘‘making a making a turn or backing and may have sample sizes of HV vehicles of all turn’’ and ‘‘backing’’ maneuvers, which had an obstructed view of the crashes are increased to 68,950 (with constitute a majority of the low speed pedestrian. Because backing crashes are 420 pedestrian crashes for all hybrid maneuvers examined in the agency’s addressed by our recent final rule to vehicle models). The earlier research crash analysis, are maneuvers during increase the field of view requirements obtained the pedestrian crash odds which it is difficult for drivers to detect of FMVSS No. 111, Rear Visibility, we ratios of HV versus ICE vehicle with pedestrians. American Honda Motor Co. have adjusted our benefits calculation much smaller sample sizes. The new (Honda) stated that NHTSA should for this rulemaking to remove those analysis showed that after the Honda examine whether there is a significant crashes addressed by FMVSS No. 111. Civic and Accord models are moved difference between HEV/EV pedestrian Also, the fact that the driver’s view may from the hybrid category to the ICE crashes and ICE pedestrian crashes for have been obstructed supports the need category the odds ratio of HV vs. ICE vehicles starting from stationary. to establish minimum sound pedestrian crashes for all speeds is 1.21 Advocates stated that elevated crash requirements for HVs and EVs so that and the odds ratio for slower speed rates between EVs/HEVs and pedestrians can detect when those maneuvers is 1.52. This analysis also pedestrians and pedalcyclists, concerns vehicles are pulling out or approaching shows that the odds ratio of HV vs. ICE of blind advocacy groups, and the in situations in which the pedestrian is pedalcyclist crashes is 1.58 for all international attention focused on the potentially obscured from the driver’s speeds including all speed maneuvers, issue support the conclusion that view. minimum sound requirements for EVs and 1.50 for slower maneuvers. Fatalities In the NPRM, the agency asked for and HEVs will reduce the rate of comments on whether the differences in pedestrian crashes involving these The Fatality Analysis Reporting pedestrian crash rates between HV and vehicles. The Insurance Institute for System (FARS) contains a census of all ICE vehicles are solely due to Highway Safety stated that, according to traffic fatalities. HVs and EVs that struck pedestrians’ inability to detect these research from the Highway Data Loss and killed a pedestrian were identified vehicles based on sound, or whether Institute (HDLI), hybrid vehicles where using the Vehicle Identification there may be other factors that we have 17.2 percent more likely to cause Numbers (VINs) contained in the 2001 not identified that affect the difference injuries to pedestrians than their ICE through 2009 FARS files. During this in crash rates. vehicle counterparts. period, there were 53 pedestrian fatalities attributed to crashes involving Ideally, in order to determine whether Agency Response to Comments this lack of sound is causing accidents, 47 HVs and three EVs. Almost all of NHTSA would have compared accident After review of the comments these fatalities (47 of the 53) involved rates for HVs and EVs with and without received on the NPRM, we utilized a vehicles that were identified as sound. However, there have not been multivariate logistic regression model to passenger vehicles. In 2008, there were enough HVs and EVs with sound for a examine whether other variables besides 10 HVs or EVs that struck and killed 10 long enough period of data to be able type of powertrain in the State Data pedestrians, and in 2009, there were 11 reasonably conduct this analysis. System contributed to increased risk of HVs or EVs that struck and killed 11 NHTSA has also been unable to directly pedestrian collisions. In addition, we pedestrians. measure the pedestrian and pedalcyclist utilized the calculated odds ratio to However, these fatalities are not crash rates per mile travelled for HVs compare HVs and ICEs using a case- included in the target population for and EVs to the rates for ICEs because the control analysis. The variables that analysis under this rulemaking for two Agency does not have data on VMT for NHTSA examined in the regression are: reasons. The first is that pedestrian HVs and EVs. Therefore, we have Whether the vehicle was an HV or ICE; fatalities are not as likely to occur at low instead used the number of other types whether the vehicle was involved in a speeds for which the rate of HV of crashes vehicles are involved in and low-speed maneuver at the time of the pedestrian collisions is significantly using that as a proxy for VMT. While crash; city size; driver age; vehicle age; higher than collisions between ICE this is a standard technique in analyzing and calendar year. The results of the vehicles and pedestrians. Today’s final regression analysis show that an HV crash risk, it does raise the possibility rule establishes minimum sound may have 1.18 times higher likelihood that there may be other explanations requirements for hybrid and electric of hitting a pedestrian than an ICE after than the lack of sound for hybrids vehicles operating at speeds up to 30 accounting for these other confounding having higher-than-average rates of km/h (18.6 mph). A majority of risk factors included in the State Data pedestrian and pedalcyclist crashes pedestrian fatalities occur when the System. NHTSA believes that our case- relative to other crashes. vehicle involved in the collision is not Various comments noted that the control analysis, the results of our travelling at a low speed. Overall, 67 agency should consider the possibility multivariate logistic regression, and the percent of the pedestrian fatalities that factors other than sound will have results of HDLI’s research show that involving HVs or EVs and with known an impact on the difference in crash there is a difference in crash rates speed limits occurred at a speed limit rates between HVs and ICE vehicles. between HVs and ICE vehicles that is above 35 mph.31 For all pedestrian Commenters stated that driver attributable to sound. We note that we fatalities with known speed limits, 62 characteristics and higher rates of were unable to calculate a statistically percent occurred at a speed limit above exposure to pedestrians were factors significant difference in crash rates 35 mph and 61 percent of those that could contribute to the higher rate between HVs and ICE vehicles for of pedestrian crashes among HVs when pedestrian crashes when the vehicle 31 For those pedestrian fatalities that occurred on compared to ICE vehicles. was starting from a stopped position roads with a posted speed limit of 35 mph or less, because of the small number of crashes we do not have any data on actual travel speed of the vehicles involved. Therefore, we are not able to 30 Wu, J., 2015, ‘‘Updated Analysis of Pedestrian involving HVs in the State Data System. tell if the vehicles involved were travelling at a and Pedalcyclist Crashes of Hybrid Vehicles with We have considered the fact that speed at which they would be required to meet the Larger Samples and Multiple Risk Factors.’’ many of the crashes in the low-speed requirements of the final rule.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00008 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90423

involving passenger vehicles occurred at avoided as a result of preventing and direction of travel. For example, a a speed limit above 35 mph.32 The goal vehicle-pedestrian crashes, NHTSA blind or visually-impaired person of this rule is to prevent injuries to believes it is important to note another moving through a parking lot can hear pedestrians that result from pedestrians dimension of safety that should be taken and avoid vehicles entering or exiting being unable to hear nearby hybrid and into account with respect to pedestrians the lot or looking for parking spaces; a electric vehicles operating at low who are blind or visually-impaired. blind person walking through a speeds. At speeds of 35 mph and above, Pedestrians who are blind or visually- neighborhood can hear when a neighbor at which a majority of fatal crashes impaired need to be able to travel is backing out of a driveway. The involving pedestrians occur, it is very independently and safely throughout vehicle sound also indicates to a blind unlikely that lack of sound is the cause their communities without fear and risk or visually-impaired pedestrian whether as the sound levels produced by hybrid of injury, both as a result of collisions a vehicle is making a turn, and if so, in and electric vehicles at those speeds are with motor vehicles and as a result of which direction. The sound of the same as the sound levels produced other adverse events in the individual vehicles also allows the by ICE vehicles. Establishing minimum environments they must negotiate. To a blind traveler to detect and react to sound requirements for hybrid and far greater extent than is the case for unusual or unexpected vehicle electric vehicles operating at speeds up sighted people, vehicle sounds help to movement. The sound of a vehicle that to 30 km/h is expected to prevent injury define a blind or visually-impaired has an activated starting system but is crashes but not necessarily have an person’s environment and contribute to stationary (usually referred to as impact on those crashes involving that person’s ability to negotiate through ‘‘idling’’ for vehicles with internal pedestrian fatalities, based on existing his/her environment in a variety of combustion engines) alerts the blind or data. situations.33 visually-impaired traveler to the fact The second reason is that the rate of The modern white cane and the that the vehicle is not simply parked pedestrian fatalities per registered techniques for its use help the user to and that it may move at any moment. If vehicle for HVs and EVs is not larger navigate and allow sighted people to a blind person is approaching a (and is in fact smaller) than that for ICE recognize that a person is blind or driveway and notes a vehicle that is vehicles. Using 2008 data, the fatality visually-impaired. Today, the stationary but running he or she will rate for pedestrians in crashes with HVs ‘‘structured discovery’’ method of wait for the vehicle to pull out, or for and EVs is 0.85 fatalities per 100,000 teaching independent travel for an indication that it will not, for registered vehicles, and the visually-impaired people emphasizes example by noting that the vehicle corresponding rate for ICE vehicles is learning to use information provided by remains stationary for some time, 1.57 per 100,000 vehicles. the white cane, traffic sounds, and other indicating that the driver has no There also could be fatalities cues in the environment to travel immediate plans to move. involving HVs and EVs that occur in anywhere safely and independently, In the NPRM, the agency described non-traffic crashes in places such as whether the individual has previously how the acoustic cues provided by driveways and parking lots. However, a visited the place or not. vehicles help blind pedestrians discern comprehensive search for HVs and EVs Whether a blind or visually-impaired changes in the road-way, determine involved in pedestrian fatalities could person uses a white cane or guide dog, whether an intersection has a traffic not be undertaken because NHTSA’s the primary purpose of both travel tools control device, and navigate Not in Traffic Surveillance (NiTS) is to help the blind traveler identify intersections with unusual system does not provide VINs, and a and/or avoid obstacles in his or her path characteristics such as three-way search for model names that indicate using the sense of touch. The remaining intersections or roundabouts. The hybrid or electric vehicles did not information needed by a blind or sounds made by traffic including the identify any crashes involving visually-impaired person to safely and sounds of idling vehicles allow blind pedestrian fatalities. independently travel is provided pedestrians to determine when it is safe primarily through the sense of hearing. Low-Speed Vehicles to cross the street and maintain a When traveling with a white cane or straight travel path while walking NHTSA has no data on pedestrian or guide dog, the primary sound cue used through the intersection. pedalcyclist crash rates for low-speed by blind pedestrians is the sound of Using the white cane or guide dog and vehicles due to the low rate of sales of vehicle traffic, which serves two the sound of traffic, people who are these vehicles as a percentage of the purposes: navigation and collision blind or visually-impaired have been light vehicle fleet. NHTSA also has not avoidance. Navigation involves not only able to navigate safely and found any examples of crashes ascertaining the proper time to enter a independently for decades. Blind and involving LSVs and pedestrians or crosswalk and maintain a straight visually-impaired people travel to pedalcyclists that appear to be caused course through an intersection while school, the workplace, and throughout by the lack of sound in LSVs. However, crossing, but also the recognition of their communities to conduct the daily we assume that the safety problem with roadways and their traffic patterns and functions of life primarily by walking these vehicles will be similar to that for their relationship to sidewalks and other and using public transportation. Safe HVs based on the acoustic profile of travel ways a blind or visually-impaired and independent pedestrian travel is these vehicles. person might use. essential for blind or visually-impaired Sound emitted by individual vehicles, individuals to obtain and maintain Need for Independent Mobility of as opposed to the general sound of employment, acquire an education, and People Who Are Visually-Impaired moving traffic, is critical. The sound of fully participate in community life. In addition to addressing the safety individual vehicles helps to alert blind Short of constantly traveling with a need in the traditional sense of injuries travelers to the vehicle’s location, speed, human companion, a blind or visually- impaired pedestrian simply cannot 32 Data particularly tied to other speeds, such as 33 National Federation of the Blind (2011) How ensure his or her own safety or navigate 20 mph, is not available because of the structure of People Who are Blind Use Sound for Independent effectively without traffic sound. To the the databases used, i.e., the relevant data variable Travel, memorandum to the docket, NHTSA–2011– is whether the speed limit was above or below 35 0148–0028, Washington, DC. That memorandum is extent that there are more and more HVs mph at the crash location. the source for this information. and EVs on the road that are hard to

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00009 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90424 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

detect, people who are blind or visually- on Quieter Cars and the Safety of Blind unstructured conversations with blind impaired will lose a key means—the Pedestrians, A Report to Congress.’’ 35 pedestrians and orientation and sound of traffic—by which they The report to Congress briefly discussed mobility specialists. Scenarios were determine when it is safe to cross the quieter vehicle safety issue, how defined by combining pedestrian streets, but also by which they orient NHTSA’s research plan would address vehicle environments, vehicle type, themselves and navigate safely the issue, and the status of the agency’s vehicle maneuver/speed/operation, and throughout their daily lives, avoiding implementation of that plan. considerations of ambient sound level. dangers other than automobiles. In 2010 through 2014 the agency The operating scenarios identified in C. Research on Vehicle Emitted Sounds continued relevant quiet car research as Phase 1 were: Vehicle approaching at and Detectability briefly discussed below. low speed; vehicle backing out (as if Phase 1 Research coming out of a driveway); vehicle Early Research on Quiet Vehicles and travelling in parallel and slowing (like Public Meeting In April 2010, NHTSA issued a report a vehicle that is about to make a turn); NHTSA began collaborating with a that began addressing the tasks listed in vehicle accelerating from a stop; and a working group within the Society of the research plan. This report, titled vehicle that is stationary. Automotive Engineers International ‘‘Quieter Cars and the Safety of Blind Pedestrians: Phase I,’’ documents the In Phase 1, NHTSA also compared the (SAE) in August 2007 to identify auditory detectability of HVs and ICE effective ways to address the safety overall sound levels and general spectral content for a selection of ICE vehicles by pedestrians who are legally issue of quiet hybrid and electric blind. Forty-eight independent travelers, vehicles. This working group included vehicles and HVs in different operating conditions, evaluates vehicle with self-reported normal hearing, representatives from the Alliance of 37 detectability for two background noise listened to binaural audio recordings Automobile Manufacturers, Global of two HVs and two ICE vehicles in Automakers, the visually impaired levels, and considers the viability of countermeasure concepts categorized as three operating conditions, and two community and NHTSA. different ambient sound levels. The On June 23, 2008, NHTSA held a vehicle-based, infrastructure-based, and systems requiring vehicle-pedestrian operating conditions included a vehicle: public meeting to bring together Approaching at a constant speed (6 government policymakers, stakeholders communications.36 mph); backing out at 5 mph; and from the visually impaired community, The results show that the overall slowing from 20 to 10 mph (as if to turn industry representatives, and public sound levels for the HVs tested are right). The ambient sound levels were a interest groups to discuss the technical noticeably lower at low speeds than for quiet rural (31.2 dB(A)) and a and safety policy issues associated with the ICE vehicles tested. Overall, study moderately noisy suburban ambient hybrid vehicles, electric vehicles, and participants were able to detect any (49.8 dB(A)). Overall, participants took quiet internal combustion engine (ICE) vehicle sooner in the low ambient noise longer to detect the two HVs tested vehicles, and the risks they present to condition. ICE vehicles tested were (operated in electric mode), except for visually impaired pedestrians. After this detected sooner than their HV the slowing maneuver. Vehicle type, public meeting, NHTSA issued a counterpart vehicles except for the test ambient level, and operating condition research plan to investigate hybrid and scenario in which the target vehicle was electric vehicles and pedestrian safety.34 slowing down. In this scenario, HVs had a significant effect on response The objectives of the research plan were were detected sooner because of the time. to identify critical safety scenarios for distinctive sound emitted by the Table 5 shows the time-to-vehicle visually impaired pedestrians, identify regenerative braking system on the HVs. arrival at the time of detection by requirements for blind pedestrians’ safe Response time to detect a target vehicle vehicle type, and ambient condition. mobility (emphasizing acoustic cues varies by vehicle operating condition, Considering all three independent from vehicles and ambient conditions), ambient sound level, and vehicle type variables, there was a main effect of identify potential countermeasures, and (i.e., ICE vehicle versus HV or EV vehicle, vehicle maneuver, and ambient describe the countermeasures’ mode). sound level. Similarly, there were advantages and disadvantages. As part of Phase 1 research, NHTSA interaction effects between vehicle type In 2009 NHTSA issued the report sought to identify operating scenarios and ambient, vehicle type and ‘‘Incidence of Pedestrian and Bicyclist necessary for the safety of visually maneuver, ambient and vehicle Crashes by Hybrid Electric Passenger impaired pedestrians. The researchers maneuver, and a three way interaction Vehicles,’’ discussed in Section II.B of identified these scenarios based on between ambient, vehicle type and this notice, and a report titled ‘‘Research crash data, literature reviews, and vehicle maneuver.

TABLE 5—AVERAGE TIME-TO-VEHICLE ARRIVAL BY SCENARIO, VEHICLE TYPE, AND AMBIENT SOUND

Low ambient High ambient Scenario ICE ICE HVs vehicles HVs vehicles

Approaching at 6 mph ...... 4.8 6.2 3.3 5.5 Backing out at 5 mph ...... 3.7 5.2 2.0 3.5

34 A copy of the research plan is available at Technical%20Publications/2010/ Crash%20Avoidance/Technical%20Publications/ www.regulations.gov (Docket No. NHTSA–2008– RptToCongress091709.pdf. 2010/811304rev.pdf. 0108–0025). 36 Garay-Vega, et al. (2010) Quieter Cars and the 37 Binaural recordings reproduce the acoustic 35 Research on Quieter Cars and the Safety of Safety of Blind Pedestrians: Phase I, Report No. characteristics of the sound similar to how a human Blind Pedestrians, A Report to Congress. U.S. Dept. DOT HS 811 304, U.S. Dept. of Transportation, perceives it. Binaural recordings reproduce a more realistic three dimensional sensation than of Transportation, Washington, DC, October 2009, Washington, DC. Available at http:// available at http://www.nhtsa.gov/DOT/NHTSA/ conventional stereo and are intended for playback www.nhtsa.gov/DOT/NHTSA/NVS/ NVS/Crash%20Avoidance/ through headphones, rather than loudspeakers.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00010 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90425

TABLE 5—AVERAGE TIME-TO-VEHICLE ARRIVAL BY SCENARIO, VEHICLE TYPE, AND AMBIENT SOUND—Continued

Low ambient High ambient Scenario ICE ICE HVs vehicles HVs vehicles

Slowing from 20 to 10 mph ...... 2.5 1.3 2.3 1.1

The Phase 1 research showed that • Provide for detection of a vehicle in parameters and criteria for sounds to be HVs were more difficult for pedestrians residential, commercial, and other detectable and recognizable as a motor to detect by hearing than ICE vehicles. suburban and urban environments in vehicle. The Phase 1 research report also which blind pedestrians would expect VRTC Acoustic Measurements discussed various countermeasures to to be able to navigate using acoustic mitigate pedestrian safety risks cues. Note: Human factors tests for The primary focus of Phase 3 research associated with quiet vehicles. The Phase 2 were conducted in an ambient conducted by VRTC was to develop an Phase 1 report also concluded that a of approximately 58–61 dB(A). objective and repeatable test procedure vehicle-based audible alert signal was As part of the Phase 2 research, Volpe to measure vehicle-emitted sound. This the countermeasure that both provided conducted a human factors study to work consisted mainly of evaluation of all the necessary information to blind compare the auditory detectability of the new SAE J2889–1, Measurement of pedestrians to make safe travel potential sounds for hybrid and electric Minimum Noise Emitted by Road decisions and produced benefits for vehicles operating at a low speed and Vehicles, test method, and several other pedestrians and for pedalcyclists. how those sounds compared to an ICE variations used to test operating control vehicle. The human factors conditions that were not included in Phase 2 Research testing in Phase 2 suggested that SAE J2889–1, and development of a In October 2011 NHTSA released a synthetic sounds resembling an ICE practical test procedure for collecting second report examining issues produce similar detection distances as test track acoustic data from HVs, EVs involving hybrid and electric vehicles actual ICE vehicles. In some instances, and ICE vehicles. The data collected and blind pedestrian safety titled the results indicated that synthetic was then evaluated to begin establishing ‘‘Quieter Cars and the Safety of Blind sounds designed according to potential performance criteria. The draft Pedestrians, Phase 2: Development of psychoacoustic principles can produce version of SAE J2889–1 used by VTRC Potential Specifications for Vehicle double the detection distances relative included recommended procedures for Countermeasure Sounds.’’ 38 The Phase to the reference vehicle. The results also measuring minimum sound pressure 2 research developed various methods suggested that synthetic sounds that levels of vehicle-emitted sound but did to specify a sound to be used as a contain only the fundamental not include any recommended vehicle-based audible alert signal that combustion noise are relatively performance requirements for minimum could be used to provide information at ineffective. None of the analyses found levels of vehicle-emitted sound. SAE least equivalent to the cues provided by a significant effect of vision ability.39 J2889–1 was still in draft form at the ICE vehicles, including speed change, Participants who were legally blind, on start of the research, but the version and evaluated sounds using human average, were no better or worse than published in September 2011 was not factors testing to examine whether the sighted participants in detecting the significantly different from the draft. sounds could be detected and approach sounds. The research was conducted using recognized as vehicle sounds. This three HVs, one EV, and four ICE Phase 3 Research research used acoustic data acquired vehicles. The vehicles were used to from a sample of ten ICE vehicles to In order to develop possible test gather sample data on the difference in examine the sound levels at which procedures and requirements for an sound pressure levels between ICE synthetic vehicle sounds used could be FMVSS proposing to establish sounds and EV or HV sounds. VRTC set, and used psychoacoustic models to minimum acoustic requirements for also gathered data to determine how examine issues of detectability and hybrid and electric vehicles, NHTSA synthetic vehicle sounds emitted from masking of ICE-like sounds and initiated a third phase of research to speakers projected around the vehicle, alternative sounds, and also included a develop an objective, repeatable test as referred to as the directivity of the human factors study to examine the procedure and objective specifications sound, and sound quality levels. Some detectability of synthetic sounds. for minimum sound requirements. of the hybrid and electric vehicles were The methods for specifying sounds NHTSA’s Vehicle Research and Test tested with multiple alert sounds. Some discussed in the Phase 2 final report Center (VRTC), as part of its effort to of the hybrid and electric vehicles were assumed that the vehicle acoustic develop a test procedure, conducted also tested with no alert sound at all, to countermeasure should: acoustic measurements and recordings examine the difference between the • Provide information at least of several HVs and EVs and those sound pressure level produced by equivalent to that provided by ICE vehicle’s ICE pair vehicles.40 Volpe used hybrid and electric vehicles and ICE vehicles, including speed change; and these recordings as well as data from the vehicles. Phase 1 and Phase 2 research to identify One of the purposes of the Phase 3 38 Garay-Vega, et al. (2011) Quieter Cars and the acoustic measurements was to gather Safety of Blind Pedestrians, Phase 2: Development 39 All participants were required to wear a additional data on the difference in of Potential Specifications for Vehicle blindfold during the study. sound levels between ICE vehicles and Countermeasure Sounds, Report No. DOT HS 811 40 Evans and Harris. (2012) Quieter Vehicle EVs and HVs operating in electric mode. 496. Dept. of Transportation, Washington, DC. Performance Test Development Research Report, Available at http://www.nhtsa.gov/DOT/NHTSA/ U.S. Dept. of Transportation, Washington, DC. For the pass-by tests at 10 km/h in NVS/Crash%20Avoidance/ Available at www.regulations.gov, Document ID: Phase 3, the ICE vehicles were between Technical%20Publications/2011/811496.pdf. NHTSA–2011–0148–0047. 6.2 and 8.5 dB(A) louder than the EV/

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00011 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90426 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

HVs without added sound. At 20 km/h vehicles. At 30 km/h the sound levels of vehicles have considerable overlap. the difference between the HV/EVs and the HV/EVs approached the levels of the Table 6 shows the results of HV/EV ICE vehicles varied, but the average ICE vehicles and the individual vehicles with no sound alert as delta was 3.5 dB(A) louder for the ICE measurements for the two types of compared to their ICE counterparts.

TABLE 6—PASS-BY SOUND LEVEL FOR HV/EV VEHICLES WITHOUT ALERT SOUND VERSUS COUNTERPART ICE VEHICLES

HV/EV Sound ICE Sound ICE minus Manufacturer Speed, km/h Level, dB Level, dB HEV/EV, dB

Nissan ...... 10 50.5 56.6 6.1 20 60.0 62.3 2.3 30 66.5 68.1 1.6 Prototype Vehicle G ...... 10 51.4 59.9 8.5 20 60.5 63.1 2.6 30 67.0 67.5 0.5 Prototype Vehicle H ...... 10 51.2 59.7 8.5 20 59.3 64.5 5.2 30 65.3 69.2 3.9 Average ...... 10 51.0 58.7 7.7 20 59.9 63.3 3.4 30 66.3 68.3 2.0

The measurements from the startup pertinent information about the signal contains detectable components and stationary but active scenarios were perceived loudness and detectability of over a wide frequency range. The used to measure the directivity of the a sound. Two approaches were used to research also explored how tones and vehicles’ sound. The purpose of identify potential detectability broadband content could enhance the measuring the directivity pattern of the specifications for alert sounds to be detectability of synthetic alert sounds. vehicles was to compare the directivity included in the NPRM: (1) Sound The report used acoustic data for pattern of ICE vehicles to those hybrid parameters based on a loudness model directivity to estimate minimum sound and electric vehicles equipped with a and detection distances and (2) sound levels for ‘reverse’ or ‘backing’ speaker system. For the ICE vehicles, parameters based on the sound of ICE maneuvers. Volpe then used the results the sound pressure level behind the vehicles. of this analysis of the detectability of vehicle was 6 to 10 dB lower than that Volpe’s work in developing the sound sounds as estimated by psychoacoustic directly in front of the vehicle. For the specifications based on a loudness models to make recommendations for hybrid and electric vehicles with a model and detection distances was potential minimum sound levels for the speaker system, the sound level behind guided by several aspects of the NPRM. the vehicle was 12 to 15 dB lower agency’s Phase 1 and Phase 2 research. In addition to using psychoacoustic behind the vehicle. There was a Volpe analyzed the acoustic data of the models to develop recommendations for systematic difference from left to right sounds used in the human factors minimum sound specifications, Volpe for some vehicles, particularly with an research in Phase 2 from a created a set of minimum sound artificial sound. psychoacoustic perspective to specifications based on the sound Volpe Acoustic Analysis determine the loudness of the sounds produced by ICE vehicles. Volpe and whether the sounds would be considered multiple minimum sound As another part of the Phase 3 detectable in several different ambient specifications in an attempt to derive at research, Volpe conducted an analysis environments. Because the response of the most optimal approach for defining of existing acoustic data and data the study participants in the human sound specification requirements in collected during the previously factors experimentation in Phase 2 order to provide recommendations for a mentioned VTRC testing to develop varied significantly due to variations in variety of sound specifications for recommendations for performance the ambient,42 Volpe determined that NHTSA to seek comment on in the requirements for minimum levels of any analysis of sounds using a loudness NPRM. Volpe created the specification vehicle emitted sound to be proposed in model should use a synthetic ambient based on the sound produced by ICE the NPRM. This work consisted of that did not vary with respect to the vehicles (using data captured during examining the frequency ranges, frequency profile or overall sound Volpe’s Phase 2 research) and minimum sound levels for selected one- pressure level. Volpe used a synthetic recordings of vehicles provided by third octave bands, and requirements for ambient sound with the loudness model automobile manufacturers. Volpe broadband noise and tones as possible during Phase 3 in developing the aggregated this data to create minimum criteria for setting minimum acoustic specifications based on the requirements for vehicle-emitted sound. specifications contained in the NPRM. mean sound levels of ICE vehicles and Evaluations were conducted using a This research showed that the mean sound levels of ICE vehicles loudness model 41 to determine when pedestrians’ ability to detect synthetic minus one standard deviation. the sounds might be detectable in a sounds would be maximized if the alert given ambient. Of the several different Agency Research and Analysis loudness models examined by Volpe, 42 Garay-Vega, et al. (2011) Quieter Cars and the Conducted Since the NPRM Safety of Blind Pedestrians, Phase 2: Development Moore’s Loudness provided the most of Potential Specifications for Vehicle After the NPRM was issued, NHTSA Countermeasure Sounds, Report No. DOT HS 811 conducted research to examine 41 Loudness models are computer simulations 496. Dept. of Transportation, Washington, DC. used to estimate the minimum sound levels needed Available at http://www.nhtsa.gov/DOT/NHTSA/ additional aspects of minimum sound for alert sounds to be detectable in the presence of NVS/Crash%20Avoidance/ requirements for hybrid and electric ambient noise. Technical%20Publications/2011/811496.pdf. vehicles. The research involved human

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00012 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90427

factors testing and acoustic modeling to identified based on a psychoacoustic trial to the instant the participant examine the detectability of sounds loudness modeling approach and safe presses the push-button as an indication with different acoustic characteristics. detection distances.43 After the NPRM he/she detected the target signal. The The research also involved acoustic was published, the agency conducted a time-to-vehicle arrival is the elapsed measurement of heavy-duty vehicles study to quantify the differences time, in seconds, from first detection of and motorcycles, analysis of indoor between predicted detection levels of a target signal to the instant the vehicle testing conducted by Transport Canada, vehicle sounds in the presence of an passes the pedestrian location. The and additional light vehicle testing to ambient (as indicated by the loudness detection distance is the separation refine the test procedure proposed in model) and the actual responses by between the vehicle and the pedestrian the NPRM. The research is documented participants listening to these vehicle location at the moment of detection. The in multiple separate research reports sounds through headphones. This was detection distance can be computed and is summarized below. In some done in order to evaluate the accuracy from the time-to-vehicle arrival and cases, as identified below, more details of the psychoacoustic model in vehicle speed. Signals meeting the of the research are provided in the predicting when sounds would be minimum sound levels, computed appropriate sub-sections of Section III of detected. The study also explored the according to the approach described in this preamble. In those cases, the agency effect of different factors such as the the NPRM, are expected to be detectable discusses the important aspects of the number of bands at threshold, adjacent at least 2.0 seconds or 5 meters away research that were utilized to make and non-adjacent bands, and signal type decisions finalized in this rule. (e.g., pure tones, bands of noise).44 In (for a vehicle approaching at 10 km/h). addition to the human factors study, Table 7 shows the time-to-vehicle Human Factors Research and Acoustic Volpe also conducted an analysis of arrival and detection distances for the Modeling acoustic data in order to predict the signals examined in this study. The In the NPRM, NHTSA proposed probability that a sound would be signals used in the study included minimum sound pressure levels for a detected in different ambients as the sounds developed by Volpe to test specific set of one-third octave bands number of one-third octave bands different hypotheses involving the that included low frequency bands (315, making up the sound changes. detection model, recordings of 400, and 500 Hz) and high-frequency The key performance metrics for the prototype synthetic sounds provided by bands (2000, 2500, 3150, 4000, and 5000 human factors study were the response vehicle manufacturers, and a recording Hz) for various operating conditions. time and associated time-to-vehicle of an ICE vehicle. The ‘‘Source’’ column These proposed specifications for arrival. Response time is the elapsed in Table 7 describes the origin of each minimum sound pressure levels were time, in seconds, from the start of the sound.

TABLE 7—SOUND STIMULI TESTED

Vehicle dis- Signal ID Significant component fre- Levels, dB(A) Source Comment Time-to-vehi- tance at detec- quencies, Hz cle arrival, s tion, m

3 ...... 315, 400, 500, 630, 2000, Threshold ...... Simulation ...... Tone @315 Hz, TNR 9 4.9 13.6 2500, 3150, 4000, 5000. dB. 6 ...... 315, 400, 500, 630, 2000, Threshold ...... Simulation ...... Tone @630 Hz, TNR 9 4.3 11.9 2500, 3150, 4000, 5000. dB. 9 ...... 315, 400, 500, 630, 2000, Threshold ...... Simulation ...... Tone @2500 Hz, TNR 4.5 12.5 2500, 3150, 4000, 5000. 9 dB. 10 ...... 315, 400, 500, 630, 2000, Threshold ...... Simulation ...... NNPRM + 630 Hz ...... 4.4 12.2 2500, 3150, 4000, 5000. 11 ...... 315 ...... Threshold ...... Simulation ...... Single Noise Band ...... 2.3 6.4 12 ...... 630 ...... Threshold ...... Simulation ...... Single Noise Band ...... 2.9 8.1 13 ...... 2500 ...... Threshold ...... Simulation ...... Single Noise Band ...... 2 5.6 14 ...... 315, 400, 500, 2000, 2500, Threshold ...... Simulation ...... NPRM ...... 4.3 11.9 3150, 4000, 5000. 15 ...... 50 to 10,000 ...... Threshold ...... Simulation ...... Noise in all Bands ...... 4.6 12.8 17 ...... 315, 400, 500 ...... 46, 54, 48 ...... Prototype Re- ASG as Recorded (No 5.8 16.1 cording. calibration). 18 ...... 315, 400, 500, 2000, 2500, Threshold ...... Prototype Re- ASN (Calibrated to 4.5 12.5 3150, 4000, 5000. cording. match NPRM). 19 ...... 2500 ...... 56 ...... Prototype Re- ASN as Recorded (No 5.8 16.1 cording. calibration). 20 ...... 315, 400, 500, 2000, 2500, Threshold ...... Prototype Re- ASV Sound4 (Cali- 6.7 18.6 3150, 4000, 5000. cording. brated to match NPRM). 23 ...... 4000, 5000, 6300, 8000, 10000 37, 36, 34, 32, ICE Recording ASF ICE (No Calibra- 3.1 8.6 31. tion). 25 ...... 315, 400, 500 ...... Threshold ...... Simulation ...... Low Frequency Noise .. 4.2 11.7 26 ...... 315, 630, 2000, 5000 ...... Threshold ...... Simulation ...... Non-adjacent Noise ..... 4.5 12.5 27 ...... 630, 800, 1000, 1250, 1600 ..... Threshold ...... Simulation ...... Mid-frequency Noise .... 3.7 10.3 28 ...... 800, 2500 ...... 39, 45 ...... Simulation ...... 1 below threshold, 1 at 2.2 6.1 threshold. 29 ...... 800, 2500 ...... 45, 39 ...... Simulation ...... both below threshold ... 1.4 3.9

43 Hastings, et al. (2012). Research on Minimum 44 Hastings, et al. Detectability of Alert Signals for and Human Subjects Experiment. (2015) Sound Specification for Hybrid and Electric Hybrid and Electric Vehicles: Acoustic Modeling Washington, DC: DOT/NHTSA. Vehicles. Docket NHTSA–2011–0148–0048.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00013 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90428 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

TABLE 7—SOUND STIMULI TESTED—Continued

Vehicle dis- Signal ID Significant component fre- Levels, dB(A) Source Comment Time-to-vehi- tance at detec- quencies, Hz cle arrival, s tion, m

30 ...... 800, 2500 ...... 50, 50 ...... Simulation ...... 1 ∼ threshold, 1 above 3.6 10.0 threshold. 31 ...... 2000, 2500, 3150, 4000, 5000 Threshold ...... Simulation ...... High Frequency Noise 3.2 8.9 32 ...... 315 ...... Threshold ...... Simulation ...... Pure Tone ...... 3.1 8.6 33 ...... 630 ...... Threshold ...... Simulation ...... Pure Tone ...... 2.9 8.1 34 ...... 2500 ...... Threshold ...... Simulation ...... Pure Tone ...... 2.4 6.7

The data showed that all signals octave band levels for detection used in reconcile the predicted detection values tested in the study exceeded the 2.0- the NPRM, were detected 5.8 seconds in the model and the performance of the second detection criterion except for away. In general, signals with a pure participants in the experiment. signal 29, which was detected 1.4 tone (signals 32, 33, 34) were detected 45 In order to ensure that the model was seconds before pass-by. Exceeding the sooner than signals with a single band as predictive of real-world experience as 2.0-second detection criterion was of noise at the same frequency (signals possible, that is, in order to obtain the expected for signals with content in 11, 12, 13). For example, the average best agreement between modeled more than one one-third octave band, time-to-vehicle arrival was 3.1 seconds detection thresholds and those of the since the modeled thresholds were for a pure tone at 315 Hz and 2.3 participants, and also to correct for based on a signal with content in a seconds for a single band of noise at the single band. Content in multiple one- same frequency. A statistical analysis frequency sensitivity differences, Volpe third octave bands could increase the also found that the interaction of sound did a series of linear regressions using time-to-vehicle arrival if subjects type (tones or noise) and frequency was different loudness metrics. The best aggregated the energy across bands or if significant. agreement between modeled and actual they utilized a ‘best’ single band The study results indicated that, participant detection times occurred strategy. That is, with more one-third except for frequency sensitivity for high when a detection threshold of 0.079 octave bands, the signal can be more frequency components, the modeling sones 46 per ERB was used 47 (see Figure easily detected either because it is approach for determining detection 1). The R-squared value achieved for stronger overall or because, given the thresholds was conservative, meaning this model was 0.72, indicating that the many possible random factors that that the study participants were able to model performs well on average could affect detectability, more detect sounds sooner than predicted by although, as anticipated, outcomes are components creates a greater probability the model. In order to correct for not always exactly the same due to that at least one band will be easier to frequency sensitivity differences and to random variation and other differences detect. develop the best agreement between between the model predictions and An ICE vehicle (signal 23), without modeled detection thresholds and those participant performance. Thus, the calibration to minimum one-third of the participants so that the minimum agency chose to use the detection octave band levels for detection used in one-third octave band levels for threshold of 0.079 sones per ERB in the the NPRM, was detected 3.1 seconds detection in the final rule more closely Moore’s model as the basis for deriving away on average. Two prototype alert align with pedestrians’ ability to detect the revised minimum levels for each of signals (signals 17, 19), without sounds in the real world, Volpe the one-third octave bands in the final calibration to minimum one-third performed a linear regression to rule.

45 Signal 29 had two components, and the levels loudness (in sones) and partial loudness (in sones 47 Hastings A.; and McInnis, C. ‘‘Detectability of were set below the minimum detection thresholds. per equivalent rectangular bandwidth or ‘‘ERB’’) Alert Signals for Hybrid and Electric Vehicles: 46 Sone is a unit of subjective loudness on a linear parameters as a basis for determining thresholds, Acoustic Modeling and Human Subjects scale. The Moore’s Loudness model used by the i.e., minimum sound levels, required for vehicle Experiment’’ Docket NHTSA–2011–0148. agency in the NPRM and this final rule utilizes detection. Washington, DC: DOT/NHTSA.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00014 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90429

The agency also conducted an while other samples are dominated by conditions. For signals with content in analysis of acoustic recordings to high frequency content or have a mix of 1, 2, 3, 4, 5, 6, and 7 one-third octave evaluate the detectability of signals with high and low frequency content. Each bands, the predicted probabilities were varying numbers of non-adjacent ambient sample was normalized to an about 55, 81, 93, 97, 98, 100, and 100 components in the presence of overall sound pressure level of 55 percent, respectively. The analysis additional ambient conditions different dB(A), so that the effect of the spectral indicates that there is a rapid increase from the standardized ambient used to content of each ambient on the in detectability as the number of develop the one-third octave band detectability of a signal could be components increases from 1 band to 4 minimum levels for detectability in the examined in isolation from other bands when each band is set at the NPRM or this final rule. The analysis variables. This analysis differs from the specified minimum detectable level. provides an estimate of how often modeling approach used to develop the Additional bands beyond 4 do not pedestrians would be able to detect a minimum one-third octave band levels appear to increase the detectability level sound signal in a 55 dB(A) ambient, for detection in the NPRM and the final significantly. An eight-band sound was with expected spectral variation, as a rule because that approach used a single not included in the analysis because function of the number of one-third ambient that was chosen for consistency eight non-adjacent one-third octave octave bands meeting the revised in development of minimum standards. bands do not fit in the frequency range minimum thresholds.48 Ambient data NHTSA refers to the resistance to over which we are establishing were collected at 17 locations along masking of a signal evaluated using this minimum requirements in the final rule. Centre Street in Newton, Massachusetts, analysis as the ‘‘robustness’’ of the This analysis also showed that some signalized and stop-controlled signal. Signals evaluated for robustness signals with content in only 2 one-third intersections (some with relatively high contained from one to seven non- octave bands are expected to be detected traffic volume and some removed from adjacent components within the 315 to with the same frequency in multiple the main road), one-way streets, and 5000 Hz frequency range. In most cases, ambients as signals with content in 4 side streets or driveways. The spectral these signals were scaled so that the one-third octave bands. Because signals shape of the ambient varies from sample components just met the minimum one- with content in 2 one-third octaves to sample, as would be expected given third octave band levels for detectability bands could be equally detectable as the different locations in which they derived from the human factors study. sounds with content in 4 one-third were collected. Some samples are This analysis predicted that, as octave bands the agency decided to dominated by low frequency content ambient conditions vary, the probability include minimum requirements for that at least one component is detectable content in either 2 or 4 one-third octave 48 For practical reasons, this analysis is limited in increases with increasing number of bands in the final rule. that it includes 17 measurement locations for the components when each component is ambient that are in one State, Massachusetts. Also, Heavy Vehicle and Motorcycle Testing ambient samples were not categorized or weighted set to the minimum detection levels according to ‘preferred crossable’ opportunities for calculated based on the human factors The research NHTSA conducted prior pedestrians. study. This is true for all operating to the NPRM focused exclusively on

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00015 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 ER14DE16.000 90430 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

light vehicles. However, since issuing As discussed in Section III.B, the agency three locations, Dayton, Ohio; the NPRM, the agency has conducted has determined that, as with other types Washington, DC; and Cambridge, some acoustic measurements on hybrid of hybrid and electric vehicles, it is Massachusetts. The four vehicles in the and electric heavy-duty vehicles (GVWR appropriate that the requirements of this screening tests were all transit buses over 10,000 lb.) and electric final rule should apply to hybrid and and included a New Flyer diesel-electric motorcycles.49 The test protocol used electric motorcycles. hybrid bus in Washington, DC; a for those measurements followed NHTSA also collected acoustic data trackless electric trolley bus and a procedures in SAE–2889–1 (May 2012). for a pure electric heavy vehicle diesel-electric hybrid trolley bus in Two electric motorcycles were tested (Navistar eStar two-axle delivery van) Dayton, and a Neoplan trackless electric at the Transportation Research Center in on a surface compliant with ISO 10844 trolley bus in Cambridge. Each vehicle Columbus, Ohio, on a test surface and suitable for heavy vehicles. No was tested in as many of the applicable conforming to ISO 10844–2011 issues were encountered in applying the operating scenarios (stationary, 10, 20, specifications. NHTSA was able to test protocol to the heavy vehicle tested. and 30 km/h pass-by) as possible. apply the proposed test procedure to the It is important to note that only this one However, due to vehicle or site motorcycles without major issues.50 The delivery truck was tested. The agency limitations, not all vehicles were tested overall sound pressure levels for a 2012 was unable to obtain electric or hybrid in all of those operating scenarios. model Brammo Enertia were 57.0, 63.2 heavy-duty vehicles with different sizes The screening data showed that the and 66.5 dB(A) for the 10, 20, and 30 and configurations for testing. The overall levels for these vehicles range km/h pass-by, respectively. The overall overall sound pressure levels for the from 55.9 to 59.0 dB(A) for a stationary sound pressure levels for a 2012 model Navistar eStar were 55.4, 64.5, 73.4, and test; 61.7 to 69.3 dB(A) for a 10 km/h Zero S were between 6.2 to 7.9 dB lower 75.2 dB(A) for the stationary, 10, 20, and pass-by test; and 66 to 70.3 dB(A) for a with 49.1, 57.0 and 59.6 dB(A) for the 30 km/h pass-by scenarios, respectively. 20 km/h pass-by test. The acoustic 10, 20, and 30 km/h pass-by, The acoustic measurements for this measurements for these vehicles were respectively. vehicle were computed and compared computed and compared to the NPRM The one-third octave band levels for to the minimum levels needed for minimum levels for detection in the the two motorcycles were computed and detection in the frequency range from frequency range from 315 Hz to 5000 compared to the minimum levels 315 Hz to 5000 Hz.52 The data showed Hz, for the eight bands included in the needed for detection (as determined in that the measured one-third octave band NPRM.53 The data showed that the NHTSA’s research described in Section levels for the e-Star heavy vehicle are measured levels for the heavy vehicles II.C 51) in the frequency range from 315 equal to or greater than the minimum tested are equal to or greater than the Hz to 5000 Hz. Results for the 2012 levels for detection in seven bands for minimum levels in five to seven bands Brammo Enertia show that the measured stationary, nine bands for the 10 km/h for stationary; five to eight bands for the levels were equal or greater than the pass-by, eight bands for the 20 km/h 10 km/h pass-by; two to five bands for minimum levels in two bands for the 10 pass-by, and seven bands for the 30 km/ the 20 km/h pass-by; and seven bands km/h pass-by and in three bands for the h pass-by. Thus, this vehicle generated for the 30 km/h pass-by. The screening 20 km/h pass-by. Sound levels for the appreciable sound at low speeds data were informative about hybrid and Enertia for the 30 km/h pass-by did not without the addition of a pedestrian electric medium-duty and heavy-duty meet the minimum levels for detection alert system, and we would expect this vehicle noise levels, but they were not in any one-third octave bands from 315 vehicle to be detectable. However, intended to be conclusive, and thus the Hz to 5000 Hz. Sound levels for the because this testing was limited to only agency did not determine from this 2012 Zero S did not meet the minimum one electric truck, the agency is not able testing that it would be appropriate to levels for detection in any of the bands to reach any general conclusions that exclude medium and heavy vehicles for all pass-by tests (i.e., 10, 20, and 30 hybrid and electric heavy vehicles from the final rule. km/h). While there is an appreciable should be exempt from the final rule. Analysis of Indoor Test Data difference between the two models The agency also collected ‘‘screening’’ NHTSA also analyzed acoustic data tested, these results indicate that both data for four hybrid and electric heavy- measured in hemi-anechoic chambers models operate quietly over all or part duty vehicles. Screening tests were of the range of speeds up to 30 km/h. equipped with a chassis conducted in the field (not on ISO dynamometer.54 The data acquired at 10844 sound pads) at convenient indoor test facilities included 49 Hastings, et al. Acoustic Data for Electric Heavy locations using portable sound level Vehicles and Electric Motorcycles. (2014) DOT/ measurements of electric, hybrid, and NHTSA. meters. We note that the test protocol internal combustion engine vehicles. 50 One notable change is that the motorcycles used for the screening tests did not NHTSA’s analyses examined ambient were run just to the right of the center of the lane fulfill all the parameters stated in SAE– noise, repeatability, and reproducibility with respect to the direction of travel. This was J2889–1, and the measurements may not of the indoor acoustic measurements. done so the motorcycles’ tires were not rolling on have been within the constraints of the the painted center line, since it was important to Acoustic data were collected at two keep the tires on the portion of the test track which SAE standard for acoustic environment, indoor facilities: The General Motors had pavement meeting the ISO specification (the operating conditions, test surface, Milford Proving Grounds (MPG), in painted center line is not intended to meet the ISO number of microphones, and Milford, MI and the International specification.) Additionally, motorcycles were not microphone position. The results tested in reverse since they did not have reverse Automotive Components (IAC) facility, capabilities. obtained from screening data therefore 51 Hastings, et al. Detectability of Alert Signals for may deviate appreciably from results 53 Hastings, et al. Detectability of Alert Signals for Hybrid and Electric Vehicles: Acoustic Modeling obtained using protocols and test Hybrid and Electric Vehicles: Acoustic Modeling and Human Subjects Experiment. (2015) conditions that strictly adhere to the and Human Subjects Experiment. (2015) Washington, DC: DOT/NHTSA. As described in this SAE standard. Data were collected at Washington, DC: DOT/NHTSA. report, the minimum levels needed for detection 54 Hastings, et al. Analysis of Acoustic Data for were determined using an acoustic loudness model Hybrid and Electric Vehicles measured on Hemi- that was adjusted for actual human hearing 52 Hastings, et al. Detectability of Alert Signals for Anechoic Chambers. Washington, DC: DOT/ responses to vehicle sounds and other sounds by Hybrid and Electric Vehicles: Acoustic Modeling NHTSA. A hemi-anechoic chamber is a specially- using the results of a series of human factors and Human Subjects Experiment. (2015) designed room with walls that absorb sound waves experiments conducted by Volpe for NHTSA. Washington, DC: DOT/NHTSA. for better acoustic analysis.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00016 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90431

in Plymouth, MI. Indoor test data was inherent test variability, as discussed in Sound levels were measured while provided to NHTSA by Transport section III.K of this document, and also vehicles were stationary and while they Canada.55 Outdoor test data were to differences in each site’s were driving or coasting past collected by NHTSA’s Vehicle Research dynamometer/tire interaction. microphones at constant speeds of 10, and Test Center (VRTC) at the In addition to comparing the two 20, and 30 km/h. Transportation Research Center (TRC), indoor test sites to one another, both The repeatability of the measurement East Liberty, OH, and NHTSA did a facilities were also compared with of the sound pressure level was assessed comparison of indoor and outdoor outdoor measurements made at TRC. by performing multiple tests with one measurements. The dataset available to Measurement reproducibility between vehicle (a 2010 Ford Fusion) on one support these analyses included eight each indoor test facility and the outdoor surface. The TRC ISO-compliant surface vehicles. Test vehicles were transported test facility was evaluated by comparing was used for this work and tests were between the Milford and Plymouth the average sound pressure levels of performed twice a month from April to facilities so that the exact same vehicles each vehicle at each one-third octave October 2012. Each test consisted of were used at both indoor test sites. band for each speed at the respective eight individual measurements for each Vehicle make and model were sites. Results showed that the indoor scenario. Results showed that the 95- consistent between indoor and outdoor facilities tend to have higher sound percent confidence interval of the testing,56 but the outdoor test results pressure levels, especially at 20 and 30 overall sound pressure level ranged have been aggregated over several km/h. Because the differences are from ±0.7 dB to ±1.9 dB for the various testing efforts and do not in all cases smaller at 10 km/h, it is not likely that scenarios. There was no significant represent the exact same test vehicles. the differences in acoustic reflections systematic change in overall sound Repeatability at each indoor test site from the indoor floor and the outdoor pressure levels over the six month was evaluated by computing the pavement are causing the difference. period. standard error of the mean for each one- Rather, it is likely that the tire/ Data were also collected at different third octave band from the sound dynamometer interaction is producing ISO 10844-compliant surfaces to pressure measurements, considering the higher sound pressure levels. examine test reproducibility. The each measurement as an estimate of the Considering confidence intervals of reproducibility of sound pressure levels mean for each vehicle. The standard estimated mean values for individual was estimated by testing the 2010 Ford errors for these two indoor test sites vehicle/speed/frequency pairs, the Fusion twice on two other ISO- were typically around 0.5 to 0.75 dB for standard deviation between TRC and compliant surfaces (at Ford Motor the 315 Hz one-third octave band and MPG was as high as 5 dB and the Company Proving Ground in Romeo, above. This indicates that about 95 standard deviation between TRC and Michigan, and at the Navistar Test Track percent of measured one-third octave IAC was as high as 4.7 dB. Therefore, in Fort Wayne, Indiana). The average band levels for a given vehicle and tolerance values associated with 95- sound pressure levels for all scenarios operating speed will be within a range percent confidence intervals would be on the other ISO surfaces fell within the of ±1 to ±1.5 dB and, when estimating as large as ±9.8 and ±9.2 dB experimental errors of the average a mean value using four samples, the respectively. sound pressure levels measured on the mean value should be within about 0.5 These confidence intervals include TRC ISO surface. The 95-percent to 0.75 dB of the true mean with 95- site-to-site differences and differences as confidence interval of site-to-site percent confidence. a result of using different vehicles and variation for overall sound pressure Measurement reproducibility between in some cases different model years. It level ranged from ±0.6 dB to ±2.1 dB the two indoor test sites was evaluated is anticipated that this confidence and the 95-percent confidence estimates by comparing the average values of each interval would be reduced if identical for reproducibility, including the vehicle at each one-third octave band vehicles were tested. This indoor/ repeatability of the measurements, for each speed. The differences between outdoor analysis involved only a very ranged from ±1.3 dB to ±2.4 dB. sites were about 2 dB on average at 10 limited amount of data and the data in To determine if acoustic testing km/h and only about 1 dB on average some cases was not from the exact same locations could include test areas with at 20 and 30 km/h. Although the average vehicle. The agency would prefer to surfaces that are not ISO-compliant, the difference is generally less than 2 dB conduct additional testing in a more agency investigated using correction between the two sites, differences for highly controlled fashion to allow for factors to adjust data from non-ISO- specific vehicle/speed/frequency pairs more conclusive results. In the absence compliant surfaces, the agency are still significant. When considering of that, we have not changed our compared overall sound pressure levels site-to-site differences, the 95-percent position on using outdoor testing as measured on ISO 10844-compliant confidence intervals for estimated proposed in the NPRM. surfaces to overall sound pressure levels means range from ±2.5 dB to ±6.7 dB measured on three other asphalt Acoustic Measurements of Hybrid and depending on the one-third octave band. surfaces of varying characteristics. The Electric Vehicles Bands at and below 400 Hz consistently alternative surfaces were located at TRC have standard deviations greater than 2 NHTSA’s VRTC conducted additional in East Liberty, OH, and included: A dB and bands 500 Hz and above acoustic measures for hybrid vehicles, new asphalt surface in the vehicle typically have standard deviations less electric vehicles, low speed electric dynamics area; a sealed asphalt surface; than 2 dB (exceptions being 630 Hz and vehicles, and internal combustion and a skid calibration lane. These 800 Hz). The reproducibility between engine (ICE) vehicles to collect pavements were appropriate examples sites appears good. We believe the additional sound measurements and to of potential test surfaces that are not measurement differences are due to evaluate the repeatability of the test ISO-compliant to examine the impact procedure proposed in the NPRM.57 that testing using different surfaces may 55 Whittal, I.; Jonasch, R.; and Meyer, N. Quiet have on measuring vehicle sound. Vehicle Sounds Test Data (2013) Transport Canada. 57 Garrott, W. R., Hoover, R. L., Evans, L. R., Overall sound pressure levels on the Docket NHTSA–2011–0148–0321. Gerdus, E., and Harris, J. R., ‘‘2012 Quieter Vehicle three asphalt surfaces were compared to 56 Indoor results from a 2012 Nissan Leaf were Testing Report: Measured Sound Levels for Electric, compared to outdoor results from a 2010 Nissan Hybrid Electric, and Low Speed Vehicles’’ the results on the TRC ISO surface using Leaf. Washington, DC, DOT/NHTSA, November 2016. the 2010 Ford Fusion, and an EV with

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00017 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90432 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

an active external sound generator, as levels that were not significantly for each one-third octave band using a well as an EV without an active external different from their ICE peer vehicles. detection model that estimated the sound generator. Results showed that One-third octave band data and distance at which a pedestrian would be one surface tended to produce overall comparisons were also reported. able hear a given sound in the presence sound pressure levels significantly In addition, the agency compared the of a given ambient sound profile. In the lower than the ISO-compliant surface at sound pressure levels of ICE vehicles in NPRM, NHTSA proposed to require 0 and 10 km/h. Researchers concluded motion with their engines running to eight one-third octave bands with the that this was due to greater absorptivity the same ICE vehicles coasting past the perspective that required sounds should of this asphalt composition. The other microphones with their engines turned be detectable in a wide variety of two surfaces tended to generate results off. These comparisons were made at 10, ambients, including ambients that had not significantly different than the ISO- 20, and 30 km/h. The sound pressure different acoustic characteristics from compliant surface when the vehicles levels for the vehicles with their engines the ambient that we used with our were stationary or traveling at 10 km/h. running were an average of 7.9 dB detection model. The NPRM also On these surfaces, sound levels higher than in the coasting (engine-off) required that sound produced by EVs increased more rapidly than for the ISO condition at 10 km/h (min. 4.3 dB, max. and HVs be recognizable to pedestrians surface as the vehicle speed increased. 11.6 dB); 2.2 dB higher than in the as motor vehicle sounds by containing The overall sound pressure levels at 20 coasting (engine off) condition at 20 km/ low frequency tones and broadband and 30 km/h tended to be significantly h (min. 0.6 dB, max. 5.7 dB); and 0.9 dB content because these are characteristics higher for these two surfaces compared higher than in the coasting (engine off) commonly associated with sounds to the ISO surface. Researchers condition at 30 km/h (min. 0.5 dB; max. produced by internal combustion concluded that these surfaces tended to 1.7 dB). engines. generate more tire noise than the ISO- D. Notice of Proposed Rulemaking The compliance test procedure compliant surface. An attempt to use the specified in the NPRM was to be In the NPRM we proposed to apply data from the Ford Fusion to normalize performed outdoors and was based in the minimum sound requirements to all the sounds from the different surfaces part on SAE J2889–1 SEPT 2011. The hybrid and electric passenger cars, light was unsuccessful. Consequently, we did compliance test procedure contained trucks and vans (LTVs), medium and not conclude that it is feasible to test on tests for stationary, reverse, and pass-by surfaces other than an ISO-compliant heavy-duty trucks and buses, low speed vehicles (LSVs), and motorcycles, that tests conducted at 10 km/h, 20 km/h, one. and 30 km/h. We explained in the To examine the sound levels emitted are capable of propulsion in any forward or reverse gear without the NPRM that NHTSA believed that by low speed electric vehicles (LSVs), outdoor pass-by testing would be VRTC tested five of examples of these vehicle’s ICE operating. The proposed minimum sound preferable to indoor testing in hemi- vehicles. LSVs typically are lighter than anechoic chambers using dynamometers EVs and often use different tires, so it requirements would apply to these HVs and EVs in three circumstances: (1) because outdoor testing is more was prudent to conduct separate representative of the real-world measurements of LSVs rather than When operating up to 30 km/h (18 interactions between pedestrians and assume they are as quiet as EVs. The mph), (2) when the vehicle’s starting vehicles. We also expressed concern sound levels produced by the LSVs system is activated but the vehicle is 58 that specifications for indoor testing were very similar to those of the EVs, stationary, and (3) when the vehicle is were not as developed and did not have with the main difference being that four operating in reverse. The NPRM also the same level of objectivity, of the LSVs were equipped with back- contained requirements for the sound repeatability, and reproducibility as test up beepers of varying sound pressure produced by hybrid and electric specifications for outdoor testing. levels. Other than during reverse vehicles to increase and decrease in acceleration, the LSVs showed overall pitch as the vehicle increases and The NPRM proposed a phase-in sound levels with standard deviations decreases speed so that pedestrians schedule consistent with the PSEA ranging from about 1 to 2.5 dB. would be able to detect those changes. which would require ‘‘full compliance To provide data for the agency’s We proposed a crossover speed of 30 with the required motor vehicle safety analysis of the crossover speed of HVs km/h because this was the speed at standard for motor vehicles and EVs, the agency tested additional which tire noise, wind resistance noise, manufactured on or after September 1st HVs and one EV as well as a number of and other noises from the vehicle of the calendar year that begins 3 years ICE peer vehicles (in cases where a peer become the dominant noise and after the date on which the final rule is vehicle was available for the HVs and eliminate the need for added alert issued.’’ In the NPRM we stated that if the EV selected for testing) and sounds.59 the final rule was issued January 4, compared the ICE peer vehicle test The agency proposed to require HVs 2014, compliance would commence on results to the HV and EV results. At 10 and EVs to make a minimum amount of September 1, 2015, which would mark km/h, the three HVs tested (none with sound in each of eight different one- the start of a three-year phase-in period. external sound generators) had an third octave bands, under each of The NPRM proposed the following average SPL 2.4 dB lower than their ICE several test conditions. The agency phase-in schedule: peer vehicles. An EV without an active developed the minimum sound levels • 30 percent of the subject vehicles external sound generator had an average produced on or after September 1 of the SPL 7.3 dB lower than its ICE peer 58 The NPRM contained minimum sound first year of the phase-in; vehicle. At 20 km/h, the three HVs requirements for the stationary but active condition • 60 percent of the subject vehicles because the definition of alert sound in the PSEA (none with external sound generators) requires the agency to issue minimum sound produced on or after September 1 of the had an average sound pressure level 1.1 requirements to allow pedestrians to detect the second year of the phase-in; dB lower than their ICE peer vehicle operation of nearby hybrid and electric vehicles, • 90 percent of the subject vehicles and the EV without external sound had including those vehicles that are operating but produced on or after September 1 of the stationary. an average sound pressure level of 3.5 third year of the phase-in; and 59 For additional details about how and why the • dB below its ICE peer vehicle. At 30 km/ agency selected the crossover speed of 30 km/h 100 percent of all vehicles h the HVs and EV had sound pressure refer to section III.D. in this document. produced on or after, by September 1 of

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00018 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90433

the year that begins three years after the that the agency should only establish manufacture, stating that the agency had date that the final rule is issued. minimum sound requirements up to 20 not shown a safety need based on crash In the NPRM, we tentatively km/h, arguing that above 20 km/h tire data. They stated that the final rule concluded that this phase-in schedule and wind noises are the dominant should not apply to those vehicles was reasonable for manufacturers and contributors to the sound produced by because hybrid and electric motorcycles allowed the fastest implementation of moving vehicles, and provide enough and heavy- and medium-duty trucks the standard for pedestrian safety. sound for pedestrians to safely detect and buses do not pose an increased risk hybrid and electric vehicles. to pedestrians over ICE vehicles. E. Summary of Comments to the NPRM NFB and ACB supported the agency’s A number of individual commenters The agency received comments to the proposal to require that hybrid and either expressed general support for the NPRM from a wide variety of electric vehicles produce sound in the rule or general opposition to increasing commenters, including trade stationary but active operating the amount of sound produced by associations,60 vehicle manufacturers,61 condition, because it would help blind hybrid and electric vehicles. Several advocacy groups,62 suppliers,63 and visually-impaired pedestrians be individuals also questioned why the academia,64 standards-development aware of nearby vehicles and avoid agency was limiting the scope of the organizations,65 governments,66 and collisions. NFB, ACB, and Advocates proposed rule to hybrid and electric approximately 225 individuals. also supported the agency’s proposal to vehicles. These commenters stated that The primary issues raised by the establish minimum sound requirements the minimum sound requirements in the advocacy groups and manufacturers for speeds up to 30 km/h, stating that NPRM should apply to all vehicles concerned our proposal to require they believe that the agency’s research including ICE vehicles that do not sound while hybrid and electric supports establishing minimum sound produce enough sound to be safely vehicles are stationary but active and requirements to those limits. detected by pedestrians. our proposal to establish minimum Manufacturers and groups that III. Final Rule and Response to sound requirements up to a speed of 30 represent manufacturers were Comments km/h. Manufacturers and trade supportive of the concept of adding association groups argued that a sound sound to EVs and HVs to enhance A. Summary of the Final Rule at stationary is not required for safety. pedestrian detection but expressed concern that the minimum sound Today’s final rule generally adopts the These commenters stated NHTSA proposed standard but modifies the should instead mandate a commencing requirements proposed in the NPRM were more restrictive than necessary to requirements in several ways. As motion sound that activated when the proposed, we will require hybrid and driver of an HV/EV removed her foot accomplish this goal. They argued that sounds meeting the requirements electric vehicles to emit sound at from the brake pedal. Manufacturers minimum levels while the vehicle is and trade associations also commented proposed in the NPRM would be annoying to consumers and might stationary (although not necessarily at all times when the vehicle propulsion 60 The trade associations representing negatively affect sales of hybrid and manufacturers that submitted comments included electric vehicles. Regarding the agency’s system is active); while the vehicle is in the International Motorcycle Manufacturers proposed compliance test procedure, reverse; and while the vehicle is in Association (IMMA), the Truck and Engine manufacturers and groups that represent forward motion up to 30 km/h. Today’s Manufacturers Association (EMA), the Electric final rule also adopts the agency’s Drive Transportation Association (EDTA), the manufacturers requested the option to Motorcycle Industry Council (MIC) and the conduct compliance testing in indoor proposal to conduct compliance testing Organization Internationale DES Constructeurs d’ hemi-anechoic chambers using outdoors. Automobiles (OICA). The Alliance of Automobile dynamometers, arguing that that is a The agency is adopting numerous Manufacturers and Global Automakers submitted a changes to the proposal in response to joint comment that is referenced here as the more accurate and consistent method of ‘‘Alliance/Global’’ comment. testing because it is a more controlled additional analysis conducted by the 61 Such as Toyota Motor North America (Toyota), environment that minimizes the kind of agency and in response to the comments Volkswagen Group of America (Volkswagen), ambient variations that are expected in on the proposal. The most significant Porsche Cars North America (Porsche), Ford Motor outdoor environments. They also raised change relates to the scope of the final Company (Ford), American Honda Motor Co. rule. This final rule only applies to (Honda), Mercedes-Benz USA (Mercedes), General issues regarding the agency’s proposed Motors Company (General Motors), Mitsubishi method of measuring a vehicle’s change hybrid and electric passenger cars and Motors R&D of America (Mitsubishi), Chrysler in pitch as it increases or decreases LTVs with a GVWR of 4,536 kg (10,000) Group LLC (Chrysler), Navistar, Inc. (Navistar), speed, commenting that pitch shifting pounds or less and LSVs. This final rule Nissan North America, Inc. (Nissan) and BMW of does not apply to medium and heavy North America, LLC (BMW). should be measured using a component- 62 The public safety advocacy groups submitting level test, i.e., a bench test procedure, duty trucks and buses with a GVWR comments to the proposal included National rather than testing the entire vehicle. over 4,536 kg (10,000) pounds or to Federal of the Blind (NFB), National Council of Manufacturers also disagreed with the motorcycles. Based on a review of the State Agencies of the Blind, the Advocates for agency’s estimate of the cost of speaker available acoustic data regarding these Highway Safety (the Advocates), Noise Pollution Clearinghouse, the Insurance Institute for Highway systems needed to produce sounds vehicles and the comments, we have Safety (IIHS), Safe Kids Worldwide, the World capable of complying with the determined that we do not have enough Blind Union, and American Council of the Blind requirements in the NPRM, stating that information at this time to apply this (ACB). speakers capable of producing the low final rule to medium and heavy duty 63 Such as Denso International America, Inc. frequency content specified in the vehicles and motorcycles. (Denso) and Hear for Yourself, LLC. 64 Such as the Mercatus Center at George Mason proposed minimum sound requirements We have determined the final rule University, Western Michigan University (Western were more expensive than the agency should apply to LSVs, because unlike Michigan), and Accessible Designs for the Blind estimated. electric motorcycles and medium and (ADB). Organizations that represent heavy duty trucks and buses with a 65 SAE International. manufacturers of motorcycles and GVWR over 4,536 kg (10,000) pounds, 66 The European Commission Enterprise and Industry Directorate-General (DG Enterprise), and heavy-duty and medium-duty vehicles we have acoustic data showing that the Disability and Communication Access Board of took issue with the agency’s basis for LSVs are quiet. Therefore, we do not Hawaii. applying the rule to the vehicles they have any justification to exclude them

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00019 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90434 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

from the coverage of the final rule given minimum requirements provides vehicle-emitted sounds contain the requirements of PSEA. additional flexibility to manufacturers broadband sound between 160 Hz and We have also made significant for designing pedestrian alert systems. 5000 Hz. This means sound across a changes to the detectability Sounds meeting these new requirements wide range of frequencies, and reflects specifications in the NPRM, i.e., what will have a similar overall sound the fact that ICE vehicles produce sounds HV/EVs are permitted to make pressure level to those meeting the broadband sound when operating at low that the agency would consider requirements in the NPRM. These speed. We agree with commenters that compliant with the standard. After changes preserve the agency’s goal of this requirement is not critical for sound further consideration of the NPRM establishing requirements that will lead recognition because we believe that specifications, we are establishing new to pedestrian alert sounds that are pedestrians will use other sound cues specifications in this final rule that detectable in ambient sound that provide more information in order provide greater flexibility for environments with different spectral to recognize sounds meeting the manufacturers in this respect, but that shapes. The detectability specifications requirements of the final rule as vehicle- will still allow pedestrians to safely are discussed further in Section III.E of emitted sounds. In addition to the detect EVs and HVs. Specifically, this final rule. revised requirement that the alert sound whereas in the NPRM we proposed that The agency originally proposed to level must increase as a vehicle HV/EVs would have to meet minimum require ‘‘pitch shifting,’’ meaning that as increases speed, we believe that acoustic requirements in eight separate HV/EVs increased or decreased in speed pedestrians would use other cues to one-third octave bands, in this final (from stationary up to the cutoff of 30 recognize EVs and HVs such as the rule, the agency is providing two km/h), the frequency of the sound location of the sound source and the alternative acoustic specifications, produced by the HV/EV had to vary up frequency and level changes caused by either of which the agency would or down with speed by one percent per the motion of the sound, so tones and consider to be compliant, and both of km/h. After further consideration, we broadband content are not essential for which reduce the number of one-third have concluded that the proposed pitch these vehicles to be recognizable. This octave bands for which there are shifting compliance test is likely to have issue is discussed more in Section III.F minimum levels. Under the first repeatability issues and may involve of this final rule. compliance option, hybrid and electric subjective assessments in compliance With regard to test procedures, the vehicles would have to meet minimum evaluations. For those reasons, and also final rule also makes a number of acoustic requirements in four one-third in response to information raised in changes from the proposal. We have octave bands instead of eight. Under the manufacturers’ comments, the agency modified the procedure for determining second compliance option, hybrid and has decided instead to require simply whether the sound produced by two electric vehicles would have to meet that the vehicle-emitted sound increase hybrid or electric vehicles of the same minimum acoustic requirements in two and decrease in volume by a specified make, model, and model year is the one-third octave bands, plus meet an amount as the vehicle’s speed increases same. After further analysis, we have overall sound pressure minimum. and decreases. The agency believes this determined that requiring the sound Under the four one-third octave band revised requirement, like the proposed produced by two hybrid or electric compliance option, the minimum sound pitch shifting requirement, will vehicles of the same make, model, and requirements for each band would be appropriately convey to pedestrians model year to be within three dB(A) for slightly lower than the values proposed when a vehicle is accelerating or every one-third octave band between in the NPRM and the overall sound decelerating. This approach also has a 315 Hz and 5000 Hz would not pressure of sounds meeting the four testing advantage in that changes in guarantee that the sound produced by one-third octave band compliance vehicle speed and corresponding the two vehicles would be the same. We option will be similar to those meeting changes in vehicle-produced sound can have instead decided to ensure that EVs the proposed requirements for eight be determined using the same data and HVs of the same make, model, and bands in the NPRM. Under the two one- collected during the stationary and model year produce the same sound by third octave band compliance option, constant-speed pass-by tests. This issue requiring that all vehicles of the same the minimum sound requirements for is discussed further in Section III.G of make, model, and model year use the each band are lower than those of the this final rule. same alert system hardware and eight one-third octave band proposal in The agency also proposed to require software, including specific items such the NPRM for the low and mid the pedestrian alert sound to contain a as the same digital sound file where frequency bands and higher than the low frequency tone under 400 Hz to aid applicable, to produce sound used to minimum values in the NPRM for the recognizability by pedestrians, stating meet the minimum sound requirements high frequency one-third octave bands that this would make the required alert in today’s final rule. We have also made centered at 4000 Hz and 5000 Hz. sounds more similar to ICE vehicle numerous other changes to the proposed Neither the four-band compliance sounds which typically include low test procedures in response to option nor the two-band compliance frequencies. Based on additional comments. option include requirements for tones or analysis indicating that low-frequency While we have retained the broadband content contained in the tones are not essential for vehicle- requirement that EVs and HVs must NPRM. emitted sounds to be recognized as generate an alert when stationary, the For both the two-band and four-band motor vehicles in operation, and final rule requires an alert only when a compliance options, the final rule manufacturer comments arguing that vehicle’s transmission gear selector is expands the range of acceptable one- low-frequency tones would be intrusive not in the ‘‘Park’’ position. We have third octave bands to include those to vehicle occupants and expensive to changed the test procedure accordingly, between 630 Hz and 1600 Hz (these reproduce, we have decided against and we will test this condition with the bands were excluded in the NPRM). including the proposed requirement in vehicle’s gear selector in ‘‘Drive’’ or any Reducing the number of required one- the final rule. Section III.F discusses forward gear. We believe that this third octave bands while expanding the this issue in more detail. modification to the stationary number of possible bands that Also to aid recognizability, we requirement will provide pedestrians manufacturers can use to meet the originally proposed to require that the with a way to detect those vehicles that

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00020 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90435

pose the greatest risk to them (i.e., those the proposed rule would apply to any about whether hybrid vehicles other vehicles that could begin moving at any vehicle with multiple independent than hybrid-electrics would be any less moment) while ensuring that EVs and means of propulsion. However, the quiet than hybrid-electric vehicles when HVs do not produce unwanted sound in definitions section of the NPRM not equipped with pedestrian alert situations in which they do not pose a regulatory text did not include a specific systems. As for hybrids other than threat to pedestrians, such as when they definition of ‘‘hybrid vehicle.’’ electric ones, if the vehicle produces are parked. The final rule requirements Alliance/Global and OICA disagreed sound levels in excess of those required and procedures also address vehicles with the agency’s proposal that the by this final rule then no additional with manual transmission. Test standard should apply to any vehicle alert would be required; if not, an procedures are discussed in more detail with multiple independent means of additional alert would be required. propulsion, and argued that it should in Sections III.J and III.K of this Vehicles With a GVWR Over 10,000 lbs. preamble. apply only to those vehicles that have With regard to the phase-in schedule an electric motor as the additional In the NPRM, we stated that the PSEA for the standard, we have simplified the means of independent propulsion. requires the agency to apply the proposed phase-in schedule by Alliance/Global and OICA stated they requirements of the standard to all shortening it to include a single year of do not believe that vehicles with non- hybrid and electric motor vehicles phase-in, rather than the three-year electric hybrid powertrains should be which includes cars, multipurpose phase-in that the agency proposed in the subject to the requirements of the final passenger vehicles, trucks, buses, low- NPRM. This simplification provides rule, because the agency has not speed vehicles and motorcycles.67 somewhat greater lead-time and demonstrated that those vehicles are However, we acknowledged that ICE responds to vehicle manufacturers’ quiet. Alliance/Global and OICA also vehicles with a gross vehicle weight comments that the proposed phase-in stated that the final rule should include rating (GVWR) over 10,000 pounds (lbs.) was unnecessarily complex. Half of each a definition of ‘‘hybrid vehicle’’ in have a lower rate of collisions involving 68 manufacturer’s HV and EV production paragraph S4 of the regulatory text. pedestrians than light ICE vehicles, and we stated that we were not able to must comply with this final rule by Agency Response to Comments September 1, 2018, and 100 percent of calculate a separate incidence rate for each manufacturer’s HV and EV We agree that a definition of ‘‘hybrid collisions between pedestrians and production must comply with this final vehicle’’ should be included in the rule hybrid and electric vehicles with a and have added one. The definition rule by September 1, 2019. The phase- GWVR over 10,000 lbs. because the appears in Section S4 of the regulatory in does not apply to multi-stage and number of those vehicles in the on-road text, and is based on the definition for small volume manufacturers: 100 vehicle fleet was extremely limited. a hybrid vehicle that was presented in percent of their HV and EV production Because we were not able to calculate a the ‘‘Application’’ section of the NPRM must comply with this final rule by separate incidence rate for collisions preamble, where we stated that a hybrid September 1, 2019. involving pedestrians and hybrid and vehicle is ‘‘a motor vehicle that has electric heavy vehicles, we did not B. Applicability of the Standard more than one means of propulsion for calculate the benefits of applying the which the vehicle’s propulsion system rule to them in the NPRM. We stated in Definition of a Hybrid Vehicle can propel the vehicle in the normal the NPRM that we believe that as the The PSEA defines ‘‘hybrid vehicle’’ as travel mode in at least one forward drive number of these vehicles in the fleet ‘‘a motor vehicle which has more than gear or reverse without the internal increases, the difference in pedestrian one means of propulsion.’’ As discussed combustion engine operating.’’ collision rate between heavy HV/EVs in the NPRM, we concluded that the In response to the industry request to and heavy ICE vehicles would be definition in the PSEA requires the limit the scope of the rule to only HVs similar to the difference in pedestrian agency to apply the standard only to with an electric motor as the additional collision rate between light HV/EVs and hybrid vehicles that are capable of means of propulsion, we are aware that light ICE vehicles. propulsion without the vehicle’s ICE some alternative hybrid vehicles may The agency also recognized at the operating, because if the ICE is always use something other than an electric time of the NPRM that we had very running when these vehicles are drive system in conjunction with an limited data about the sound levels operating, then the fact that these ICE, for example, a hybrid that uses produced by hybrid and electric heavy vehicles may not provide sufficient hydraulic or flywheel energy storage in vehicles. We also acknowledged that sound for pedestrians to detect them place of electric motor and batteries, there are a limited number of test pads cannot be attributed to the type of although we currently are not aware of having pavements that meet ISO 10844, propulsion. Under the agency’s hybrid vehicles other than hybrid- Acoustics—Specification of test tracks interpretation of the definition of electrics that are for sale in the U.S. for measuring noise emitted by road ‘‘hybrid vehicle’’ in the PSEA, more Regardless of whether such vehicles vehicles and their tires, that can than one means of propulsion therefore are currently available for sale, however, accommodate the extra weight of heavy means more than one independent we continue to believe that any hybrid vehicles. means of propulsion. This definition of operating under an independent, non- Manufacturers and organizations that ‘‘hybrid vehicle’’ would exclude from ICE means of propulsion should be represent manufacturers of heavy-duty the applicability of the proposed required to meet the minimum sound vehicles stated that NHTSA should not standard those vehicles that are requirements of this standard because apply the final rule to heavy-duty equipped with an electric motor that we have no evidence that they may not vehicles because the agency had not runs only in tandem with the vehicle’s be capable of operating as quietly as established that these vehicles are quiet, ICE to provide additional motive power, electric hybrids. From a safety could not demonstrate a safety need to for example a vehicle that cannot perspective, the agency is concerned operate in a purely electric drive mode. with all hybrids that might operate 67 The PSEA specifically excludes trailers from The NPRM also stated that the PSEA quietly, regardless of the power source the scope of the required rulemaking. 68 For the purposes of this document we refer to did not limit the definition of ‘‘hybrid for their non-ICE propulsion, and all motor vehicles with a GVWR over 10,000 lbs. vehicle’’ to hybrid-electric vehicles, so commenters provided no information as ‘‘heavy-duty vehicles.’’

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00021 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90436 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

merit applying the requirements of the procedures to all hybrid and electric so that pedestrians and bicyclists can proposal to these vehicles, and had not vehicles. differentiate backing heavy vehicles developed appropriate requirements EMA stated that NHTSA does not from other vehicles. and compliance tests for these vehicles. have any acoustic data that shows that ACB and NFB stated that the final Safety advocacy organizations and heavy-duty hybrid and electric vehicles rule should apply to heavy-duty hybrid organizations that represent individuals are quieter than heavy ICE vehicles and and electric vehicles because these who are blind and visually-impaired, in pose a safety risk to blind and other vehicles pose the same safety risks to contrast, stated that NHTSA should pedestrians. EMA stated that the NPRM pedestrians as light vehicles, and the apply the requirements of the final rule did not contain any data comparing the number of these vehicles in the fleet to heavy-duty vehicles because these sound produced by heavy-duty ICE will likely increase in the future. vehicles would pose an increased risk of vehicles to heavy-duty hybrid and Western Michigan University stated collision with pedestrians if they were electric vehicles. EMA stated that that if the intent of the rule is to address quiet. without acoustic data on heavy vehicles, potential hazards to the travel of blind EDTA stated in its comments that NHTSA is unable to know what the pedestrians, then potentially quiet NHTSA should defer application of crossover speeds are for heavy-duty hybrid and electric heavy-duty vehicles minimum sound requirements in the vehicles or whether heavy-duty vehicles should be required to meet the final rule to heavy-duty vehicles, produce sufficient sound that they do minimum sound requirements in the motorcycles and low-speed vehicles not need to be equipped with a sound final rule. WMU stated that it was not until the agency establishes a more generation device. In addition, EMA aware of research on the audibility of complete record showing the need for stated that the crossover speed hybrid and electric buses or light rail these vehicles to meet those developed for light vehicles might be vehicles but that it seemed better to err requirements. EDTA further stated that inappropriate for heavy-duty vehicles. on the side of caution and include if the agency found that the Because these vehicles have larger tires heavy-duty hybrid and electric vehicles requirements in the final rule should than light vehicles and often have more in the coverage of the final rule. apply to heavy-duty vehicles, tires and have a less aerodynamic body motorcycles and low-speed vehicles, the design they produce more sound than Agency Response to Comments agency should develop audibility light vehicles under the same operating Despite what was proposed in the specifications that reflect the conditions. NPRM, we have decided not to apply technologies, duty cycles and uses, and EMA stated in its comments that the requirements of this final rule to sound profiles specific to these types of applying the requirements in the NPRM heavy-duty hybrid and electric vehicles. vehicles. to heavy-duty vehicles would violate We reached this decision because we do EMA and Navistar stated that NHTSA the PSEA because NHTSA has not not believe that we currently have should exclude hybrid and electric determined a separate crossover speed enough information to determine vehicles with a GVWR over 10,000 lb. for heavy vehicles. EMA stated that to whether the acoustic requirements or from the scope of this rulemaking until comply with the PSEA NHTSA must the crossover speed in this final rule are the agency identifies a potential determine the crossover speed for each appropriate for heavy-duty hybrid and unreasonable risk to safety caused by type of heavy-vehicle to which the final electric vehicles. Therefore, we plan to the quiet nature of these vehicles, rule would apply. EMA stated further conduct further research on sound develops acoustic requirements that applying the NPRM to heavy-duty emitted by heavy-duty hybrid and specifically for these vehicles, and vehicles violates the Vehicle Safety Act electric vehicles before issuing a new develops appropriate compliance test because the NPRM did not assess NPRM proposing acoustic requirements procedures. whether a different standard was for these vehicles. EMA stated that, in addition to the needed for heavy vehicles. As described in Section II.C, after incidence rate of collisions between Advocates commented that NHTSA NHTSA issued the NPRM, we pedestrians and heavy vehicles, NHTSA should apply the final rule to hybrid conducted testing to examine the sound also should consider the exposure level and electric heavy vehicles. Advocates levels produced by heavy-duty electric of pedestrians to being struck by heavy- suggested that as advances in alternative and hybrid vehicles. The agency tested duty vehicles. EMA stated that certain energy increase, there will be a greater the Navistar eStar Electric Heavy heavy vehicles such as truck tractors do number of these types of vehicles. Vehicle following the procedures in not typically operate in environments Advocates stated ‘‘the agency should SAE J2889–1, MAY 2012, using an ISO where pedestrians are present, so their consider its findings that pedestrians asphalt pad meeting the specifications risk of collision with pedestrians is and pedalcyclists, especially the of International Standards Organization much lower than the risk for passenger visually-impaired, utilize the different (ISO) 10844 ‘‘Acoustics—Specification cars. In addition to having lower rates sound of heavy vehicles when of test tracks for measuring noise of exposure to pedestrians, heavy-duty compared with light vehicles to modify emitted by road vehicles and their vehicles make up a small fraction of the their estimation of when it is safe to tyres.’’ 70 The agency compared the on-road vehicle fleet when compared to undertake a movement, like crossing a acoustic recordings of the Navistar eStar light vehicles. EMA suggested that the road, which may vary with vehicular 69 to the four-band acoustic specifications risk of a pedestrian being struck by a traffic.’’ For that reason, Advocates in today’s final rule. The eStar met or heavy-duty vehicle is much lower than suggested NHTSA should consider exceeded a number of minimum one- the risk of a pedestrian being struck by establishing different acoustic third octave levels at the 10, 20, and 30 a light vehicle when the percentage of requirements to ensure that pedestrians km/h pass-by test conditions. According heavy vehicles in the on-road fleet and and others can accurately identify and to the agency’s detection model, given a their exposure to pedestrians are distinguish between heavy and light background noise level at the standard considered. EMA further suggested that EVs and HVs. Advocates further stated ambient, a vehicle is detectable if it lower rate of collisions with pedestrians that NHTSA should standardize the and the low exposure show that NHTSA backing sound across all heavy vehicles 70 Hastings, et al., (2014) Acoustic Data for Hybrid should not apply a single and Electric Heavy-duty Vehicles and Electric countermeasure with the same test 69 Document No. NHTSA–2011–0148–0270. Motorcycles.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00022 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90437

meets or exceeds the minimum levels according to the detection model. At the whether the incidence rate of collisions for detection in at least one of thirteen 20 km/h pass-by, three of the vehicles between pedestrians and electric one-third octave bands. So the eStar would be expected to be detectable motorcycles is different than the without any noise enhancements would according to the detection model, and incidence rate of collisions between be expected to be detectable at least in two would have met the requirements of pedestrians and motorcycles with ICEs, the standard ambient at the tested pass- the final rule.73 but stated that we expected that the by speeds. For the stationary test, the This heavy vehicle screening data difference in pedestrian collision rates eStar had acoustic content that met or showed that some hybrid and electric between electric motorcycles and their exceeded the minimum values in three heavy-duty vehicles may already make traditional ICE counterparts would be non-adjacent one-third octave bands. So sufficient sound in some operating similar to the difference in pedestrian in many ambient environments, in conditions to be detected by pedestrians collision rates between light HVs and addition to the standard ambient, the according to the agency’s model. light ICE vehicles should the number of eStar without any enhancements would Because the data the agency collected electric motorcycles in the fleet match be expected to be detectable at during screening testing is limited in the current market penetration of light stationary. scope and was not obtained on an ISO HVs and EVs. Additionally, while we The agency also conducted screening 10844 compliant surface, the agency did not have data on the extent to which tests in the field of the sound levels of needs to conduct further evaluation in electric motorcycles are quieter than ICE a selection of other heavy-duty EVs and this area before we can draw motorcycles of the same type, we also HVs using a simplified procedure. For conclusions regarding the sound levels noted that neither did we have these screening tests, NHTSA measured produced by these vehicles. information indicating whether electric four different electric or hybrid-electric Furthermore, the agency does not motorcycles produced sound levels transit buses, as described in the have any data on the crossover speed of sufficient to allow pedestrians to detect agency’s report ‘‘Acoustic Data for heavy vehicles. Given that heavy these vehicles in time to avoid Hybrid and Electric Heavy-duty vehicles have very different tires and collisions. The NPRM did, however, cite Vehicles and Electric Motorcycles’’ 71 wind noise characteristics than light crash statistics contained in BMW’s which provides details of those vehicles, and these factors heavily comments on the NOI regarding measurements.72 These screening tests influence crossover speed, it is possible incidents of motorcycle collisions with were basic evaluations of the sound that the light vehicle crossover speed is pedestrians. BMW cited data from characteristics of these vehicles, and inappropriate for heavy vehicles. The NHTSA’s General Estimates System they were conducted at facilities agency anticipates conducting further (GES) for the period between 2005 and belonging to transit agencies or at other research and evaluation to make these 2009 shows that 1.07 percent of the suitable locations. Therefore they did determinations and, if it proves pedestrians injured in motor vehicle not utilize an asphalt pad meeting the necessary, to develop separate acoustic crashes were injured in crashes specifications in ISO 10844. requirements for these vehicles. involving motorcycles to illustrate the Additionally, for these screening tests Regarding EMA and Advocates low rates of crashes between the agency used hand-held (or tripod- comments that the agency should motorcycles and pedestrians.74 mounted) sound level meters rather develop a separate acoustic We also stated in the NPRM that the than the requisite microphone array specification for heavy-duty vehicles, proposal was technology-neutral and specified in SAE J2889–1. for the reasons discussed above NHTSA that it would be possible for electric In conducting these screening agrees and plans to conduct further motorcycles to meet the requirements in measurements, the agency only evaluations on this issue. the NPRM without the use of a speaker recorded results for the eight one-third Given that NHTSA has not yet system if they already produced octave bands for which we proposed established that heavy hybrid and sufficient sound to meet the requirements in the NPRM. The agency electric vehicles are too quiet to be performance requirements. We sought compared the measurements to the detected without a pedestrian alert comment on whether the minimum revised minimum detectability system, and the agency has not sound requirements should be applied thresholds based on our human factors determined that the same acoustic to electric motorcycles. research. requirements and crossover speed for The comments that the agency Of the three vehicles the agency light vehicles in today’s final rule are received in response to the NPRM from evaluated in the stationary condition, all appropriate for heavy vehicles, we are organizations that represent motorcycle had sound content in several bands, and excluding both those categories from the manufacturers for the most part all would have been detectable in some applicability section of today’s final reiterated the concerns expressed by ambient conditions according to the rule, and we anticipate conducting a MIC and BMW in response to the NOI. agency’s detection model. At the 10 km/ separate rulemaking effort to address the BMW and MIC stated in their comments h pass-by, all of the vehicles tested potential need for pedestrian alert to the NOI that, because of the unique would be expected to be detectable systems on those vehicles. attributes of motorcycles, there is no Electric Motorcycles safety need for NHTSA to establish 71 Hastings, et al., (2014) Acoustic Data for Hybrid minimum sound levels for electric and Electric Heavy-duty Vehicles and Electric In the NPRM, we stated that we had motorcycles. MIC reiterated this point in Motorcycles. tentatively concluded that the proposed their NPRM comments. According to 72 Using the informal measurement procedures to rule should apply to electric capture these recordings allowed the agency to gather data on heavy-duty hybrid and electric motorcycles, because Congress defined 74 BMW’s comments on the NOI. Available at vehicles without the difficulty and expense of ‘‘electric vehicle’’ broadly in the PSEA http://www.regulations.gov, Docket No. NHTSA– transporting these vehicles to a location where they and did not exclude motorcycles from 2011–0100–0020. Referring to the data cited, BMW could tested on a sound pad meeting the the definition. We acknowledged that argued in its NOI comments that based on the specifications of International Standards number of crashes between motorcycles and Organization (ISO) 10844 ‘‘Acoustics— the agency was not able to determine pedestrians and the percentage of all pedestrian Specification of test tracks for measuring noise crashes involving motorcycles, there is no safety emitted by road vehicles and their tyres’’ as 73 The agency only tested one of the four vehicles need for minimum sound requirements for electric required by SAE J2889–1. at 30 km/h. motorcycles.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00023 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90438 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

MIC and BMW, motorcycle riders are crash rates. However, according to MIC, evidence that these vehicles pose a able to better see and avoid pedestrians that is not the case, and the difference safety risk to pedestrians. MIC stated than automobile drivers because their in overall crash rates is supported by that if NHTSA does decide to establish view is unobstructed by pillars and sun FARS data which indicate that the minimum sound requirements for visors and they are more alert because overall crash rate for HEVs is only half motorcycles, it should extend the they themselves are vulnerable road of the overall crash rate for ICEs. MIC exemption for small-volume users. BMW and MIC maintained that stated that the higher incidence rate of manufacturers indefinitely. because motorcycles are unstable at low HEV-pedestrian collisions is likely to be IMMA suggested that electric speeds, riders are required to maintain artificial and driven by demographic motorcycles do not introduce a new a high level of alertness, which factors other than sound, mainly that threat to blind and visually impaired minimizes the likelihood of collisions HEV drivers actually tend to be safer pedestrians because blind and visually with pedestrians during low speed drivers on average, which makes their impaired pedestrians already are maneuvers. overall crash rate lower and which exposed to pedalcyclists on both the Also in their NOI comments, both inflates their rate of pedestrian crashes road and on sidewalks (and bicycles BMW and MIC stated that adding a as a percentage of all crashes. MIC would not be any louder than electric speaker system to a motorcycle could pointed out that motorcycle pedestrian motorcycles). Operators of electric involve technical challenges not present crash frequency is actually no higher motorcycles, like pedalcyclists, have the for other vehicles because there is less than for ICEs. MIC stated that crash rate advantage of greater awareness of space on the motorcycle to install the differences due to demographic factors nearby pedestrians and greater ability to speaker and the weight of the speaker are not uncommon and are, for example, avoid them. would have a greater impact on the what explain large differences in fatality IMMA stated that limited data exists vehicle’s range. MIC and BMW also rates between different types of on crashes between motorcycles and suggested that electric motorcycles motorcycles (e.g., touring bikes pedestrians and pedalcyclists but that should not be subject to the minimum compared to sport bikes). Overall, MIC there are a significant number of sound level requirements in this concluded that, because motorcycles incidences of crashes involving proposal because electric motorcycles have a lower overall crash rate than motorcycles and four-wheeled vehicles, are not quiet.75 four-wheeled vehicles, the risk they which it argued showed the high MIC commented in response to the pose to pedestrians is actually lower vulnerability of motorcycle riders and NPRM that motorcycles should be than the incidence rate of motorcycle- their inherent alertness to other road exempt from meeting the minimum pedestrian crashes might indicate. users including pedestrians. They also sound requirements in the final rule MIC also argued that it is logical that commented that motorcycles by design because motorcycles, both electric and motorcycles should have a lower rate of provide the operator with better vision ICE, pose less of a risk to pedestrians collisions with pedestrians because of the surrounding environment which than other vehicles, citing statistics that motorcycles require two hands to increases awareness of nearby the collision rate between motorcycles operate so there is a lower chance of the pedestrians and pedalcyclists. and pedestrians is 0.27 percent operator being distracted, which should IMMA commented that studies have compared with 0.76 percent for other decrease the risk to pedestrians. shown that pedestrians are at greater vehicles under conditions most likely to MIC stated that, in addition to having risk of being struck by HVs while the pose a threat to pedestrians (backing up, a low rate of crashes involving vehicle is operating in reverse, but this turning, entering or leaving parking pedestrians, electric motorcycles are not is not a concern for motorcycles because spaces, starting, or slowing).76 quiet. MIC referenced a report submitted the vast majority of motorcycles do not MIC argued that NHTSA’s assumption in response to the NPRM by Brammo, have a reverse gear and those that do that electric motorcycles will show a Inc., a manufacturer of electric cannot move quickly in reverse. similar increase in rate of pedestrian motorcycles, that MIC believes shows IMMA stated that preliminary data collisions as four-wheeled ‘‘HEVs’’ that by design, electric motorcycles are shows that electric motorcycles are not 77 (MIC’s term for hybrid and electric not silent vehicles when moving. MIC quiet and suggested that this data, vehicles, collectively) is invalid because stated that unlike EV automobiles, the coupled with the fact the electric four-wheeled HEVs in fact do not pose engine and drivetrain are open and motorcycles do not pose an increased a greater threat to pedestrians than ICE exposed to the surrounding risk to pedestrians, shows that electric vehicles. MIC stated that the higher environment, and will produce sound motorcycles should not be subject to the incidence of collisions between levels that exceed the sound level minimum sound requirements in the pedestrians and HEVs does not mean minimums proposed by NHTSA. MIC final rule. that HEVs collide with pedestrians at a stated that two motorcycles tested by DG Enterprise stated that the higher frequency, arguing that NHTSA’s Brammo, the Empulse and the Enertia detectability parameters determined for comparison of incidence rates of Plus, produced sound levels that were 8 EVs and HEVs in the NPRM may require pedestrian collisions between ICEs and to 18 dB(A) higher than the minimum the installation of an alert sound system HEVs to determine the overall frequency requirements in the NPRM. on other quiet vehicles such as electric of pedestrian crashes between each MIC also stated that the NPRM did motorcycles and mopeds as well as group of vehicles is only valid if both not take into account that motorcycles electrically assisted bicycles. DG classes of vehicles have similar overall do not have a reverse gear and therefore Enterprise inquired whether NHTSA do not collide with pedestrians while plans to mandate the installation of and 75 MIC submitted measurements of overall sound backing. ‘‘AVAS’’ (Acoustic Vehicle Alerting pressure level of two electric vehicle models MIC stated that NHTSA should not Systems) in all these vehicle categories. recorded at 8 km/h (5 mph) and 16 km/h (10 mph) establish minimum sound requirements in its comments to the NOI. MIC did not provide Western Michigan stated that all quiet for electric motorcycles until there is any measurements of overall sound pressure level vehicles traveling at the slow speeds for ICE motorcycles as a comparison. Available at, www.regulations.gov, Docket No. NHTSA–2011– 77 The report submitted by Brammo, Inc. is covered by the NPRM, whether they are 0100–0028. available through www.regulations.gov, Docket No. light-duty EVs and HVs or electric 76 Docket No. NHTSA–2011–0148–0268. NHTSA–2011–0148–0268. motorcycles, have the potential of

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00024 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90439

causing harm to pedestrian who are compared to the agency’s detection would allow pedestrians to specifically blind. model, the Brammo Enertia would be recognize a vehicle as a motorcycle. expected to be detectable in the 55 Furthermore, motorcycles do not need a Agency Response to Comments dB(A) standard ambient at 10 and 20 backing sound since they generally are Although the agency proposed in the km/h. According to the agency’s model, not driven in reverse.80 For these NPRM to include motorcycles in the the Zero S would not be expected to be reasons, this final rule does not apply to final rule, we have decided not to apply detectable in the 55 dB(A) ambient at motorcycles, and we anticipate the requirements of this final rule to any of the three speeds tested. conducting a separate rulemaking effort electric motorcycles. As is the case with When compared to the average overall to address the potential need for heavy hybrid and electric vehicles, we sound pressure level of four-wheeled pedestrian alert systems on electric currently do not have enough ICE vehicles, the sound level produced motorcycles. information to determine whether the by the Brammo Enertia was similar, light vehicle acoustic requirements or based on a broad selection of ICE Low Speed Vehicles the crossover speed in this final rule are measurement data which the agency In the NPRM, we stated that we had appropriate for electric motorcycles. acquired from its own testing and from tentatively concluded that Low Speed Instead, the agency is planning to other sources (shown in Table 13 of the Vehicles (LSV) should be required to conduct further research on sound NPRM). The Zero S produced a lower meet the minimum sound requirements emitted by electric motorcycles before overall sound level than the ICE mean in the proposed standard. We stated that issuing a new NPRM, if needed, to and also was lower than the mean- while we had not conducted any propose acoustic requirements for these minus-one-standard-deviation of the acoustic testing of these vehicles and vehicles. same ICE data (shown in Table 14 of the had limited real-world data on crashes As described in Section II.C of this NPRM.) involving LSVs and pedestrians, we notice, after issuing the NPRM the Based on comparing the one-third expected LSVs equipped with electric agency conducted acoustic testing on octave band data to the agency’s motors would be extremely quiet. two electric motorcycles following the detection model and comparing the EDTA stated that NHTSA should procedures in SAE J2889–1, MAY overall sound pressure levels to the defer application of minimum sound 2012.78 The agency compared the one- sound produced by four-wheeled ICE standards to LSVs until a more complete third octave band measurements of vehicles, the agency believes the record establishing the need for these electric motorcycles to the acoustic data from these two electric standards for these vehicles exists. minimum levels needed for detection motorcycles are inconclusive as to EDTA suggested that if the agency based on the agency’s detection model. whether electric motorcycles might be documents a need for LSVs to meet the The first motorcycle, the 2012 Brammo too quiet for pedestrians to detect by minimum sound requirements in the Enertia, had two one-third octave band hearing. Furthermore the agency has not final rule, the agency should then measurements at the 10 km/h pass-by collected any data or conducted any develop audibility specifications that that met or exceeded the minimum analysis regarding the crossover speed reflect the technologies, duty cycles and levels for detection out of the thirteen for electric motorcycles, which might be uses, and sound profiles specific to one-third octave bands in the range of different from that of four-wheeled these types of vehicles. interest (315Hz to 5kHz); for the 20 vehicles. Because our acoustic data Western Michigan stated that LSVs km/h pass-by, the Enertia met or show that one of the two electric should be required to meet the exceeded the minimum in three of the motorcycles would be detectable by requirements in the final rule because thirteen bands. The second motorcycle pedestrians within a safe detection they could pose a potential hazard to that the agency evaluated, the 2012 Zero distance, but the other one would not blind pedestrians. NFB stated that the S, did not have any one-third octave be, we believe that further evaluation of rule should apply to LSVs. bands that were equal to or greater than electric motorcycles is needed before we Agency Response to Comments can determine if it is appropriate that the minimum levels for detection at the We have decided to apply the they be subject to the same acoustic speeds tested. The overall sound minimum sound requirements in pressure levels for the Brammo Enertia requirements and crossover speed as today’s final rule to LSVs. The PSEA in the 10 km/h, 20 km/h, and 30 km/ four-wheeled vehicles. requires NHTSA to establish minimum Commenters stated that adding an h pass-bys were 57 dB(A), 63.2 dB(A), sound requirements for all motor alert system to a motorcycle would be and 66.5 dB(A). The overall sound vehicles that are hybrid or electric a technical challenge because pressure levels for the Zero S in the 10 motor vehicles. Because trailers are the motorcycles are very different from cars km/h, 20 km/h, and 30 km/h pass-bys only vehicles excluded from the scope in terms of layout and architecture, and were 49.1 dB(A), 57 dB(A), and 59.6 of the required rulemaking, NHTSA’s a pedestrian alert system which dB(A). interpretation is that Congress intended According to the agency’s detection includes a speaker is a significant for the agency to apply minimum sound model, a vehicle is detectable in the 55 amount of hardware to integrate into a requirements to all other vehicles that dB(A) standard ambient utilized in the motorcycle. NHTSA has not determined are HVs or EVs including LSVs. agency’s acoustic evaluations if it meets if this design burden would make it The agency tested five LSVs to or exceeds the minimum levels for impracticable for electric motorcycles to determine the sound levels produced by detection in at least one of the thirteen be required to meet today’s final rule. these vehicles. The sound levels one-third octave bands.79 When The agency also needs to further evaluate whether electric motorcycles 80 One or more models of touring motorcycle are 78 Hastings, et al., (2014) Acoustic Data for Hybrid require distinct specifications separate fitted with a reverse feature that uses the engine and Electric Heavy-duty Vehicles and Electric from four-wheeled vehicles. For starter motor to assist in backing, for example when Motorcycles. example, there is nothing in the the rider is unable to walk the motorcycle out of 79 While a sound with one one-third octave band an inclined parking space. This feature is intended at the detectable threshold would be expected to be minimum sound requirements that for limited use. Currently this feature is not present detectable in the 55 dB(A) ambient utilized in the on any electric motorcycles. As a result, reverse agency’s research, such a sound may not be same overall sound pressure level depending on the operation is not considered to be a safety issue for detectable in other ambient conditions with the spectral shape of the ambient. motorcycles as it is with passenger cars.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00025 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90440 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

produced by the LSVs for the 10 km/h, low level of sound that they produce Agency Response to Comments 20 km/h, and 30 km/h pass-bys were when operating at low speeds. We We have chosen to limit the similar to the sound levels produced by stated in the NPRM that the agency application of the final rule to hybrid the electric passenger cars that the would decide whether to apply the and electric vehicles. The PSEA agency evaluated during VTRC’s testing minimum sound requirements required NHTSA to establish minimum 81 in 2012. The sound levels produced established for HVs and EVs to ICE sound requirements for hybrid and by the LSVs when operating in reverse vehicles after completing the Report to electric vehicles. After completing the varied significantly because four of the Congress on ICE vehicles, as required by rulemaking to establish minimum sound five LSVs were equipped with back-up the PSEA. requirements for hybrid and electric beepers. We also stated in the NPRM that vehicles, NHTSA is required to Results of the acoustic testing of these complete a study and submit a report to LSVs confirmed the agency’s while some of the ICE vehicles the agency tested during our research did congress on whether there is a safety understanding that these vehicles need to apply the final rule to ICE produce similar sound levels as EVs and not meet the proposed requirements, these vehicles emit sound in areas of the vehicles. If NHTSA subsequently HVs. Also, they operate in locations determines that there is a safety need to where pedestrian exposure is similar to audible spectrum not covered in the proposed requirements. We stated that apply the rule to ICE vehicles, the that of EVs and HVs. Therefore, the agency is required to initiate a this characteristic of ICE vehicles made agency believes that electric LSVs pose rulemaking to do so. Because we have it difficult to compare the detectability an increased risk to pedestrians when not yet completed the required report to of ICE vehicles to hybrid and electric they are operating at low speed when Congress, we have not yet determined compared to conventional vehicles. vehicles solely based on acoustic whether a safety need exists to apply the Vehicles in the LSV category have a measurements. requirements of today’s final rule to ICE maximum speed limitation of 25mph, so In response to the NPRM, we received vehicles. Because they agency has not by definition LSVs operate at low several comments from members of the yet determined whether a safety need speeds. These speeds are reflective of general public stating that if the agency exists for quiet ICE vehicles to produce those for which HVs and EVs have the chose to establish minimum sound additional sound, we have no basis at highest risk of involvement in requirements for hybrid and electric this time to subject these vehicles to the pedestrian crashes when compared to vehicles it should also establish requirements of today’s final rule. ICE vehicles, as noted in Section II.B of requirements for quiet ICE vehicles. We are aware that some ICE vehicles today’s final rule. The agency is not These commenters stated that NHTSA do not meet the requirements of the aware of any factors related to the use should make the determination final rule, and that this could lead to the of LSVs that would mitigate the risk to inference that some ICE vehicles do not pedestrians created by the low sound regarding which vehicles will be subject to the final rule based on whether the produce sufficient sound to allow levels produced by these vehicles. pedestrians to detect these vehicles. We Because of the low sound level vehicle poses an increased risk to pedestrians when operating at low do not think that it is appropriate, produced by LSVs and the fact they however, to make the assumption— operate primarily at low speeds, the speed not based on the vehicle’s propulsion type. These commenters based solely on the data mentioned agency believes that it is necessary for above—that some ICE vehicles must hybrid and electric LSVs to meet the suggested that requiring only hybrid and electric vehicles to meet the produce additional sound to be safely minimum sound requirements in detected by pedestrians. As we stated in today’s final rule. This is in contrast to requirements of the final rule the NPRM, ICE vehicles produce sounds electric motorcycles and EVs/HVs with discriminates against those types of in areas of the audible spectrum that a GVWR over 10,000 for which our test vehicles. make it difficult to draw conclusions data were inconclusive regarding the DG Enterprise inquired whether about how detectable they are by sound levels those vehicles achieve NHTSA had plans to require quiet ICE comparing them to the requirements in before having any sound added. vehicles to meet the requirements of the In response to the comment submitted today’s final rule. In addition, the sound final rule. DG Enterprise further produced by an ICE includes acoustic by EDTA, NHTSA believes that acoustic inquired whether the agency considered requirements for light duty EVs and HVs characteristics such as modulation that that the minimum sound requirements enhance detectability that are not are appropriate for LSVs. LSVs are not in the final rule might influence the sufficiently different from vehicles that included in the final rule. Therefore, it installation of alert sound systems on is likely that ICE vehicles that are are not speed limited when those quiet ICE vehicles. vehicles are traveling at low speeds, so readily detectable by pedestrians might LSVs do not require a separate acoustic WMU stated that, although increases not meet the requirements of the final specifications in order for pedestrians to in the number of hybrid and electric rule. detect them. vehicles in the on-road fleet have The agency will examine whether brought about an increased awareness of there is any crash data that shows that Quiet ICE Vehicles the safety risks to pedestrians posed by ICE vehicles that produce a lower sound In the NPRM, we chose not to apply quiet vehicles, there are many modern level have an increased risk of crashes the proposed requirements to ICE vehicles that are too quiet to be with pedestrians as part of the agency’s conventional ICE vehicles for the time safely detected by pedestrians who are investigation of whether there is a safety being. We acknowledged that it is blind. ADB stated that pedestrians who need to apply the requirements of possible that some ICE vehicles may are blind are at just as much risk from today’s final rule to ICE vehicles as part pose a risk to pedestrians because of the a quiet ICE as they are from an EV or of the agency’s report to Congress. HV. ADB believes that quiet ICE C. Critical Operating Scenarios 81 Garrott, W.R., Hoover, R.L., Evans, L.R., vehicles should be subject to the final Gerdus, E., and Harris, J.R., ‘‘2012 Quieter Vehicle rule because the agency has not Stationary but Active Testing Report: Measured Sound Levels for Electric, Hybrid Electric, and Low Speed Vehicles.’’ conducted enough research about the The agency proposed to require Washington, DC, DOT/NHTSA, November 2016. detectability of these vehicles. hybrid and electric vehicles to meet

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00026 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90441

minimum sound requirements in the the stationary but active condition. A accurately gauge the amount of time ‘‘stationary but active’’ condition. The sound emitted by an HV or EV when needed to safely cross.86 agency used the term ‘‘stationary but stationary but active is analogous to the The NPRM further stated that the active’’ to describe the state of a sound produced by an ICE vehicle agency did not believe that there would stationary hybrid or electric vehicle that idling while at a standstill. We stated be any incremental increase in cost that has its propulsion system active. This is that this requirement ensures that the would result from requiring a sound at an important scenario to include responsibility to avoid a collision the stationary but active operating because these vehicles typically do not between a vehicle and a pedestrian is condition for vehicles already equipped with an alert sound system and that the idle in the way that an ICE vehicle does. shared between the driver of the vehicle draft EA showed that requiring sound at The NPRM explained that the and the pedestrian by providing ‘‘stationary but active’’ condition stationary would not have any pedestrians with an acoustic cue that a appreciable impact on ambient noise included any time following activation vehicle may begin moving at any of the vehicle’s starting system without levels. moment. While there are some scenarios regard to the transmission gear position In their comments to the NOI and in in which a driver starting from a or any other factor affecting the meetings with agency staff prior to the vehicle’s ability to begin moving (i.e., stopped position should be able to see NPRM, representatives from several parking brake application). The NPRM a pedestrian in front of the vehicle and auto manufacturers said that the agency proposed requiring EVs and HVs to thus avoid a crash, the driver may not should not establish minimum sound meet the minimum sound requirements always be relied upon, especially in requirements for the stationary but for the stationary but active condition situations where the driver may have an active condition. These manufacturers beginning 500 milliseconds after the obstructed view. A driver pulling out of did not believe there was a safety need vehicle’s starting system is activated.82 a parking space in a crowded parking lot for an alert sound when vehicles are In the NPRM, we explained that the is an example of a situation in which a stationary. They were concerned that PSEA required the agency to establish driver might not be able to see a the sound of EVs and HVs standing in minimum sound requirements for this pedestrian and the pedestrian may step highway traffic and other scenarios in operating condition. The PSEA states into the path of a vehicle just as the which pedestrians would not be that the required safety standard must vehicle is beginning to move. If the expected to be present would unnecessarily contribute to increases in allow pedestrians ‘‘to reasonably detect pedestrian is able to hear the vehicle environmental noise. Advocacy a nearby electric or hybrid vehicle in before it begins to move, the pedestrian organizations for individuals who are critical operating scenarios including, would be able to exercise caution and but not limited to constant speed, blind or visually impaired, in contrast, avoid a collision by not stepping in the argued prior to the NPRM that NHTSA accelerating, or decelerating.’’ 83 This path of the vehicle. should establish minimum sound encompasses the possibility that The agency also discussed incidents requirements for the stationary but ‘‘stationary but active’’ could be a active condition. These organizations ‘‘critical operating scenario.’’ Also, the of HVs colliding with pedestrians when stated that sound made by stationary PSEA defines ‘‘alert sound’’ as ‘‘a starting from a stopped position that vehicles is necessary for the safety of vehicle-emitted sound to enable appear in the data that the agency used blind or visually impaired pedestrians pedestrians to discern vehicle presence, for the statistical analysis of crashes to avoid collisions with EVs and HVs direction, location and operation.’’ 84 between hybrid vehicles and operating at low speeds because it Thus, in order for a vehicle to satisfy the pedestrians. The NPRM noted that allows individuals who are blind to requirement in the PSEA to provide an instances of HVs starting from a stopped proceed with caution when they hear a ‘‘alert sound,’’ the sound emitted by the position and colliding with pedestrians 85 are present in our data although the nearby ‘‘idling’’ vehicle. vehicle must satisfy that definition. The NPRM also discussed and sought sample size is not large enough to prove We explained in the NPRM that in order comment on a suggestion from a statistically significant incidence rate. to satisfy the definition of alert sound in Mercedes for alerting nearby pedestrians the PSEA the agency was required to We stated that this limited data showed that a hybrid or electric vehicle was establish minimum sound requirements there could be a safety risk which, if about to begin moving without requiring for EVs and HVs in the stationary but correct, would grow commensurate with a sound in the stationary but active active operating condition. the population of HV/EVs, such that it condition. Mercedes had suggested that We also stated that, in addition to would be appropriate to require that instead of emitting sound when the being a required operating condition vehicles provide adequate sound cues vehicle was stationary with the under the PSEA, the agency believed while stationary. propulsion system active, hybrid and that there was a safety need for hybrid electric vehicles should be required to and electric vehicles to emit a sound in In the NPRM, we also noted that sound cues produced by idling ICE emit a ‘‘commencing motion sound’’ that would activate when the vehicle 82 The NPRM proposed that vehicles with manual vehicles are critical for safe navigation transmissions meet the stationary but active by blind pedestrians. The sound was in ‘‘drive’’ and the driver released requirement when the vehicle’s gear selection produced by vehicles idling while his or her foot from the brake pedal. control is in ‘‘neutral.’’ waiting to pass through an intersection 83 Public Law 111–373, 124 Stat. 4086 (January 4, 86 provides a reference to visually- The NPRM also discussed how NHTSA staff 2011). traveled to the headquarters of the National 84 Id. impaired pedestrians so they are able to Federation of the Blind in Baltimore, Maryland to 85 Given that the language of the PSEA definition cross a street in a straight line and arrive receive training on white cane travel techniques of ‘alert sound’ uses the conjunction ‘and’ when safely at the other side. The sound of used by individuals who are blind. This allowed listing the circumstances of vehicle operation that NHTSA staff to experience firsthand the necessity a pedestrian must be able to discern, i.e., ‘‘presence, vehicles idling on the far side of the of sound at stationary to the mobility of individuals direction, location, and operation,’’ it is apparent street while waiting to pass through an who are blind. When approaching intersections, that a pedestrian must be able to discern any intersection also provides visually- NHTSA staff found the sound of idling vehicles vehicle operation, which would include the impaired pedestrians with a reference necessary for determining whether there was a condition in which the vehicle could imminently vehicle present at the intersection and whether it be in motion and present a risk to a pedestrian. for how wide a street is so they can was safe to cross.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00027 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90442 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

When the driver released the brake that as an option for manufacturers. of EVs and HVs.’’ 87 Volkswagen stated pedal, the vehicle would emit a sound Some manufacturers, including that because ICE vehicles will be for a brief period that would be Mercedes and Nissan, said that sound at present a majority of the times that noticeably higher than the sound stationary can mask the sound of other blind pedestrians are attempting to cross required at low speed. According to vehicles that are in motion. Mercedes at signal-controlled intersections, the Mercedes, this brief, elevated sound stated that it had enlisted researchers to sound produced by the idling ICE would uniquely signal the onset of conduct some experimentation on this vehicles will provide the acoustic cues vehicle motion. Once the vehicle began topic. They found in preliminary trials needed to ‘‘shoreline.’’ 88 Volkswagen to move, the alert sound would revert to that it was easier for pedestrians to stated that, by the time the market a low-speed sound which would have to detect when a vehicle begins to move if penetration of EVs and HVs increases to comply with the acoustic requirements the vehicle did not produce sound when the level at which they would make up proposed for speeds up to 10 km/h. The stationary, and that this might be the majority of vehicles idling at an agency sought comment on using a because the sound activates just as the intersection, technology will eliminate ‘‘commencing motion sound’’ approach. vehicle initiates movement. Nissan also the need for pedestrians who are blind The NPRM also solicited comment on conducted trials that they said indicated to rely on vehicle-emitted sound to whether the final rule should allow the that blind pedestrians were less aware safely navigate intersections. sound at stationary to be reduced or of traffic moving adjacent to an alert- Alliance/Global stated that NHTSA deactivated if the vehicle had been emitting stationary vehicle, i.e., when should follow the European and stationary for a prolonged period of the stopped vehicle emitted no sound, Japanese guidelines for pedestrian alert time. sound systems which concluded that Many industry commenters the pedestrians were more aware of the nearby moving traffic. there is no safety need for hybrid and responding to the NPRM raised many of electric vehicles to emit sound while the same points raised in their Volkswagen stated that vehicles that stationary. Alliance/Global also comments to the NOI and in meetings are not moving do not pose a threat to suggested that requiring a commencing with agency staff prior to the agency pedestrians or pedalcyclists. motion sound as an alternative to issuing the NPRM. Auto manufacturers Volkswagen argued that it is unlikely requiring sound in the stationary but and groups that represent them that drivers will fail to make sure that active condition ‘‘would lower the commented that sound at stationary is the vehicle’s path is clear of pedestrians ambient noise level at intersections, not necessary for safety, and that Europe when starting up from a full stop, and thus making it easier for pedestrians to and Japan do not require sound at that in the rare case in which an detect the presence and operating stationary. Industry commenters inattentive driver begins to accelerate patterns of other moving vehicles.’’ 89 expressed concern that requiring sound from a stop toward a pedestrian who is In general, commenters pointed out a in the stationary but active condition in or about to enter the vehicle’s path in number of reasons why sound in the could annoy drivers, which would harm that case, a ‘‘commencing motion’’ stationary operating condition should EV and HV sales, and that it also would sound would provide the pedestrian not be required. They stated that EVs lead to increases in environmental noise with a warning that the EV or HV is and HVs should only be required to pollution. These commenters also beginning to move, so that the emit sound when they are capable of argued that a sound at stationary would pedestrian could take appropriate moving, because vehicles with their gear mask the sound of other approaching action. selector in the ‘‘park’’ position and vehicles. Industry commenters including EMA commented that it is vehicles with the parking brake engaged Alliance/Global, Denso, EDTA, unreasonable to require heavy vehicles are not capable of motion so NHTSA Mercedes, Mitsubishi, OICA, and to emit sound continuously while idling should not establish minimum sound Volkswagen requested that NHTSA because many types of heavy-duty requirements for these conditions. For require a ‘‘commencing motion sound’’ vehicles must idle for extended periods instance, Toyota stated that, according rather than establishing minimum in order to power a variety of utility to NHTSA’s interpretation of the PSEA, sound requirements for either when a functions such as operating on-board a vehicle is capable of being ‘‘operated’’ vehicle is in ‘‘park’’ or when the vehicle equipment like hydraulic lifts or pumps. even without an operator being present in the vehicle, and that a vehicle that is is in ‘‘drive’’ but is stationary. Some of Industry commenters also commented stationary is inherently incapable of these commenters pointed out that the that the level of sound for the stationary striking a pedestrian, and therefore NPRM did not define ‘‘active’’ and condition proposed in the NPRM is too should not be required to emit sound.90 argued that NHTSA should define high, and sound level is higher than that A number of commenters expressed ‘‘stationary but active’’ specifically as of ICE vehicles at idle. They stated that, concern about the environmental noise the condition in which the vehicle’s if NHTSA did decide to establish that would be created by alert sounds gear selector is in the ‘‘drive’’ position minimum sound levels for when a and the driver has released the service emitted by stationary vehicles. Alliance/ vehicle is stationary with an active Global stated that if EVs and HVs are brake. Alliance/Global commented that propulsion system, those levels should requiring a commencing motion sound required to produce an alert sound as be lower than the levels in the NPRM. soon as the starting system is activated, that activates when a vehicle begins In addition, the sound should be moving would satisfy the requirement required only when the vehicle’s gear 87 in the PSEA that the alert sound allow See document no. NHTSA–2011–0148–0250, selector is in the ‘‘drive’’ or ‘‘reverse’’ available at www.regulations.gov. pedestrians to discern the presence, position and not when the gear selector 88 ‘‘Shoreline’’ refers to the practice by which direction, location, and operation of the is in the ‘‘park’’ position. pedestrians who are blind use walls, handrails, vehicle. Honda and Nissan, in addition curbs or other features parallel to their direction of to opposing a requirement for stationary Volkswagen noted, ‘‘for the travel to help guide them. They may also use traffic sound without further research on the foreseeable future, it is exceedingly sound for shorelining. 89 See document no. NHTSA–2011–0148–0251, need for it, commented that NHTSA unlikely that a blind pedestrian will encounter a line of vehicles stopped at available at www.regulations.gov. should not require a commencing 90 See document no. NHTSA–2011–0148–0272, motion sound and should instead leave a traffic light that is comprised entirely available at www.regulations.gov.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00028 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90443

they will be required to make noise vehicles are stationary with their Agency Response to Comments under conditions for which there is no propulsion systems active. Among the As described in Section II.A of this threat to pedestrians, which in turn will comments from these organizations was final rule, NHTSA has concluded that needlessly increase environmental noise the contention that the sound of the PSEA requires NHTSA’s safety levels. Volkswagen stated that requiring ‘‘idling’’ vehicles is useful for standard to specify that vehicles must EVs and HVs to emit a sound at navigation by pedestrians who are blind have sound when stationary. However, stationary would cause many hours of in a number of scenarios and makes based on careful review of the unnecessary sound emissions, which them aware of the presence of a nearby comments received, we have decided to will annoy vehicle owners and add to vehicle that is likely to start moving at modify the proposed sound at stationary overall noise pollution. Volkswagen also any moment so the pedestrian has the requirement to apply only when a claimed that requiring sound at opportunity to react safely once that vehicle’s gear selection control is not in stationary would lead to unnecessary vehicle begins to move. These the ‘‘Park’’ position. wear and tear on the sound generation organizations stated they do not believe The definition of ‘‘alert sound’’ in the system components. that a ‘‘commencing motion sound’’ is PSEA requires the agency to establish Representatives from Nissan, Toyota, sufficient to replace the acoustic cues minimum sound requirements to allow Honda, GM, and Mitsubishi conducted provided by ‘‘idling’’ vehicles. However, pedestrians to detect the presence of a demonstration attended by NHTSA some of these commenters suggested nearby vehicles that are in operation. Of 91 staff to show that a vehicle that emits that they would not be opposed to a the comments that suggested that the sound when stationary could mask the commencing motion sound if it is agency define ‘‘stationary but active’’ as presence of other vehicles. They provided in addition to, not in place of, the condition in which the vehicle’s conducted the demonstration to a stationary sound. Advocates gear selection control is in ‘‘drive’’ and highlight situations in which they commented that the sound required for the driver is not applying the brake believed pedestrians would be able to a stationary vehicle in ‘park’ could be at pedal, none of those comments better detect other approaching vehicles a lower acoustic level until such time as explained how that approach would if nearby hybrid and electric vehicles the brake pedal is applied. fulfill the mandate in the PSEA that the did not emit sound while they are WMU stated ‘‘pedestrians who are minimum sound requirements allow stationary. Their contention was that blind gain important information pedestrians to detect the ‘‘presence’’ and requiring a stationary hybrid or electric regarding vehicle presence from the ‘‘operation’’ of a nearby vehicle, vehicle to emit sound could mask the sounds of idling vehicles’’ 93 and ‘‘blind including one that is stationary. sound of a moving vehicle that was pedestrians often rely heavily on the The agency believes that adopting the approaching in an adjacent lane. sound of vehicles starting up from a sound at stationary requirements will Representatives from Nissan met with stop at an intersection (signalized or mitigate the potential risk to pedestrians NHTSA staff and presented their not) to decide when to cross and to from HVs and EVs starting from a analysis of when a sound at stationary understand the geometry and operation stopped position. As we stated in the would be beneficial to pedestrians and of the intersection.’’ 94 These assertions NPRM, there is evidence in the crash when it would mask the sound of an were reflected to a great extent in data that these types of crashes do approaching vehicle that actually posed comments from other organizations 92 occur. A sound at stationary would help a threat to pedestrians. In this among this group. both blind and sighted pedestrians analysis, Nissan examined thirty WMU also stated that its research has because it would alert them to the different traffic scenarios. Nissan stated shown that blind pedestrians have great presence of a vehicle that might start that it had found that requiring EVs and difficulty detecting hybrid and electric moving so they could avoid walking HVs to emit a sound at stationary would vehicles (without an alert system) into the vehicle’s travel path. We are make it more difficult to detect an starting from a stopped position and, concerned that a ‘‘commencing motion’’ approaching vehicle that posed a threat consequently, sound in the stationary sound would not always give a to pedestrians in twenty of the thirty but active condition should be required pedestrian who was entering the path of scenarios, would have no impact in when the hybrid or electric vehicle’s a vehicle sufficient time to react to eight of the scenarios, and would aid the gear selection control is in ‘‘park’’ to avoid a collision, as argued by ACB and pedestrian in detecting the threat alert blind pedestrians of potential NFB. While we agree that the onset of vehicle in only two of the scenarios. conflict. WMU expressed concern that a an alert sound coincident with the Nissan indicated that it would be more hybrid or electric vehicle could be put commencement of motion on a vehicle difficult for pedestrians to detect an into ‘‘drive’’ and begin moving quickly that was not emitting sound when it was approaching vehicle that posed a threat enough that a pedestrian walking near stationary might be of some benefit, in these twenty scenarios because a the vehicle would not have time to because the contrast provided by the stationary EV or HV producing an react. activation of the sound might better ‘‘idle’’ sound would mask the help pedestrians who are blind detect approaching vehicle that posed the WMU also stated that, while a commencing motion sound does not when the vehicle begins to move, we do threat. not believe that this outweighs the fact Organizations that represent replace sound at stationary, it does that requiring sound at stationary will individuals who are blind or visually allow pedestrians to more easily help all pedestrians avoid collisions impaired and safety advocates including identify vehicles starting from a stopped with vehicles starting from a stopped NFB, ACB, ADB, NCSAB, WBU, WMU, position. WMU suggested that, if a position by providing an audible and Advocates stated that the agency vehicle has been stationary for a long indication of a nearby vehicle that could should require hybrid and electric time, that vehicle is less likely to begin begin moving at any time. vehicles to produce sound when those moving and should not be required to produce a sound for a prolonged period. While it may be some time in the future before it becomes likely that a 91 See document no. NHTSA–2011–0148–0240, available at www.regulations.gov. 93 See document no. NHTSA–2011–0148–0180, pedestrian who is blind will encounter 92 See document no. NHTSA–2011–0148–0051, available at www.regulations.gov. traffic that is comprised exclusively of available at www.regulations.gov. 94 See id. EVs and HVs (as VW’s comment

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00029 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90444 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

suggested), a sound at stationary can vehicle equipped with sound at but Active’’ as used in the NPRM is no assist pedestrians who are blind with stationary. longer accurate because this final rule navigation and orientation tasks before NHTSA believes manufacturers will allows EVs and HVs to be ‘‘active’’ that scenario becomes a reality. A sound install alert sounds on vehicles that are without emitting an alert sound. That is, at stationary can assist pedestrians who acceptable to drivers because they do the ignition of an HV or EV can be in are blind in performing orientation and not want to annoy current or potential the ‘on’ position while the vehicle is not mobility tasks in commonplace customers. We do not know whether the emitting an alert, assuming the vehicle’s situations such as when a pedestrian second study conducted by Nissan gear selector is in Park. This scenario encounters a single EV or HV at an could have been influenced by the fact would not have been allowed under the intersection where the traffic flow is that the testing in question occurred proposed requirement. Therefore, we light. As stated above, a sound at indoors, and we would expect the have chosen to simply use the term stationary also would provide circumstances under which a vehicle ‘‘stationary’’ rather than ‘‘stationary but immediate benefits to pedestrians who would be making a sound at stationary active’’ for this operating condition. are blind by allowing them to avoid indoors to be limited. We do not believe Furthermore, the regulatory text collisions with EVs and HVs starting that this second study is representative adequately specifies the conditions for from a stopped position. of the real-world situations in which a stationary tests, and the words ‘‘but NHTSA does not believe that the driver would be exposed to a sound at active’’ do not clarify any aspects of possibility that a sound at stationary stationary. Given our questions about testing. For these reasons, the phrase might mask the sound of other vehicles the findings of Nissan’s second study, ‘‘stationary but active’’ is not used in the operating in the vicinity outweighs the the fact that we do not have any other final rule. benefits of requiring a sound in the data on this issue from other We believe that requiring sound at stationary but active condition. After manufacturers, and the fact that stationary only if a vehicle’s gear reviewing Nissan’s analysis of scenarios, Nissan’s original study showed that over selector is not in the ‘‘Park’’ position NHTSA is unable to determine whether 60 percent of customers would accept a will still allow pedestrians to avoid a pedestrian who is blind would attempt sound at stationary, we do not have crashes with HVs and EVs starting from to cross in the situations in which enough information to indicate that the stopped position, while also Nissan claimed that a sound at concerns regarding public acceptance of minimizing sound in situations in stationary would mask the sound of an a sound at stationary are sufficient to which vehicles may pose no immediate approaching vehicle. For example, some outweigh the safety justifications for a risk to pedestrians, such as when they of those scenarios involve a pedestrian sound at stationary or the requirements are parked with their ignition turned on. who encounters a stationary vehicle that of the PSEA. Furthermore, a vast HVs and EVs that are stationary pose a is being passed by another vehicle majority of ICE vehicles make a sound risk to pedestrians only if they could travelling in the same direction in an at stationary, and that sound does not begin moving at any moment. When a adjacent lane. The agency is unsure deter customers from buying those vehicle is in Park, the driver must step whether upon encountering a stationary vehicles. on the brake and move the gear selector vehicle, a pedestrian who is blind In reference to comments about to Drive or Reverse and then release the would proceed to cross in front of the stationary alert sounds having brake in order to begin moving, which vehicle without waiting for the vehicle environmental impact, the agency takes some time. Although there are to move away so the pedestrian can be conducted an environmental assessment situations in which a driver could sure no other traffic is present and that and concluded that the requirements quickly shift a vehicle into Drive and it is safe to cross. overall will have a minor impact on begin moving, there also are situations Nissan presented data showing that environmental noise.97 in which a vehicle in Park with its some of the company’s customers would After reviewing the comments and all ignition turned on will remain find the sound at stationary to be information provided in response to the stationary for a prolonged period of unacceptable. In one Nissan study, over NPRM on this issue, the agency has time. Without data to indicate which of 60 percent of the subjects found an alert decided to limit the requirements for the these scenarios is predominant, we sound at stationary to be acceptable stationary but active condition to when believe that requiring an alert sound when the overall sound pressure level an HV or EV’s gear selector is not in while HVs and EVs are stationary but was similar to that of sounds meeting ‘‘Park.’’ As stated in Section II.A, the are not in ‘‘Park’’ appropriately balances the requirements of today’s final rule.95 term ‘‘operation’’ means a state of being pedestrian safety, as provided for in the In a second Nissan study, which was functional or operative. The agency PSEA, with concerns about producing conducted indoors, the number of believes that it is reasonable to conclude sound when it is not necessary to alert participants who found an alert sound that Congress intended the term pedestrians. Such concerns were at stationary unacceptable was 50 ‘‘operation’’ in the PSEA to be the expressed by a number of commenters percent with the windows of the vehicle condition in which a driver is operating including vehicle manufacturers but rolled up when the overall sound the vehicle as opposed to the operation also by a large number of individuals pressure level was similar to that of of the vehicle’s propulsion system. It is who commented on the NPRM and who sounds meeting the requirements of the operation of the vehicle by the stated that adding alert sounds to today’s final rule.96 No other commenter driver, not the operation of the vehicle’s vehicles will create noise in provided data or survey results showing propulsion system, that creates the environments and circumstances that that a sound at stationary would affect safety risk to pedestrians who are otherwise would be quiet. As with automatic-transmission HVs customer acceptance. Nissan did not unable to detect hybrid and electric and EVs, our intent is that the stationary submit any data that would indicate that vehicles. requirement will ensure that manual- customers would decline to purchase a We note that, as a result of this decision, the terminology ‘‘Stationary transmission HVs and EVs also emit an alert sound in all routine in-traffic 95 See document no. NHTSA–2011–0148–0051, available at www.regulations.gov. 97 ‘‘Environmental Assessment—Minimum Sound situations but not when they are parked. 96 See document no. NHTSA–2011–0148–0320, Requirements for Hybrid and Electric Vehicles,’’ However, for manual-transmission available at www.regulations.gov. docket no. NHTSA–2011–0100. vehicles, there is no gear selector

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00030 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90445

position exactly analogous to the Park We also stated that HVs and EVs As discussed in Section III.C, we are position; the Neutral position is similar, should be required to produce a sound requiring the sound levels when the but not the same. Automatic- while operating in reverse despite the vehicle is in reverse to be slightly higher transmission vehicles typically remain agency’s rear visibility requirements in than when the vehicle is stationary and in Drive, i.e., not in Park, as long as they FMVSS No. 111. lower than the levels required for are in traffic, but they typically are in The NPRM stated that NHTSA’s vehicles moving forward at more than Park when stationary for more than a report on the incidence rates of crashes 10 km/h because the vast majority of short time. In contrast, manual- between HVs and pedestrians found 13 vehicle operation in reverse is likely to transmission vehicles may routinely be collisions with pedestrians when an HV be limited to speeds around 10 km/h. In in Neutral both in traffic (e.g., vehicles is backing up.99 We explained in the addition, drivers may be less aware of waiting at traffic lights) as well as when NPRM that while we could not establish pedestrians passing behind their vehicle parked. If we were to specify that an a statistically significant incidence rate because of obstructed visibility to the alert sound is required on manual- for backing crashes for HVs to compare rear. transmission HVs and EVs only when to backing crashes involving ICEs due to For the reasons discussed in Section the gear selector is in a position other the limited sample size, these accident III.G, the final rule no longer contains than Neutral, that would fail to achieve reports do show that these crashes requirements for pitch shifting, so there the desired safety outcome because occur. We also stated that backing will be no such requirements when the some routine in-traffic situations would incidents occur in parking lots, garages, vehicle is operating in reverse. We note not be covered (e.g., vehicles waiting at and driveways, as well as other ‘‘off that the requirement in the final rule traffic lights). Consequently, we have roadway’’ locations that would not be that the volume of the sound produced decided to focus on parking brake usage captured in the State Data System, and by the vehicle increase as the vehicle as an alternative factor to determine thus they might be underreported. increases speed does not apply when when an alert is needed on a stationary Because of difficulties in conducting the vehicle is operating in reverse. HV or EV with a manual transmission. tests with the test vehicle is in motion The agency has considered the We are specifying in the stationary in reverse, the NPRM stated that the potential impact on today’s final rule of requirement that the alert sound on agency would test the minimum sound the NHTSA rulemaking on FMVSS No. manual transmission-equipped HVs and 111 to expand the required rear field of requirements for reverse while the 100 EVs must activate any time the ignition vehicle is stationary but with the reverse view. The expanded field-of-view is turned on and the parking brake is not gear engaged. requirements will reduce pedestrian in the applied position. Thus, a vehicle Alliance/Global stated that HVs and crashes involving backing vehicles of all with a manual transmission that is EVs should not be required to make propulsion types. On the other hand, it parked and idling will not be required sound while stationary in reverse. will not eliminate those crashes. As we stated in the NPRM, establishing to emit an alert sound as long as the Alliance/Global also stated that HVs and minimum sound level requirements for parking brake is applied. We believe EVs should emit the same overall sound reverse operation will ensure that both that this approach responds to pressure level as in the stationary but the pedestrian and the driver continue comments, that it is within the scope of active condition when in reverse and to have the ability to avoid pedestrian- the proposal, and that it meets the goal only when the vehicle is in motion. vehicle collisions. Nevertheless, we of improving safety for blind and other Honda stated that the agency should have adjusted the target population in pedestrians while minimizing non- not require pitch shifting when HVs and our assessment of benefits to reflect the essential vehicle noise. EVs are operating in reverse. Honda also As discussed elsewhere in today’s recent amendments to FMVSS No. 111 stated that NHTSA should consider the final rule, the minimum sound level under which many vehicles will be role of pending changes to the requirements for the stationary equipped with rear vision cameras. requirements of FMVSS No. 111 that condition are based on the agency’s The proposed requirements in the should serve to increase the driver’s detection model. These minimum NPRM for operation in reverse allowed level of awareness of pedestrians who requirements represent the sound levels the use of back-up beepers that most may be present while operating a that a pedestrian would need in order to heavy vehicles are equipped with as a vehicle in reverse. hear a vehicle at a distance of two means of compliance with the meters. For more discussion of the Agency Response to Comments pedestrian alert safety standard. As minimum sound requirements, see We have decided to establish noted elsewhere in this preamble, this Section II.C in this notice. minimum sound requirements final rule does not apply to medium and heavy vehicles, so the proposed Operation in Reverse applicable to HVs and EVs with their requirement to allow the use of back-up In the NPRM, we stated that reverse gear selection control in reverse, both when stationary and when moving. We beepers is not included in this final is a critical operating scenario for which rule. the agency should issue minimum are requiring HVs and EVs to produce sound requirements for HVs and EVs to a sound in reverse for the reasons stated Acceleration and Deceleration provide acoustic cues to pedestrians in the NPRM and in our discussion In the NPRM, we did not include when the vehicles are backing out of regarding sound at stationary. An HV or separate test procedures to measure parking spaces or driveways, to prevent EV with its gear selection control in vehicles when they are accelerating or collisions between EVs and HVs and reverse could start moving at any time decelerating. We stated that we chose pedestrians, and to satisfy the and pedestrians should be aware of the not to propose separate requirements requirements of the PSEA.98 presence of such a vehicle so they can when EVs and HVs are accelerating and avoid walking into the vehicle’s path. decelerating because of concerns that it 98 Because the PSEA requires NHTSA to issue was not feasible to test accelerating or minimum sound levels to allow pedestrians to 99 Wu et al. (2011) Incidence Rates of Pedestrian decelerating vehicles accurately and discern vehicle presence and operation, and a And Bicyclist Crashes by Hybrid Electric Passenger vehicle moving in reverse is unquestionably Vehicles: An Update, Report No. DOT HS 811 526. repeatably. We stated that the proposed operating, a minimum sound level is required for Dept. of Transportation, Washington, DC. Available this condition. at http://www-nrd.nhtsa.dot.gov/Pubs/811526.pdf. 100 See 79 FR 19178, April 7, 2014.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00031 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90446 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

pitch shifting requirements would allow mph). The proposal contained In the NPRM, we explained that the pedestrians to detect the acceleration minimum acoustic requirements up to peer vehicle method measures the speed and deceleration of HVs and EVs, so the speed of 30 km/h because, for the at which the sound level produced by separate acoustic requirements are not reasons discussed in the NPRM, the an HV or EV and the sound level necessary. In the responses to the agency believed that 30 km/h was the produced by the vehicle’s ICE ‘‘peer’’ NPRM, the topic of acceleration and appropriate crossover speed. The agency become indistinguishable from one deceleration was not commented on believed that it was necessary to include another in terms of overall sound separately from the topic of pitch pass-by tests at speeds up to and pressure. We stated that this should shifting which is covered in Section including the crossover speed to ensure establish the crossover speed, although III.G of this final rule. that EVs and HVs meet the minimum that speed may differ depending on the For the reasons stated in Section III.G, sound level requirements for all speeds make and model of the test vehicles. we have not included a requirement for within the range of speeds covered by This method estimates the speed at pitch shifting in today’s final rule. the requirements. which an HV or EV generates a sound Today’s final rule instead contains a The agency received no comments level equivalent to the sound level that requirement that the sound produced by related specifically to the proposed would be generated if the HV or EV was a vehicle must increase and decrease in constant speed pass-by performance powered by an ICE rather than by loudness as the vehicle changes speed. requirements or associated tests. electric power. We stated that our The agency believes that a change in However, many commenters including measurements of vehicles showed that a sound level produced by EVs and HVs manufacturers, manufacturer gap in sound level between HVs or EVs as their speed changes will provide an organizations, and advocacy groups and their ICE peer vehicles still existed acoustic cue for pedestrians to detect argued either for or against the proposed at 20 km/h (12.4 mph) and became acceleration and deceleration. crossover speed of 30 km/h. The details much smaller or negligible in most tests In the NPRM, the required minimum of the comments on crossover speed are at 30 km/h. For that reason, NHTSA level in each one-third octave band was discussed in the next section (Section tentatively concluded in the NPRM that greater at higher speeds to allow III.D). ensuring EVs and HVs produce a pedestrians to detect faster moving minimum sound level until they reach vehicles from farther away and to Agency Response to Comments a speed of 30 km/h will ensure that account for increased stopping distance If a lower crossover speed had been those vehicles produce sufficient sound at higher speeds. The NPRM, however, to allow pedestrians to detect them. We selected for the final rule, the agency did not contain any maximum sound requested comment specifically on would have modified the pass-by test requirements, only minimums, at each whether the crossover speed should be sequence to replace the 30 km/h test operating condition so it would have 20 km/h instead of 30 km/h. been possible for an EV or HV to meet speed with the lower crossover speed. We also stated in the NPRM that the the acoustic requirements in the NPRM However, the agency has decided to difference in rates of involvement in by producing the same, unvarying maintain the 30 km/h crossover speed. pedestrian crashes between HVs and sound level from stationary up to 30 Because of this decision, the constant ICEs is highest, according to our crash km/h. If a manufacturer chose this type speed pass-by scenarios in the final rule analysis, when the vehicle involved was of design, pedestrians would not have will remain as proposed in the NPRM. executing a low speed maneuver prior any acoustic cues to determine if the D. Crossover Speed to the crash.101 Low-speed maneuvers vehicle was changing speed if the sound do not have a defined speed range, but produced by the vehicle also did not In the NPRM, we stated that the they include making a turn, slowing or change in pitch. We believe this would agency had tentatively concluded that stopping, backing, entering or leaving a make it more difficult for a blind EVs and HVs should be subject to parking space or driveway, and starting pedestrian to distinguish a stopped or minimum sound requirements until in traffic. Because vehicle noise very slow-moving vehicle from one that they reach a speed of 30 km/h. The increases as a vehicle goes faster, the is moving faster, and to determine if an NPRM explained that the PSEA defined agency tentatively concluded in the approaching vehicle is slowing to a crossover speed as ‘‘the speed at which NPRM that a crossover speed of 30 km/ stop. To avoid this situation, the agency tire noise, wind resistance, or other h would ensure that EVs and HVs will is requiring that the sound level factors eliminate the need for a separate produce sufficient sound up to the produced by EV and HV pedestrian alert alert sound.’’ We decided to propose a speed at which pedestrians can safely systems must increase as vehicle speed crossover speed of 30 km/h (18.6 mph) detect EVs and HVs without the aid of increases and must decrease as speed by examining the speed at which EVs an alert system. decreases. This requirement is and HVs produce a similar overall We noted in the NPRM that the implemented in Section S5.2 of the sound pressure level as their peer ICE agency was conducting an regulatory text of this final rule. vehicles, to determine the speed at Environmental Assessment (EA) in which the powertrain noise of the ICE connection with the rulemaking and the Vehicles in Forward Motion at Constant vehicle was no longer the dominant draft EA showed that the difference in Speed source of the vehicle sound. This peer ambient sound levels if the agency were In the NPRM, the agency proposed vehicle method was one that NHTSA to establish a crossover speed of 30 km/ that EVs and HVs produce sound had used in research prior to the h compared to a crossover speed of 20 sufficient to allow pedestrians to detect enactment of the PSEA. As far as the km/h was expected to be negligible. these vehicles at all speeds between 0 agency was aware, this method was a Several commenters to the NOI and and 30 km/h (18.6 mph). The agency reasonable way to identify an participants in United Nations proposed to ensure that EVs and HVs appropriate crossover speed. We also Economic Commission for Europe produce a minimum sound level examined the crash statistics from the necessary for safe pedestrian detection State Data System to determine if there 101 Wu, et al. (2011) Incidence Rates of Pedestrian at constant speeds by measuring vehicle was a speed above which the rate of And Bicyclist Crashes by Hybrid Electric Passenger Vehicles: An Update, Report No. DOT HS 811 526. sound output at 10 km/h (6.2 mph), 20 pedestrian crashes for HVs and ICE Dept. of Transportation, Washington, DC. Available km/h (12.4 mph), and 30 km/h (18.6 vehicles were the same. at http://www-nrd.nhtsa.dot.gov/Pubs/811526.pdf.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00032 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90447

(UNECE) informal working group equivalent ICE vehicle. Alliance/Global higher, 30 km/h, crossover speed in the meetings 102 stated that the agency claims that at 20 km/h the EV or HV in final rule. should adopt a crossover speed of 20 electric power mode is only slightly DG Enterprise stated that a 30 km/h km/h. quieter than an ICE vehicle. Alliance/ crossover speed would be excessive In the NPRM we discussed research Global also stated tire noise above 20 because most EVs and HVs already presented by JASIC. JASIC determined km/h interferes with the alert sound, produce sufficient sound in the 20–25 the crossover speed for several vehicles making the detection and measurement km/h speed range to be detected by by measuring when the tire noise was of specific sound content in one-third pedestrians. DG Enterprise believes dominant over engine noise. In this octave frequencies much more difficult. these vehicles make enough sound to be research JASIC compared the sound Alliance/Global stated that a crossover detectable because they use low-rolling produced by a vehicle when tested a speed above 20 km/h is not needed to resistance tires that produce more noise constant speed with the vehicle’s ICE on fulfil the safety goals of the final rule. than conventional tires. to the sound produced by the same The European Union commented that Advocacy groups for individuals who vehicle when tested with its ICE off. The the limits on crossover or ‘‘threshold’’ are blind stated in their comments that purpose of this test was to determine the speed indicated in the NPRM—30 km/ the crossover speed should be 30 km/h point at which the vehicle produce a h for forward motion and 18 km/h for and that NHTSA had provided similar sound level with its ICE off as reverse motion [the agency notes, sufficient data to justify that decision. it did with its ICE on. JASIC concluded however, that the latter figure does not NFB stated that the agency should from its research that tire noise was reflect any proposed requirement, and establish a crossover speed of 30 km/h which would ensure that EVs and HVs dominant for every ICE and hybrid may have been an oversight in the EU are detectable when operating on vehicle tested at speeds that exceeded comment letter]—are considered quieter paved surfaces and/or when 20 km/h. Honda and Nissan mentioned excessive as many if not most EVs and using quieter tires. the JASIC data as adequate justification HEVs produce sufficient noise for a 20 km/h crossover speed. The data emissions in the 20–25 km/h and 10–12 Agency Response to Comments indicated that JASIC evaluated six km/h speed ranges for forward and In this final rule, the agency has different vehicles, each found to have a reverse motions, respectively. This can decided to maintain the crossover speed crossover speed very close to 20 km/h. be attributed to the fact that EVs and of 30 km/h as proposed in the NPRM. At the time the NPRM was issued, the HEVs use low-rolling resistance tires In development of the NPRM and agency did not believe the JASIC data which produce more noise emissions final rule the agency carefully was sufficient for a 20 km/h crossover than conventional ones as well as to the considered the term ‘‘crossover speed,’’ speed determination. increased drivetrain/powertrain noise what it means, and how it should be In the NPRM, the agency solicited emissions when the vehicle is in determined. The PSEA requires an alert comments on whether 20 km/h should reverse. be added to electric and hybrid vehicles be the crossover speed instead of the Honda said that acoustic data shows up to the ‘‘crossover speed.’’ The PSEA proposed speed of 30 km/h. The agency a convergence of the vehicle’s sound defines crossover speed as ‘‘the speed at also requested additional research data profiles between the engine-on and which tire noise, wind resistance, or that could be used to support a 20 km/ engine-off condition at 20 km/h, and other factors eliminate the need for a h crossover speed decision. that acoustic sound requirements at 20 separate alert sound as determined by All of the vehicle manufacturers and km/h or more might not be necessary. the Secretary.’’ ‘‘Alert sound’’ was itself the organizations that represent Toyota explained that data presented defined as ‘‘a vehicle-emitted sound to manufacturers stated in their comments by the Quiet Road Transport Vehicles enable pedestrians to discern vehicle that NHTSA should adopt a crossover (QRTV) group have indicated that the presence, direction, location, and speed of 20 km/h in the final rule. These appropriate crossover speed is 20 km/h, operation.’’ commenters stated that a crossover because tire and wind noise exceed the To date, it has been a common speed of 30 km/h is overly burdensome noise of traditional ICE vehicle engines understanding that when ICE vehicles and would lead to increases in traffic above this speed. Toyota mentioned that are operated at low speeds, they are noise. They also stated that the existing Japanese and European detectable primarily due to the sounds difference in sound of HVs and EVs guidelines have adopted 20 km/h as the generated by their internal combustion compared to ICE vehicles is marginal at appropriate crossover speed and engine and drivetrain, and secondarily 20 km/h, and that a crossover speed of recommended that NHTSA do the same. due to tire noise and wind resistance 30 km/h is not necessary to achieve Volkswagen stated that the crossover noise, which are speed dependent, and safety goals. Manufacturers stated that at speed in the final rule should be 20 km/ to other factors. At higher speeds, the speeds higher than 20 km/h, tire and h. Volkswagen stated that for customer sound generated by an ICE vehicle’s wind noise interfere with measurement satisfaction reasons it will design the tires, wind resistance, and other factors of the alert sound. These commenters alert sound to fade out gradually above become the primary sound source, and also stated that the agency should adopt the crossover speed, rather than the engine sound becomes secondary 20 km/h as a crossover speed to align abruptly shutting off immediately upon (there are exceptions, such as vehicles with UNECE and Japanese government reaching the crossover speed. designed to have prominent noise from recommended practices for pedestrian (Otherwise a driver travelling at the a tuned exhaust system.) Therefore, ICE alert systems. specified crossover speed would be vehicles generally are detectable at Alliance/Global stated that by the highly aware of, and almost certainly lower speeds because of the sound time an EV or HV reaches a cruising annoyed by, a sound that toggled on and produced by the ICE and are detectable speed of 20 km/h, the sound it makes off abruptly as the vehicle crossed and at higher speeds because of sound is practically indistinguishable from an re-crossed this speed.) Volkswagen produced by the vehicle’s tires, wind suggested that other vehicle resistance, and other factors. A vehicle 102 For more information about the agency’s reaches its crossover speed when it can participation in the UNECE Quiet Road Transport manufacturers will also implement alert Vehicles informal working group see NPRM, 78 FR sounds that fade out gradually, further be detected based on these other, non- 2848. weakening the rationale for setting a ICE sound sources. The effort to

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00033 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90448 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

determine the speed at which this by the eleven ICE vehicles listed below exceeded the threshold in the 1600 Hz occurs is complicated by the fact that with their IC engines turned off, we band. conventional vehicles emit a complex were able to assess if any of the A- Our conclusion from this analysis is composition of sounds and tones at weighted one-third octave band levels that at 20 km/h few HVs and EVs make various overall sound pressure levels, from any of the test vehicles met or sufficient sound to be detectable to such that crossover speed might not be exceeded the 20 km/h band threshold pedestrians without the aid of a that same from one vehicle model to levels needed for a vehicle to be pedestrian alert system. another. Furthermore, it would be detectable in a standardized 55 dBA impractical for the agency to set ambient, and to compare that outcome In light of this, and given other different crossover speeds for different to the number of bands that met or uncertainties discussed below, the vehicles. Thus, in order to ensure that exceeded the thresholds at 30 km/h. agency has decided in this final rule to all vehicles to which this rule applies (We note that this was a re-analysis of maintain the 30 km/h crossover speed can be safely detected by pedestrians, vehicle data already collected, i.e., this proposed in the NPRM. the agency believes it must set crossover evaluation did not involve additional Regarding the different analysis relied speed at a value that captures the higher vehicle testing.) Whereas the peer upon by JASIC and other commenters to end of the range of crossover speeds that vehicle and JASIC methods are relative support a 20 km/h crossover speed, we exists among light vehicles. measures because they compare one sought additional data because the The agency explained in the NPRM vehicle’s overall sound to another JASIC data was limited to a small that, in the absence of a detailed vehicle’s overall sound, this most recent number of test vehicles. So, in addition analysis supporting another crossover NHTSA evaluation compared vehicle to the agency’s detection-based analysis speed, the agency tentatively concluded sounds directly to detection criteria. discussed above, in order to address that a crossover speed of 30km/h would The results of this analysis are crossover speed comments, NHTSA ensure that pedestrians will be able to summarized below according to test conducted tests using the same method safely detect EVs and HVs in situations speed and vehicle model. The one-third that JASIC had used to derive its in which these vehicles pose an octave bands listed are those for which recommended 20 km/h crossover speed. increased risk to pedestrians because of the given test vehicle met or exceeded As described previously in this section, their quiet nature. the threshold in NHTSA’s final rule: After considering the comments the method involves comparing sound 10 km/h with the IC engine off— pressure levels from the same vehicle received and evaluating vehicle • 2012 Mini Cooper at 2000, 2500, measurements utilizing the method measured on the track during coast- 4000,and 5000 Hz down (engine off), which approximates proposed by JASIC, as well as an • 2012 Ford Focus at 5000 Hz analysis utilizing the agency’s vehicle 20 km/h with the IC engine off— an EV or HV in electric mode, and pass- detection criteria, we have decided to • 2012 Ford Focus at 800, 1000, and by (engine on) performance tests. Under require a crossover speed of 30 km/h in 1600 Hz this analysis, the speed at which coast- this final rule as proposed in the NPRM. 30 km/h with the IC engine off— down sound level is similar to the pass- No new compelling data was submitted • 2010 Buick LaCrosse at 1000, and by sound level is considered the to the agency that can be used to 1600 Hz crossover speed for that particular conclude that reducing the crossover • 2012 Mini Cooper at 630, 800, 1000, vehicle. This method identifies the speed from the proposed 30 km/h to 20 1600, 2000 Hz speed at which the sound level due to • km/h is justified. 2012 Ford Focus at 800, 1000, 1600 all factors including tire and wind Because other methods (i.e., the peer and 2000 Hz resistance noise, which are factors cited vehicle method and JASIC method) used • 2012 Lexus RX 350, 2011 Cadillac in the PSEA, is very close to the sound to determine the crossover speed were CTS, 2011 Honda Odyssey, 2012 level of the same vehicle with its ICE inconclusive, as discussed later in this Honda Fit, 2012 Toyota Camry, operating. This method is similar to the section, and did not directly answer the 2012 Toyota Corolla, and 2012 VW peer vehicle method that the agency question of when the vehicles in the Golf ICE at 1600 Hz used in the NPRM, but it uses a single analysis produced enough sound to be These results show that at 20 km/h test vehicle in two operating conditions detected by pedestrians, NHTSA did only one of the eleven tested vehicles (engine-on and engine-off). some additional evaluation of sounds had any one-third octave bands that met or exceeded the corresponding In other words, at any speed higher produced by ICE vehicles with their IC than the crossover determined engines turned off using the one-third threshold for detection.104 Therefore, ten of the eleven vehicles would not be according to this method there is no octave band detectability thresholds perceived difference between the sound from our acoustic model. The model detectable to pedestrians at 20 km/h produced by an HV or EV without an used was the same one that was the only based on the tire and wind noise alert and the same vehicle with an ICE source of the agency’s minimum produced by the vehicle. This indicates because the predominant sound in both detection requirements in this final rule. that at 20 km/h it is unlikely that test conditions comes from the tires and We conducted this analysis after the pedestrians would be able to detect a NPRM comment period had closed to majority of EVs and HVs without an aerodynamic noise, and these factors are assist in considering the comments we alert sound. Therefore, according to this consistent for both test conditions. had received. A technical paper on this data, a crossover speed of 20 km/h does NHTSA measured coast-down and crossover speed analysis has been not meet the requirements of the PSEA. pass-by sound pressure levels for eleven included in the docket.103 At 30 km/h, four models had multiple different ICE vehicles at 10, 20 and 30 By applying the detectability model to bands that met or exceeded thresholds, km/h test speeds. The results are shown the measurements of sounds produced and another seven models met or in Table 8.

103 Quiet Car Coast Down Analysis (Final Rule) important one is that the vehicle data is for coasting operation. Consequently, this evaluation should not (June 2015). ICE vehicles (because the goal is to measure tire and be used to judge the sound level in actual operation 104 There are several important caveats in the use wind noise), and thus it does not include the engine of any of the test vehicles. Other caveats are of this crossover speed analysis. The most noise that the test vehicles would have in normal enumerated in the docketed analysis paper.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00034 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90449

TABLE 8—PASS-BYVS. COAST-DOWN MEASUREMENTS FOR ELEVEN VEHICLES AT 10, 20, AND 30 KM/H 105

Overall SPL (dBA) 10 km/h 20 km/h 30 km/h Pass-by Coast-down Pass-by Coast-down Pass-by Coast-down (engine on) (engine off) (engine on) (engine off) (engine on) (engine off)

1 ...... 2012 Toyota Camry ...... 57.8 48.4 62.1 60.3 67.2 66.6 2 ...... 2012 Toyota Corolla ...... 56.5 48.5 61.4 59.8 67.2 66.6 3 ...... 2012 VW Golf ...... 57.0 49.4 62.3 60.9 68.3 67.4 4 ...... 2012 Mini Cooper ...... 58.7 50.8 65.6 59.9 68.3 67.2 5 ...... 2011 Cadillac CTS ...... 56.7 50.4 62.0 60.2 68.1 66.7 6 ...... 2012 Toyota Yaris ...... 56.1 46.2 59.9 57.8 65.1 64.4 7 ...... 2012 Honda Fit ...... 56.6 48.3 62.2 59.3 66.6 66.1 8 ...... 2010 Buick Lacrosse ...... 55.8 49.9 63.8 60.4 68.4 66.7 9 ...... 2011 Honda Odyssey ...... 56.5 52.2 63 62.4 69.4 68.8 10 ...... 2012 Lexus RX 350 ...... 59.7 48.1 61.7 60.1 67.3 66.5 11 ...... 2012 Ford Focus ...... 57.5 49.3 62.6 60.8 68.0 67.1 Average ...... 57.2 49.2 62.4 60.2 67.7 66.7

Table 9. As explained in the NPRM,106 such that they could be directly From these data, coast-down differences in sound pressure level of compared). Based on this measurements were subtracted from less than 3dB generally are not understanding, differences identified in pass-by measurements to determine if, distinguishable to humans (differences Table 9 of less than 3 dB would indicate and at what speed, crossover occurred of 3dB might be noticeable only if two that the vehicle crossover speed has for each vehicle. The data are shown in sounds were heard one after the other been achieved.

These results indicate that at the speed by 30 km/h (when using the not reach the crossover speed as defined vehicle speed of 10 km/h all eleven criterion that the results from the two by the agency. These two vehicle vehicles had coast-down sound pressure test conditions are within 3 dB.) models had the highest pass-by sound levels significantly less than their The results at 20 km/h were less pressure levels of the eleven vehicles, associated pass-by levels, meaning that conclusive. Of the eleven vehicles and their coast-down sound pressure none of the vehicles had attained its tested, all had coast-down sound was close to the average level for all respective crossover speed. At 30 km/h, pressure levels below their respective eleven vehicles. While we note that it is all eleven vehicles had coast-down pass-by test levels. However, all but two possible to interpret this narrow data sound pressure levels close to or within of the vehicles got to within a 3-dB sample as demonstrating that a lower 3 dB of their associated pass-by levels, differential, and the average differential crossover speed may be sufficient for a meaning that every vehicle had reached of all vehicles was 2.2 dB. The two portion of the HV/EV fleet, we also its respective crossover speed. Thus, the vehicles that did not were the Mini and conducted additional analysis and additional testing clarified that 10 km/ Buick Lacrosse, which had sound considered additional factors in arriving h would not be sufficient and that all differentials greater than 3 dB (5.7 dB at our decision to maintain the approach vehicles would reach their crossover and 3.4 dB, respectively) and thus did to require the pedestrian alert sound up

105 Garrott, W.R., Hoover, R.L., Evans, L.R., Hybrid Electric, and Low Speed Vehicles’’ 106 see NPRM, 78 FR 2838. Gerdus, E., and Harris, J.R., ‘‘2012 Quieter Vehicle Washington DC, DOT/NHTSA, November 2016. Testing Report: Measured Sound Levels for Electric,

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00035 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 ER14DE16.001 90450 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

to 30 km/h, provided that vehicles are agency also concludes that a crossover premise that they do not contribute as not able to satisfy the performance speed of 30 km/h (18 mph) will ensure much to detection. In addition, alert requirements without an alert sound. that EVs and HVs will produce system devices, particularly speakers, This comparison of the engine-on and sufficient sound to allow pedestrians to that are able to produce high level, low- engine-off measurements for these safely detect them during low-speed frequency sounds would most likely vehicles does not directly answer the maneuvers in which these vehicles have to be larger, heavier, and more question of when a vehicle makes would otherwise pose a risk to costly. Specifications for the low-to- enough sound to be detected by pedestrians because of the low sound mid-range frequency bands between 315 pedestrians. We believe that it also level they produce. Because we believe and 500 Hz were included to assist demonstrates that at 20 km/h there is a that drivers may execute these low pedestrians in detecting HVs and EVs in question of whether some vehicles speed maneuvers at speeds up to at least ambient noise environments such as produce enough sound based on tire 30 km/h, and these maneuvers represent areas near construction activity with and wind noise alone to be detected by the highest risk of crash between an EV significant high frequency noise. In the pedestrians. or HV and a pedestrian, more injuries NPRM, the agency omitted mid- Other factors we considered include will be avoided due to this rule with a frequency bands from 630 to 1600 Hz the difference in pavements crossover speed of 30 km/h than with a because many common ambient encountered in traffic compared to the crossover speed of 20 km/h. conditions include frequencies within ISO sound pad that is needed for As a further consideration, we note this range. One-third octave band testing, and the use of tires with low that a vehicle is not required to have standards in this range would have to be rolling resistance. The test data used to added alert sound at any speed at which set at a relatively high level to evaluate crossover speed were obtained it meets the minimum detection effectively compensate for the masking on an ISO sound pad with a specified requirements in this final rule. It would effects caused by ambient noise asphalt pavement. On public roadways, be acceptable for an alert system to be conditions. But these bands contribute varying pavement conditions will be designed to turn off at some speed more than other bands to a vehicle’s encountered that can increase or below the 30 km/h crossover speed if it overall alert sound level for the same decrease a vehicle’s acoustic sound could be demonstrated that, between increase in detectability. By omitting profile. Also, low rolling resistance tires that lower cut-off speed and 30 km/h, it minimum requirements for the one- may tend to increase vehicle sound meets the detectability specifications third octave bands in the 630 to 1600 Hz profiles, but not all vehicles will be without the assistance of an alert frequency range in the proposal, the operated with low rolling resistance system. agency was attempting to ensure that tires. While these factors could increase E. Acoustic Parameters for Detection of alert sounds allow pedestrians to safely vehicle noise, they also might decrease detect nearby EVs and HVs without it. Selecting the higher crossover speed Motor Vehicles unnecessarily increasing overall would ensure safety is not compromised In the NPRM, the agency proposed ambient noise levels.108 The high- when real-world roadway conditions minimum sound levels for a specific set frequency bands up to 5000 Hz provide result in the latter case. 107 of one-third-octave bands that good detectability for pedestrians with Another consideration is that included low-to-mid-frequency bands limitations in available crash data do normal hearing. (315, 400, and 500 Hz) as well as high- The proposed sound specifications not permit the agency to make frequency bands (2000, 2500, 3150, determinations regarding safety benefits were based on a psychoacoustic 4000, and 5000 Hz) for various vehicle modeling approach in combination with at specific speeds. Because the vehicle operating conditions including speed at the time of a crash into a safe detection distances. The inherent stationary, reverse and forward motion assumptions for this analytical approach pedestrian is not available in the data up to 30 km/h. These one-third octave set, the agency is not able to quantify were that: 109 bands were selected in an effort to • A vehicle should be detectable in what portion of the safety benefits maximize the detectability of the associated with today’s final rule would the presence of a moderate suburban proposed alert sounds while taking into ambient, i.e., ambient at 55 dB(A); 110 be lost if we were to adopt a value for consideration the masking effects of • a psychoacoustic model can be used crossover speed below the real-world common ambient noise and the to determine minimum levels for values for some specific vehicle models. degraded hearing of some pedestrians. However, we continue to believe that detection of one-third octave bands in Specifying minimum sound pressure this rule will prevent some the presence of an ambient; levels for a wide range of one-third unqualifiable number of additional • octave bands means that sounds sounds should be detectable in injuries by adopting a 30 km/h meeting the specifications will be multiple one-third octave bands to crossover speed as opposed to a 20 km/ detected in a wider range of ambient increase the likelihood that a pedestrian h crossover speed. As discussed conditions with various acoustic will be able to detect the sound in previously, our crash analysis indicated profiles. multiple ambients with differing that the odds ratio of an HV being Low frequency bands (below 315 Hz) acoustic profiles; and involved in a crash with a pedestrian were not included in the proposed was 1.52 when the vehicle in question specifications due to the expected 108 NPRM, ‘‘Federal Motor Vehicle Safety was executing a low speed maneuver Standards; Minimum Sound Requirements for strong masking effects of the ambient Hybrid and Electric Vehicles, 78 FR 2829, (Jan. 14, immediately prior to the crash. This noise at low frequencies and the means that HVs and EVs are 52 percent 2013). 109 Hastings, et al. (2012). Research on Minimum more likely to be involved in an 107 Octave band and one-third octave band scales Sound Specification for Hybrid and Electric incident with a pedestrian than an ICE facilitate identifying the specific frequencies of Vehicles. Docket NHTSA–2011–0148–0048. vehicle under these circumstances. sounds. Octave bands separate the range of 110 In the NPRM we stated that we chose an Low-speed maneuvers include making a frequencies audible to humans into ten bands, and ambient with a 55 dB(A) overall sound pressure the one-third octave bands split each of the ten level because this represented a reasonable level turn, slowing or stopping, backing, octave bands into three smaller frequency bands. below the 60 dB(A) ambient in which pedestrians entering or leaving a parking space or Each scale in the breakdown provides more would no longer be able to reasonably rely on driveway, and starting in traffic. The information about the sound being analyzed. hearing to detect approaching vehicles.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00036 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90451

• minimum detection distances can would be unnecessarily loud for most be based on vehicle stopping distances conditions. and driver reaction times. Based on calculations using these The agency used Moore’s Partial Where: values, the agency determined that the 111 Loudness model to estimate the t = brake reaction time, sec. desired detection distances were 5 V = design speed, km/h minimum sound levels needed for a meters in front of the vehicle for the 10 sound to be detectable in the presence a = deceleration rate, m/s2 km/h (6.2 mph) pass-by, 11 meters for of an ambient. The first step in our We explained that we chose a reaction the 20 km/h (12.4 mph) pass-by, and 19 approach was to determine the time of 1.5 seconds because that is the meters for 30 km/h (18.6 mph) pass-by. minimum levels for detection, using mean reaction time for surprise The results of these computations were Moore’s model and a simplified events 113 such as an object suddenly ambient, for a pedestrian at the vehicle moving into a driver’s path. We chose rounded to the nearest meter. Moore’s location. We stated that the distance at the 5.4 m/s2 deceleration rate Partial Loudness Model was then used which a pedestrian would need to hear corresponding to dry pavement braking to derive the minimum sound levels a vehicle is at least as long as the because most of the pedestrian crashes required for detection for each driving distance travelled during the driver’s that the agency identified occurred in condition and one-third octave band. reaction time, plus the vehicle’s clear conditions. If we had decided to Levels were increased by 0.5 dB to stopping distance. We calculated these use instead a slower deceleration rate provide a small safety factor, and were distances from the guide on highway for wet pavement conditions, we believe then rounded up to the nearest integer design 112 of the American Association the necessary sound profile for for simplicity. The resulting NPRM of State Highway Transportation detection would have to be louder and levels are shown in Table 10. Officials (AASHTO) according to the for a longer period because it would following formula: take a greater distance to stop, and thus

TABLE 10—NPRM MINIMUM SOUND LEVELS FOR DETECTION

Stationary but One-third octave band center frequency, Hz activated Backing 10 km/h 20 km/h 30 km/h

315 ...... 42 45 48 54 59 400 ...... 43 46 49 55 59 500 ...... 43 46 49 56 60 2000 ...... 42 45 48 54 58 2500 ...... 39 42 45 51 56 3150 ...... 37 40 43 49 53 4000 ...... 34 36 39 46 50 5000 ...... 31 34 37 43 48 Overall A-weighted SPL Measured at SAE J2889–1 PP’ line ...... 49 52 55 62 66

We explained in the NPRM that while because of practical reasons we would need to be 2 meters from the vehicle’s we were setting the sound pressure measure sound emission for compliance path to be detected within the levels for each one-third octave band purposes at a distance of 2 meters and prescribed stopping distance. Table 11 based on the distance from the vehicle scale the required levels accordingly. shows how the sound produced by a at which we wanted pedestrians to be We used the following method to vehicle attenuates when measured using able to hear approaching vehicles, calculate what the sound level would the procedure in SAE J2889–1.

TABLE 11—SPL ADJUSTMENT (dBA) FROM SOURCE TO SAE MICROPHONE LOCATION

Speed, km/h ...... 10 20 30 X source, meters ...... 5 11 19 Y source,* meters ...... 2 2 2 r0,** meters ...... 2.3 2.3 2.3 r1,** meters ...... 5.5 11.2 19.1 r doubling ...... 1.2 2.3 3.0 Attenuation, dB ...... ¥5.8 ¥12.3 ¥16.8 * Assume effective source is at center of vehicle since propagation is forward. ** Assume Z = 1.2.

‘X’ represents the horizontal distance J2889–1. ‘Z’ represents the height of the from the source to the P–P’ line while microphone in meters as specified in ‘Y’ is the 45perpendicular distance from SAE J2889–1. The values in Table 11 the source to the microphones in SAE were calculated using the following

111 Moore, B.C.J., Glasberg, B.R., and Baer, T. 112 American Association of State Highway and 113 Green (2000) How Long Does It Take to Stop? (1997). A Model for the Prediction of Thresholds, Transportation Officials, A Policy on Geometric Methodological Analysis of Driver Perception-Brake Loudness and Partial Loudness, J. Audio Eng. Soc. Design of Highways and Streets, Chapter 3 Elements Times.’’ Transportation Human Factors 2(3) 195– 45, 224–240. of Design (2004). 216.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00037 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 ER14DE16.002 90452 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

formula and assuming a value of 1.2 meters for Z:

In the NPRM, the agency also one-third octave bands. On the topic of required in a low frequency range (less indicated its intent to conduct acoustic parameters for detection, the than 1000 Hz) and one band would be additional research before issuing a final agency received a joint comment from required in a high frequency range (1000 rule to confirm that sounds meeting the Alliance/Global, as well as comments Hz up to 3150 Hz), separated by at least proposed requirements would be from OICA, Chrysler, Ford, GM, Honda, one one-third octave band. Alliance/ detected as predicted by the model, and Mercedes, Nissan, Porsche, Toyota, the Global suggested the following levels we sought comments on the following National Federation of the Blind, the (Table 12) but noted that further topics (NPRM pp. 2832–2833): American Council of the Blind, the discussion within the QRTV group that • What improvements would make World Blind Union, the National is developing a GTR is needed before the acoustic specifications more Council of State Agencies for the Blind, these values can be fully recommended: effective and make alert sounds more the Disability and Communication detectable? Access Board, the Insurance Institute for TABLE 12—ALLIANCE/GLOBAL • Should NHTSA require vehicles to Highway Safety, Advocates for Highway RECOMMENDED TWO-BAND LEVELS emit sound that meets the four one-third and Auto Safety, Accessible Design for octave band requirements only at 2000 the Blind, and Western Michigan Test condi- Individual Hz and above as an alternative to University. Subsequent to the NPRM 116 Overall SPL band SPL tion (two bands) requirements for eight one-third octave comment period, NHTSA also received bands? a late comment submitted jointly by the Stationary/ 48 dB ...... 44 dB • What is the optimum number of Alliance, Global, the NFB, and the ACB, Backing. bands that should contain minimum and the agency had additional 10 km/h ...... 53 dB ...... 46 dB sound level requirements, and what correspondence with those commenters, 20 km/h ...... 58 dB ...... 51 dB should the corresponding levels be? which is recorded in the docket. In addition to requirements with Four main issues were discussed by Alliance/Global stated that NHTSA’s minimum content in the eight one-third the commenters relating to the acoustic target for detectability performance can octave bands between 315 Hz and 500 parameters proposed for detection: (1) be achieved with two one-third octave Hz and 2000 Hz and 5000 Hz, the NPRM The number and level of one-third bands set at the levels proposed in the also considered acoustic requirements octave bands required; (2) the methods NPRM, and the minimum levels for with minimum content in two one-third used to determine detection distances additional bands can be reduced while octave bands with a minimum and associated sound specifications; (3) maintaining the same detectability requirement for the overall sound the range of frequencies used; and (4) performance. Alliance/Global stated that pressure level of the sound. NHTSA vehicle marketability. if NHTSA chooses to require in the final stated, when discussing this possible Fifteen of the above commenters rule that sounds emitted by EVs and two-band approach in the NPRM, that it discussed the first issue about the HVs must have content in more than was seeking comment on the acoustic number and levels of one-third octave two one-third octave bands, the agency profile of the minimum sound bands required. Alliance/Global 115 should reduce the minimum levels for requirements, as well as on the number stated that NHTSA’s proposed each one-third octave band according to of one-third octave bands for which the specification in the NPRM is too the total number of required bands. agency should establish requirements. conservative. They suggested deleting Chrysler, GM, Honda, and Mercedes We stated in the NPRM that the reason the requirement for frequency content in stated that they support the two-band we were not proposing to adopt eight one-third octave bands and approach suggested by Alliance/Global. requirements for content in two one- replacing it with a simplified two-band Ford argued that based on its study of third octave bands was that a sound approach. Specifically, they this subject, not all eight one-third with content in only two one-third recommended using a minimum overall octave bands are needed for a sound to octave bands would not be detectable in SPL and minimum sound levels in at be detectable 5 meters away. Ford’s as many ambient noise environments as least two octave bands. In their study consisted of a human factors test sounds with minimum content in eight suggested approach, one band would be where audio recordings of vehicle

114 Attenuation rate = 4.5 dB for the first distance 116 The Alliance/Global recommendations did not accordance with their comments that crossover doubling and 6 dB per distance doubling thereafter. include suggested minimums for 30 km/h in speed should be limited to 20 km/h. 115 NHTSA–2011–0148–0251.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00038 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 ER14DE16.003 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90453

sounds were presented to participants dB(A).117 The test vehicle was a Toyota a pedestrian with hearing loss content at using headphones. Sounds tested by Prius V approaching at 20 km/h. The lower frequencies is needed and that Ford were an ICE vehicle sound, an study indicated that the overall level of potential sounds should have a fairly electric vehicle without an alert sound, the test vehicle was 58 dB(A) with broadband frequency spectrum. WMU and three alert sounds, but those sounds sound energy in multiple bands. The suggested that identifying two frequency did not meet all of the agency’s sound level in the 800 Hz and 2000 Hz bands that are most useful for detection, proposed minimum one-third octave bands were each 51 dB(A), which similar to Nissan’s approach, may be bands levels. Sounds were mixed with accounted for nearly half of the sound’s appropriate. a 55 dB(A) masking noise. Twenty-four acoustic energy. Toyota reported that As mentioned above, NHTSA also Ford employees and four visually the measured detection distance received a joint letter, submitted to the impaired individuals participated in the exceeded the NHTSA target detection docket and treated as a late comment, distance in the NPRM for this operating from the Alliance, Global, the NFB, and study. Ford stated that all vehicles were 119 detected before the 5-meter critical condition. the ACB. These commenters agreed OICA stated that the proposed on several technical and policy issues. distance, except for the vehicle without specification for eight bands will force They stated that the number of bands an alert. They also reported that very loud devices with unpleasant should be reduced from a minimum of participants recognized the vehicles sounds. They suggested that the sound eight to at least two, between 160 Hz with alert sounds at least at the same specifications within the UNECE–GTR and either 3150 or 5000 Hz, and that at rate as the ICE vehicle sound. development group. They stated that least one band should be below either Nissan stated that a sound with a NHTSA should consider requiring a 1000 or 1600 Hz. Within each sound pressure level equivalent to the specific number of tones which could be individual frequency band, they stated ICE fleet minimum with a two-peak in the same one-third octave band, that sound levels should be revised with sound profile is appropriate for rather than requiring a specific number input from available research. They also detectability. Nissan stated that having of one-third octave bands.118 suggested establishing limits on overall one peak frequency component between The American Council of the Blind sound pressure level, but did not 600 and 800 Hz helps detectability for (ACB) stated that the most appropriate provide specific values. aging pedestrians with high frequency approach to the sound specifications The second main topic discussed by hearing loss. A second peak frequency would be to set the minimum sound the commenters concerned the methods component between 2000 and 5000 Hz level based on the levels produced by used by the agency to determine would provide detectability for light ICE vehicles because this is the detection distances and associated pedestrians with normal hearing. Nissan sound pedestrians currently use for safe sound specifications. Eleven of the also suggested that the required navigation. ADB stated ‘‘octave bands commenters listed above provided frequency content of alert sounds at are not as great at predicting detection comments on this topic. In their joint comment, the Alliance, around 1000 Hz (the typical frequency as overall sound levels’’ based on Global, NFB and ACB agreed with the for road traffic noise) should be reduced research conducted by WMU. WMU detection distance methodology in the to avoid additional contribution to stated that their research has shown that NPRM and with the values used for the traffic noise. individual octave bands are not as useful in determining detection as is the deceleration rate and the brake reaction Porsche stated that the specified overall sound level and that, while some time. The World Blind Union (WBU), levels in the NHTSA proposal will lead regulatory direction in octave band the National Council of State Agencies to very loud and unpleasant alert make-up of alert sounds might be for the Blind (NCSAB), the Disability sounds. They suggested specifying at useful, there is limited justification for and Communication Access Board, and least two bands, but allowing up to eight a requirement as restrictive as the the Insurance Institute for Highway bands. Porsche explained that the levels NHTSA proposal. WMU stated that their Safety, all agreed that the methodology to be met should be a function of the previous research had shown a limited used by NHTSA to set the minimum number of bands selected. They advantage for content in the 500 Hz sound levels seemed reasonable and explained that if more bands are used, band in some situations, and their appropriate. OICA stated that the NPRM the levels per band can be lower to statistical analysis showed significant approach to establish detection distance achieve the same detectability. They predictive value for overall sound as a function of vehicle speed is suggested that, for example, if eight pressure levels rather compared to reasonable but only when applied to the bands are used, then the levels in each content in any particular band. WMU overall sound pressure level. band should be reduced by 6 dB (e.g., also commented that detecting a single Advocates for Highway and Auto the agency’s proposed minimum level of approaching vehicle may not be the Safety also generally agreed with 43 dB(A) for the 500 Hz one-third octave same as detecting quiet vehicles when specifications based on detection band for the stationary condition would other vehicles are present. In response distance. They commented on the driver be reduced to 37 dB(A)), and if four to the request for comments on reaction time used in the detection bands are used, the levels in each band requiring vehicles to emit sound that distance computation and suggested should be reduced by 4 dB. meets only the one-third octave band that the 1.5 sec. used by NHTSA may be requirements for 2000 Hz and above as too short. They indicated that NHTSA Toyota supported the use of an overall should examine reaction times for level and at least two one-third octave an alternative to meeting all eight one- third octave bands, WMU stated that for drivers in relation to pedestrians and bands, consistent with the Alliance/ pedalcyclists in establishing this value. Global recommendation. Toyota 117 The Toyota comment did not include details Accessible Design for the Blind (ADB) provided results from a study that it about the spectral shape of the ambient, which expressed support for the NPRM conducted to confirm the detectability would be important to better understand the approach to minimum sound levels but performance of the suggested approach. possible masking conditions and their impact on questioned the detection distance used In that study, 33 individuals (from 20 to the test vehicle alert sound acoustic profile. 118 in NHTSA’s analysis. ADB questioned 49 years old) participated. The ambient We note here that this suggestion could result in an alert signal with only one distinct component, noise level varied from 51 to 59 for example, a single amplitude-modulated tone. 119 NHTSA–2011–0148–0322

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00039 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90454 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

whether the detection distance used in Agency Response to Comments to determine whether the model NHTSA’s formulation represents Detectability Model Conclusions accurately predicted when sounds distances that are sufficient for would be detected by human listeners at pedestrians to detect, recognize, judge After considering all comments the distances predicted by the model. distance and trajectory, decide to received in response to the NPRM, and Volpe conducted a human factors the results of agency research conducted initiate a crossing, and initiate a study to quantify differences between since the NPRM was issued, we have crossing, particularly at busy predicted detection levels (as indicated decided to modify the proposed intersections. They also indicated that by Moore’s Partial Loudness model) of minimum specifications for detection of vehicle sounds in the presence of a the specifications proposed in the vehicles subject to this rule. While the NPRM are based on the detection of a standardized ambient used to calculate number of one-third octave bands for the minimum requirements proposed in single vehicle in the absence of other which the agency is establishing the NPRM and actual responses of vehicles, which they believe is not requirements for minimum content and participants listening to these vehicle realistic. the requirements related to detection of sounds through headphones.121 The WMU indicated that the detection changes in vehicle speed differ from the study also evaluated the effect of several distance used in the development of the NPRM, the underlying analytical factors on detectability, including the sound specification may be too short framework on which the minimum number of one-third octave band because it may not correspond to the acoustic requirements in the final are components contained in a sound, time needed to detect a vehicle, process based has not changed. The minimum adjacency of bands, and signal type the information, and decide to take acoustic requirements for each one-third (e.g., pure tones, bands of noise). Fifty- action. WMU explained that the octave band in the final rule remain two demographically diverse subjects detection distance formula used does based on the same formula used to were exposed to a simulation of a not account for variability among develop the requirements proposed in vehicle passing by them (as a pedestrians including those with the NPRM albeit with slightly different pedestrian) at 10 km/h, in ambient noise inputs to that formula. Furthermore, the hearing loss. conditions of 55 dB(A). In the study, a overall sound pressure level and one- selection of 24 different sound signals On the third issue about the range of third octave band levels of sounds were played back over the participants’ frequencies used, the Alliance/Global, meeting the requirements of the final headphones. The signals were based on OICA and NFB provided comments. rule will be similar to the corresponding synthesized and recorded sources and Alliance/Global said that one-third levels of sounds meeting the eight one- included pure tones, single noise bands, octave bands from 630 to 1600 Hz third octave band requirements in the multiple adjacent noise bands, multiple should not be excluded from the useable NPRM. non-adjacent noise bands, tones mixed range as NHTSA did in the NPRM After considering the comments and with noise, a signal based on a recorded because ‘‘these frequencies will clearly the agency’s further evaluations ICE, and signals from prototype alert contribute to the detectability.’’ OICA conducted in response to comments, we systems. Signals with various numbers recommended that no sound be required decided to reduce the number of one- of bands were included in the study, above 2 kHz as they believe that is not third octave bands for which we are ranging from one to four non-adjacent representative of vehicle sounds. OICA requiring content from the eight one- bands and from one to twenty-four stated that manufacturers should be third octave band requirement proposed continuous or semi-continuous bands. allowed to use the range from 125 Hz to in the NPRM to either a four one-third With the exception of the ICE vehicle 3000 Hz and suggested that low octave band compliance option or a two sound, the two recorded prototype alert frequencies could aid with detectability one-third octave band compliance signals, and the three two-band samples, but may have cost implications. OICA option, the latter including an overall all signals were calibrated to just meet recommended that low frequencies SPL specification. the NPRM specifications for safe Under the four one-third octave band should be an option for manufacturers detection in each band with signal compliance option, the minimum sound and if used, believe the regulatory content.122 requirements for each band would be scheme should give credit to The study results indicated that, slightly lower than the values proposed except for frequency sensitivity of high manufacturers for using low in the NPRM, and the overall sound 120 frequency components, the modeling frequencies. NFB stated that pressure of sounds meeting the four manufacturers should have flexibility to approach for determining the minimum one-third octave band compliance level needed in each one-third octave create sounds that are pleasant and not option will be similar to those meeting annoying to vehicle occupants and band was conservative, meaning that the the proposed requirements for eight participants responded to signals requested that the agency consider not bands in the NPRM. Under the two one- requiring sound in the lowest one-third third octave band compliance option, 121 Hastings A.; and McInnis, C. ‘‘Detectability of octave bands. NFB stated that the minimum sound requirements for Alert Signals for Hybrid and Electric Vehicles: manufacturers can limit the sound each band are lower than those in the Acoustic Modeling and Human Subjects inside the vehicle and meet the safety eight one-third octave band proposal in Experiment,’’ (2015) Washington, DC: DOT/ need of pedestrians without including NHTSA. the NPRM for the low and mid 122 The NPRM did not include specifications for content in each of the eight proposed frequency bands and higher than the the one-third octave bands from 630Hz–1600Hz. one-third octave bands. minimum values in the NPRM for the Some alert signals considered by Volpe during the high frequency one-third octave bands human factors study did include one-third octave The fourth main issue raised in bands in this range. Volpe derived the appropriate comments relates to vehicle centered at 4000 Hz and 5000 Hz. level for those bands the same way the minimum marketability. These comments are In the NPRM, NHTSA stated that it levels for the bands included in the NPRM were addressed in section III.I of this notice. planned to conduct additional research developed. For details, refer to the Volpe research once the NPRM was issued to validate report, Hastings A.; and McInnis, C. (2015). ‘‘Detectability of Alert Signals for Hybrid and 120 No explanation was provided by OICA about the model used to develop the Electric Vehicles: Acoustic Modeling and Human how or why vehicle manufacturers should be given minimum sound requirements in the Subjects Experiment’’. Washington, DC: DOT/ credit for using low frequencies. NPRM. The purpose of this research was NHTSA.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00040 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90455

somewhat sooner on average than the minimum acoustic content in two one- pressure level of 55 dB(A) without model predicted. With an third octave bands. In response to these affecting the spectral variation. Volpe understanding that the model was comments and the joint comment then used the adjusted acoustic model conservative overall but less accurate at submitted by the Alliance, Global, NFB to test how signals with different the higher frequencies, model and ACB recommending that the agency numbers of components perform across adjustments were made as discussed in require minimum content in only two this wide variety of ambient conditions. section II.C of this preamble to provide bands, NHTSA decided to conduct This approach of testing signals in more accurate results necessary for additional analysis to determine the varying ambient conditions but at a development of the final minimum one- likelihood that sounds with content in consistent overall level allowed us to third octave band levels specified in this fewer than eight bands would be determine the performance of signals as rule. masked in different ambient a function of the number of components Although not directly tested in the environments. in the signal. Specifically, this method study, we found a general trend that the The resulting analysis provided an provides a measure of ‘‘robustness’’ of minimum one-third octave band levels estimate of how often a sound signal the signal which is the metric we use to as proposed in the NPRM could be would be detected as a function of the gauge how likely it is that one or more reduced when increasing the number of number of one-third octave bands. Real- of the signal components will be heard one-third octave bands. We also found world ambient conditions are not by pedestrians in a range of ambient that using non-adjacent one-third octave consistent, and we wish to draw conditions. bands instead of adjacent bands conclusions about detectability beyond NHTSA’s approach in evaluating maintained the detectability of sounds the standardized 55 dB(A) ambient used various signals was to set the band more effectively while limiting the to create the proposed requirements in levels for each component at the overall level. Consequently, we have the NPRM. The ambient data used in appropriate psychoacoustic thresholds incorporated non-adjacency as one of this analysis was recorded at 17 according to the modified Moore’s the specifications in the final rule alert locations along Centre Street in Newton, model after the model had been 123 requirements. We have decided not to Massachusetts. Ambient samples adjusted using the results of Volpe’s adjust the minimum one-third octave were taken at intersections (signalized human factors experiment. The adjusted band levels to account for the number and stop-sign-controlled), one-way acoustic model was used to measure the of required bands because in this final streets, side streets, and driveways. performance of signals having various rule we have reduced the number of Samples had a mix of low, mid, and numbers of frequency components from required bands from eight bands to high frequencies. Some samples were one up to seven one-third octave bands either two or four bands. dominated by low frequency content, by evaluating how readily each signal The study results also indicate that i.e., the environment had other vehicles was detected in the presence of a broad sounds with minimum content in eight, in close proximity operating at and/or range of measured ambients normalized four, and two one-third octave bands accelerating from low speeds, while to the 55 dB(A) level. were all detected by study participants other samples were dominated by high prior to the two-second time-to-vehicle frequency content, i.e., the environment The ambient used also had a standardized one-third arrival point necessary for safety. had other vehicles in close proximity octave band frequency composition. To analyze the As discussed above, NHTSA received operating at higher constant speeds. robustness of various alerts, the multiple ambients several comments from manufacturers Each ambient sample was collected had various overall SPLs, either less than and groups that represent manufacturers 124 or greater than 55dB, and various frequency normalized to an overall sound compositions. For a proper evaluation of the stating that agency should adopt the various ambients, each ambient’s overall SPL had acoustic requirements with content in 123 Ambient data were collected in 2010 to be normalized, that is adjusted to 55 dB, while two one-third octave bands plus a (Hastings, et al. 2011). Walkthroughs were maintaining each individual sample’s unique requirement for a minimum overall conducted with different orientation and mobility frequency profile. To normalize each ambient instructors; data were collected on different days of sample, the sample was broken down into its one- sound pressure level discussed in the the week and time of day. third octave band levels and then each level was NPRM. These commenters believed that 124 Each ambient sample had to be normalized to decreased or increased the same percentage until NHTSA’s goal in the NPRM of ensuring an overall SPL of 55dB(A) to ensure a comparable the overall level for that particular ambient sample that sounds produced by hybrid and analysis was conducted for detectability utilizing equaled 55dB(A). For consistent comparisons of different numbers of one-third octave bands. As vehicle alert sounds in these different ambients, the electric vehicles are detectable to discussed in the NPRM and this final rule, a key data was the frequency composition, or acoustic pedestrians in a variety of ambients standardized 55dB(A) ambient was used to derive profile, across the one-third octave bands for each could be accomplished by requiring the minimum one-third octave band specifications. ambient collected.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00041 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90456 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

Figure 2 shows the ‘‘robustness’’ 125 of requirements of the NPRM were In order to comply, the alert signal single and multiple one-third octave resistant to masking, NHTSA is must meet or exceed the given levels in band alert specifications, and includes reducing the number of bands in at least four non-adjacent bands for each up to seven bands because that is the response to comments suggesting that given vehicle operating condition. Also, maximum number that can be non- requiring minimum content in eight the four bands must span a range of at adjacent over the 315 to 5000 Hz range. one-third octave bands it not necessary least nine one-third octave bands. This analysis shows that, on average, for safety. As the latest NHTSA research NHTSA believes that the four one-third signals with minimum content in four demonstrated, reducing the number of octave band compliance option achieves one-third octave bands can be detected bands with minimum requirements the goals articulated in the NPRM of in 97 percent of ambient environments from eight to either four or two one- ensuring that sounds meeting this examined. This analysis also shows that third octave bands would not impact the standard are detectable in a variety of sounds with content in only two one- effectiveness of sounds meeting the ambients and responds to comments third octave bands show strong minimum requirements of the final rule submitted to the NPRM claiming that resistance to masking if the minimum in providing alerts to pedestrians. the requirements in the NPRM were too content is in certain bands. We believe that the four-band restrictive and would require Additionally, this analysis shows that requirements and the two-band unpleasant sounds. sounds with content in more than four requirements have equivalent Because of the number of comments one-third octave bands are only performance in terms of detectability by received on this issue, NHTSA also marginally more resistant to masking pedestrians and will be equally decided to explore allowing the two than sounds with four bands. Based on detectable in a variety of different one-third octave band compliance this analysis, NHTSA agrees with the ambients. option discussed in the NPRM. Under commenters that the agency can Under the four-band compliance the two-band compliance option, accomplish the goals articulated in the option, the agency is requiring that the minimum sound pressure levels are NPRM of ensuring that sounds four bands used to meet the required in two non-adjacent one-third produced by EVs and HVs are detectable detectability requirements must be non- octave bands from 315 to 3150 Hz. One to pedestrians in a variety of ambients adjacent one-third octave bands in the of the two bands must be below 1000 Hz by requiring minimum content in fewer frequency range from 315 Hz to 5000 and the second band must be at or above than eight one-third octave bands. Hz. This range includes the eight one- 1000 Hz. The two bands used must each Given that the rationale for specifying third octave bands for which we meet the minimum requirements and minimum content in eight one-third proposed requirements in the NPRM. In together must also meet a specified octave bands in the NPRM was to response to comments, NHTSA has overall SPL. ensure that sounds meeting the decided that the final rule will also By including both a four-band allow manufacturers to comply with the specification and a two-band 125 We use the term ‘‘robustness’’ to indicate how minimum acoustic requirements by specification in this final rule, NHTSA resistant a signal is to masking by background noise from a wide selection of different normalized placing acoustic content in the mid- is providing vehicle manufacturers with ambient conditions covering a range of spectral range frequency bands excluded from the flexibility to choose either content. the NPRM. compliance option in the new safety

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00042 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 ER14DE16.004 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90457

standard. We believe this approach the minimum one-third octave band levels indicated by the responses of the adequately addresses a great majority of levels as a result of our human factors experiment participants when listening comments concerning the eight-band testing and acoustic model adjustments to various signals (see Figure 1) Table 13 detectability specification proposed in discussed above. As explained, these provides the final rule minimum one- the NPRM. slight changes provide better agreement third octave band levels for each In addition, based on the foregoing, between the modeled levels and the operating condition.126 we have implemented slight changes to

TABLE 13—FINAL RULE MINIMUM SOUND LEVELS FOR DETECTION

One-third octave band center frequency, Hz Stationary Reverse 10km/h 20 km/h 30 km/h

315 ...... 39 42 45 52 56 400 ...... 39 41 44 51 55 500 ...... 40 43 46 52 57 630 ...... 40 43 46 53 57 800 ...... 41 44 47 53 58 1000 ...... 41 44 47 54 58 1250 ...... 42 45 48 54 59 1600 ...... 39 41 44 51 55 2000 ...... 39 42 45 51 55 2500 ...... 37 40 43 50 54 3150 ...... 34 37 40 47 51 4000 ...... 32 35 38 45 49 5000 ...... 31 33 36 43 47 Overall A-weighted SPL Range...... 43–47 46–50 49–53 55–59 60–64

The minimum one-third octave band agency chose to use a standardized the least possible impact on overall requirements in the final rule for the ambient 128 at a level of 55 dB(A) as an environmental noise while still eight one-third octave bands for which alternative to recordings of actual providing pedestrians with the vehicle the agency proposed requirements in traffic.129 Based partly on research sounds they need to navigate traffic the NPRM are slightly lower than the conducted by Pedersen et al. 2011,130 situations. values proposed in the NPRM for all test NHTSA selected an ambient with a 55 In this final rule, for the reasons conditions. Alert signals just meeting dB(A) noise level and a specific spectral discussed above, the agency has decided these requirements are expected to have shape (see Figure 2, p. 2818 in the to reduce the eight one-third octave overall levels similar to sounds meeting NPRM) that the Pedersen research had band requirement as proposed in the the proposed requirements of the found to be representative of many NPRM to a four one-third octave band NPRM, ranging from 43 to 47 dB(A) for common urban ambients. Because alert requirement. The agency is requiring stationary; 46 to 50 dB(A) for reverse; 49 sounds that are detectable in the that the four bands used to meet the to 53 dB(A) for 10 km/h; 55 to 59 dB(A) standardized 55 dB(A) ambient also detectability requirements must be non- for 20 km/h; and 60 to 64 dB(A) for 30 would be detectable in ambients with adjacent one-third octave bands in the km/h. similar spectral shapes and lower frequency range from 315 Hz to 5000 Hz As proposed, our detectability overall sound pressure levels, the 55 because the results of the human factors requirements were set so that EVs and dB(A) standardized ambient was study suggests that signals with non- HVs are detectable in an ambient with appropriate for detectability adjacent bands are more detectable than a 55 dB(A) overall sound pressure level. computations and was utilized signals with adjacent bands. Also, these It has been our understanding that throughout NHTSA’s development of bands must span a range of at least nine pedestrians who are blind use sound for the minimum sound levels included in one-third octave bands. This is navigation in environments for which this final rule. consistent with comments made by the ambient is at or below 55dB(A), and Our approach of using human subject Alliance/Global. Signal components in they rely on more than just sound when responses to set detection thresholds adjacent one-third octave bands can the ambient increases above that indicates how quiet alert sounds can be mask each other more effectively than level.127 The NPRM explained that, in before they can no longer be heard and signal components in non-adjacent one- NHTSA’s development of requirements ensures that the alert sound third octave bands. Masking reduces the for minimum vehicle sound levels, the requirements in the final rule will have effectiveness of the alert signal. Further,

126 These levels are based on a single one-third 127 In the NPRM we stated that we chose an 129 The NPRM included a lengthy discussion of octave band of noise producing a detectable signal ambient with a 55 dB(A) overall sound pressure how masking of vehicle sounds by ambient noise assuming a threshold of 0.079 sones per ERB for the level because this represented a reasonable level (also called background noise) is a fundamental maximum of the partial specific loudness which is below the 60 dB(A) ambient in which pedestrians factor in developing minimum vehicle sound levels. the threshold value that provides the best fit would no longer be able to reasonably rely on For research purposes, background noise can come between modeled detection times and those of the hearing to detect approaching vehicles. from recordings of actual traffic, but such recordings are likely to include random fluctuations experiment participants. The adjustments account 128 for model biasing for specific operating conditions, The standardized ambient is a ‘‘synthetic’’ or peaks from transient sources like the passage of repeatability/reproducibility as discussed in section background noise consisting of white noise filtered nearby traffic, construction noise, or aircraft that III.K of this final rule, and calculation rounding. For to have the same spectrum as what a pedestrian introduce variability when conducting human details see: Hastings A.; and McInnis, C. would hear in real traffic but without the variations factors testing or when applying detectability ‘‘Detectability of Alert Signals for Hybrid and in amplitude over time. This synthetic noise is models. Electric Vehicles: Acoustic Modeling and Human similar to actual traffic noise but is more consistent 130 Pedersen, et al. (2011). White paper on Subjects Experiment,’’ (2015) Washington, DC: and repeatable and thus is better suited to the external sounds for electric cars— DOT/NHTSA. acoustic research that NHTSA conducted. Recommendations and guidelines.

VerDate Sep<11>2014 19:19 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00043 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90458 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

four components that span nine bands typical ambient conditions encountered addresses comments about the will be more widely spaced than four by pedestrians, were excluded in the exclusion of those frequencies in the components in adjacent bands. This will NPRM in an effort to reduce the overall NPRM. In order to comply with the four increase the probability that pedestrians level since components in this one-third octave band compliance will be able to detect at least one signal frequency range would need to be set at option, the alert signal must meet or component. This is especially true for higher sound pressure levels. However, exceed the given levels in at least four pedestrians with age-related hearing our decision to require only four bands non-adjacent bands for a given operating loss. Signals in the mid-range one-third in the final rule and to include those condition. Figure 3 provides an example octave bands from 630 Hz to 1600 Hz, mid-range frequencies provides of a four-band signal. which are most strongly masked by the manufacturers with more flexibility and

In response to commenters who utilize the entire range from 315 to 5000 levels perform well above the average believe that sounds meeting the NPRM Hz will provide manufacturers with the and do achieve a high degree of requirements will be too loud and will flexibility to design alert sounds that are detectability in the range of ambients. contribute to increases in environmental non-intrusive and are acceptable to their For this reason we have determined that noise, we believe that our human factors customers. alert sounds with content in fewer than testing has confirmed our analysis in the four one-third octave bands can be NPRM that sounds produced by EVs Two One-Third Octave Band acceptable choices but need additional and HVs need to have content meeting Compliance Option specifications to ensure that they are as the minimum thresholds we have Because of the number of commenters detectable as signals with content in specified to ensure detectability. At the stating that the agency should adopt four or more bands. same time, the agency has determined final rule with minimum content The two-band alternative that the in its Environmental Assessment that requirements in two one-third octave agency is including in this rule closely the impact of alerts meeting the bands, NHTSA decided to explore a two matches the two-band approach requirements of this final rule are one-third octave band compliance suggested by commenters to the NPRM, expected to be negligible. option in addition to the four-band but with a few important differences Several auto manufacturers also compliance option discussed above. As which are discussed below. By commented that sounds meeting the shown in Figure 2 above, the average including both a four-band specification proposed requirements in the NPRM detectability of a vehicle sound in the and a two-band specification in this would intrude into vehicle interiors and presence of a range of ambients starts to final rule, NHTSA is providing vehicle be annoying to drivers. We believe that decrease if there are fewer than four manufacturers with the flexibility to reducing the number of required bands one-third octave bands with content at choose either alternative for compliance and including frequencies from 630 Hz threshold levels. However, Figure 2 also with the new safety standard. In this to 1600 Hz in the eligible range for shows that some of the signals with section of today’s preamble, we discuss compliance so that alert systems can fewer than four bands at threshold how the agency concluded that a two-

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00044 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 ER14DE16.005 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90459

band alternative is warranted and how In their NPRM comments, Alliance/ level along with a minimum level in we developed the two-band alternative Global suggested an acoustic two one-third octave bands.131 The using specifications suggested in NPRM specification for HVs and EVs that following were the particular levels they comments. consisted of a minimum overall sound recommended:

TABLE 14—LEVELS SUGGESTED BY ALLIANCE/GLOBAL

A-weighted dB Minimum level in each of Overall 2 Bands SPL level

0 km/h, Reverse ...... 44 48 10 km/h ...... 46 53 20 km/h ...... 51 58

Two other criteria were part of two-band approach to finalize the levels band approach. Because their suggested Alliance/Global’s suggested approach: and the frequency ranges may be needed approach did not specify different levels —That one of the two one-third octave and should be based on discussion for different frequency bands, there are bands should be in a frequency region among interested parties. They stated limitless possibilities for two-band below 1000 Hz and the other should that those discussions should take place signals that would meet the Alliance/ be at or above 1000 Hz; in the QRTV working group responsible Global method. However, the range of —That the two components of the signal for developing the GTR. possible signals just meeting the should not be in adjacent one-third In developing the four-band approach requirement can be categorized octave bands. that is included in today’s final rule, according to the following four signal A number of other NPRM NHTSA evaluated signals with different type scenarios: commenters, particularly vehicle numbers of bands including signals manufacturers, endorsed the two-band with two bands. The details of that (1) Scenario A: The level of the lower approach as suggested by Alliance/ evaluation are discussed above and frequency band of the two bands is set Global. shown in Figure 2. As discussed, at the suggested minimum, and the level In a follow-up letter submitted to the NHTSA’s approach in evaluating of the higher frequency band is set such docket in February 2014 (treated as a various signals was to set the band that the combination of the two bands late NPRM comment) a group of levels for each component at the meets the overall level (see Figure 4); commenters (Alliance, Global, the appropriate psychoacoustic thresholds (2) Scenario B: The level of the higher National Federation of the Blind, and according to Moore’s model which was frequency band of the two bands is set the American Council of the Blind) adjusted using the results of Volpe’s at the suggested minimum level and the expressed their agreement on human factors experiment. The adjusted level of the lower frequency band is set recommending a general approach of acoustic model was used to analyze the such that the combination meets the specifying two bands with an overall performance of signals having various overall level (similar to Figure 4); SPL level. In that comment letter, the numbers of frequency components from (3) Scenario C: The two bands both suggested parameters were somewhat one up to eight by predicting how are set at the suggested minimum level, less specific compared to the original readily each signal would be detected in and there is low level content over Alliance/Global suggestion or the the presence of the standardized 55 many frequencies that on its own may compliance option discussed in the dB(A) ambient. NPRM. The letter provided no As discussed previously, Figure 2 not be audible but that, when combined minimum band levels for the two bands demonstrates the robustness of single- with the two prominent bands, brings and left undecided the upper limit band and multiple-band alerts when the signal up to the specified overall frequency (either 3150 Hz or 5000 Hz) each band is set at the minimum level (see Figure 5); as well as the breakpoint between the threshold levels for detection based on (4) Scenario D: The two bands are low and the high frequency (either 1000 the acoustic model the agency used. We equal and their level is set such that the Hz or 1600 Hz). The joint commenters used this same robustness methodology combination of the two bands meets the indicated that further refinement of the to evaluate the Alliance/Global two- overall level (see Figure 6).

131 See docket NHTSA–2011–0148–0251, Alliance/Global comment, p. 5.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00045 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90460 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

60 ~ Ill ~-50 Q) • • • ~ • • • • • • • • ~ • • • iil 40 • ~ 0... ""0 5 30 • 0 • U) • ]l • -5, 20 • ! • ----·Test n o~====~~~==~ ~ r3> c§> .._cy:,.._~ .._ro~"'vcy:, ~~ n;,"~ ~~#~"'Y::ro\;);~~~~:~:>;~;);~,;>;;s:~~~.,~"' One-third octave band center frequency, Hz

Figure 4. Scenario A Two Components with the Lower Frequency at the Recommended Minimum Level and the Higher Frequency Increased to Meet the Overall SPL Requirement for 0 km/h

60 ~ Ill ~-50 Q) • • • iii -' • • • • • • • ~ • • • • • • iil 40 • "'~ • 0... -.:::> • 5 30 • 0 • U) • ]l • -5,20 • ·~ .._ro~r6>~-t>~<>:> .... ~ro~~~'%>~~ .... :' n~;t-' One-third octave band center frequency, Hz

Figure 5. Scenario C Two Components at Equal Levels Plus Additional Low Level Content Adjusted to Meet Overall SPL Requirement for 0 km/h

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00046 Fmt 4701 Sfmt 4725 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 ER14DE16.006 ER14DE16.039 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90461

The range of all possible signals shows the percentage of times that each Alliance/Global approach changes at meeting the criteria will fall somewhere signal category would be detected in the higher speeds. within these four signal types. For normalized sampled ambient From Figure 7 it can be seen that, at simplicity, we have considered these conditions. Note that three vehicle idle, two-band signals meeting the four types in our analysis. It is expected speeds plus stationary are indicated in Alliance/Global approach are robust that the robustness of other signals will Figure 7. In the suggested specifications regardless of which type of signal is be within the range observed for these provided in the Alliance/Global considered. However, as vehicle speed four types. comment, the minimum band values increases, robustness decreases. Figure 7 The results of our robustness analysis increased with increasing speed but indicates that the robustness of two-band signals meeting the only enough to partially account for the performance of certain two-band Alliance/Global suggested method are increase in sound level needed to signals, particularly those in the shown in Figure 7. Two-band signals are maintain adequate detection time over Scenario C category, declines plotted according to which of the four significantly to the point that, on the whole speed range. Consequently, signal categories (Scenarios A, B, C, or average, they would be detected only unlike in NHTSA’s acoustic D, above) they fall in, with averages about 35 percent of the time at 20 km/ indicated for each category. Again, this specifications, the performance of the h in the sampled ambient conditions.132

132 Figure 7 includes values plotted at 30km/h. derived by VOLPE because Alliance/Global’s suggested alert requirements went up to only 20km/ The data depicted at 30km/h is hypothetical data h.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00047 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 ER14DE16.007 90462 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

This analysis led us to conclude that signals to that of the NHTSA four-band and compensate by setting the second adopting the two-band Alliance/Global approach: component high enough to reach the approach as it was suggested in their (1) Instead of expressing the required required minimum band sum level. sound level in terms of overall SPL, we comments would allow some poor- In order to optimize the Alliance/ can use a band sum that accounts only performing alert signals to comply with Global’s suggested two-band approach the final rule. However, this analysis for the sound energy in the two required bands; this criterion would negate the using these modifications, the minimum also led us to conclude that some two- band sum levels at each speed were band signals perform as well by our possibility ability to augment the two bands with acoustic energy that may not iteratively determined. The results are measures as the signals meeting the be audible, i.e., that may not contribute shown in Table 15. We refer to this four-band requirements in this final to detectability and robustness. specification as an ‘‘optimized’’ two- rule, and that a two-band approach (2) We can adjust the required band approach because it excludes two- would be acceptable as long as it is minimum band sum to achieve band signals that have lower robustness specified in such a way as to exclude robustness equal to that of the four-band (those signals that would be detectable poor-performing two-band signals. Our specification. This provides a high in a lower number of ambients analysis of two-band signals highlights degree of flexibility in signal design. For according to our analysis) while two minor changes that we can make to example, a system designer can make preserving the levels suggested by the modify the Alliance suggestion in order the two components equal, or can set Alliance/Global to the greatest extent to increase robustness of two-band one component at the minimum level possible.

TABLE 15—OPTIMIZED LEVELS FOR TWO-BAND SIGNALS

A-weighted dB Minimum level in each of Band sum of 2 bands the 2 bands

0 km/h ...... 44 48 10 km/h ...... 46 55 20 km/h ...... 51 61 30 km/h ...... 56 66

Figure 8 shows the robustness in Figure 7. Scenario C that used optimized criteria will now be performance of two-band signals that broadband content to enhance the two detectable in upwards of 97 percent of meet this optimized approach. Note that bands is no longer viable under the the normalized sampled ambient there now are three sound scenarios (A, optimized approach. It can be seen that conditions and, on average, they reach B, and D) instead of the four discussed all two-band combinations meeting the

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00048 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 ER14DE16.008 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90463

at least the level of robustness achieved by the four-band approach.

Also note that the optimized The overall levels for both the each overall SPL value given for the specification includes levels for 30 km/ optimized two-band specification and optimized two-band approach, the level h because, as discussed in the crossover the four-band specification (‘‘S4 is within the ranges for the four-band speed section of today’s final rule Bands’’) are summarized in Table 16. specification. (Section III.D), the agency has decided For comparison, Table 16 also shows to include acoustic requirements for the levels suggested in the Alliance/ vehicle speeds up to 30 km/h. Global comment. It can be seen that for

TABLE 16—OVERALL LEVELS OF THREE APPROACHES

Minimum level, dB(A) * Stationary Reverse 10 km/h 20 km/h 30 km/h

S4 Bands** ...... 47–50 49–53 52–56 59–62 63–67 Alliance/Global ...... 48 48 53 58 NA Optimized 2-band ...... 48 *** 52 55 61 66 * Based on Partial Specific Loudness Threshold = 0.079 sones/ERB. ** Overall SPL depends on which four bands are selected. *** SPL for 10 km/h with 3 dB subtracted.

For the Reverse specifications, the dB from the 10 km/h level yields a value we are using the higher value. This will Alliance/Global comment set the band that is about the same as the band account for the fact that sound level for minimum levels and the overall level minimum the Alliance/Global suggested Reverse operation needs to be higher equal to the corresponding levels for the for Reverse, so the value we are than sound level in the Stationary stationary operating condition. In the adopting is the same as the one they condition, as explained in Section III.C optimized two-band specification, to be suggested. For the overall level, of this preamble. consistent with the four-band approach subtracting 3 dB from the 10 km/h level The modifications we have discussed and the method used in the NPRM, we yields a value for band sum that is to make two-band signals as robust as are setting the band minimum and somewhat higher than the overall SPL four-band signals will not make the two- overall SPL by subtracting 3 dB from the for Reverse suggested in Alliance/ band and four-band options the same in level required at 10 km/h. That method Global’s comment, as shown in Table all respects. For example, the four-band is the same one NHTSA employed in 16. To be consistent with the 4-band option is somewhat less restrictive the NPRM to set the levels for Reverse. requirements and the method used in because the minimum levels for the one- For the band minimum, subtracting 3 the NPRM to set Reverse requirements, third octave bands are lower than the

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00049 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 ER14DE16.009 90464 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

levels required with the two-band need to have available content at lower levels are essential to establish option. Also, the two-band approach is frequencies and that any potential minimum requirements for detection, more likely to result in a signal that has sound should have a fairly broad and that specifying overall sound an individual component that exceeds frequency spectrum. WMU suggested pressure level alone would not be an minimum detection thresholds in a that identifying two frequency bands acceptable approach for this final rule. particular band due to the need to meet that are most useful for detection, Stopping Distance the overall SPL requirement, which similar to Nissan’s approach, may be would make that component relatively appropriate. Many of the commenters agreed with prominent. We note that this does not The agency has reviewed the research the agency’s approach for using mean that environmental noise will be cited by ADB and conducted by WMU stopping distance for determining increased because, as shown in Table on the correlation between overall detectability requirements. Two of the 16, the band sum levels for the two- sound pressure level and detectability. commenters, however, ADB and WMU, band approach are lower at all speeds While this research does show that questioned the distance calculated and than the overall sound pressure levels overall sound level had a good used. ADB and WMU questioned that can be reached by alerts meeting correlation with detectability, it does whether the detection distances used the four-band approach. As discussed in not appear that it addressed whether are sufficient for pedestrians to detect, Section V.D of today’s final rule, our specifying levels in multiple octave recognize, judge distance and trajectory, environmental assessment indicates that bands influences the detectability decide to initiate a crossing, and initiate neither the two-band nor four-band outcome. The agency does not believe a crossing, particularly at busy approach would have significant that the cited studies adequately intersections. WMU explained that the environmental noise impact. support the proposition that overall detection distance formula used does In summary, we have decided that sound pressure level is a better metric not account for variability among including both compliance options in than one-third octave band sound pedestrians including those with this final rule allows manufacturers the pressure level. Furthermore, the WMU hearing loss. flexibility to choose the approach that comments about specifying low After considering the ADB and WMU best suits their design goals, while frequencies to assist with hearing loss, comments, we have decided to continue accomplishing the agency’s goals in the and about requiring a broad frequency to follow the approach used in the NPRM by providing a robustly spectrum, and also that specifying two NPRM where we derived stopping detectable signal for pedestrians without frequency bands may be appropriate, distance using a driver reaction time of significant environmental impact. The implies that they did not conclude that 1.5 seconds and a deceleration rate of detection requirements for compliance an overall specification by itself 5.4 m/s2. The agency’s main premise for of alert systems designed to meet the necessarily would be sufficient. the calculation of the time that should four-band and two-band specifications During the course of developing be allowed for detection of approaching are given in the regulatory text of FMVSS No. 141, the agency has vehicles was the total vehicle stopping today’s final rule. carefully considered overall sound distance needed to avoid pedestrian pressure levels and corresponding collisions. While the pedestrian’s Overall Sound Pressure Level individual one-third octave band sound reaction time is important, as is In the NPRM, the agency specified pressure levels. The agency agrees that providing as much time as possible for alert requirements at the one-third there can be a strong correlation pedestrians to make crossing decisions, octave band level and not at the overall between overall sound pressure level the critical factor is that the pedestrian sound pressure level. NHTSA’s position and detectability. However, we also should hear the alert of an approaching was that the overall sound level may be believe that regulating only the overall vehicle no later than the time and sufficient for ICEs, which intrinsically sound pressure level leaves open the distance the driver would need in order produce sound over a broad range of possibility of alert signals that may be to react and stop the vehicle before frequencies at all speeds and have undetectable in many common colliding with the pedestrian. acoustic characteristics such as situations. Agency research indicates Furthermore, the alert requirements modulation that enhance detectability, that alert sounds with the same overall specified in the final rule include a but not sufficient for inherently quiet sound pressure level often do not small safety margin that will extend the vehicles operating solely on electric provide the same degree of detectability timing and distance for both the driver motors at low speeds. The agency or robustness. This topic is discussed in and the pedestrian. As discussed continues to believe that one-third sections that follow in this preamble previously, the minimum one-third octave band requirements assure that a where we identify how the agency octave band levels derived for vehicle’s total sound is detectable by a derived the two compliance options detectability were increased by 0.5 dB broad range of pedestrians over many specified in this final rule. Through our and rounded up to the closest whole ambient conditions. research, the agency has determined decibel. Also, because our minimum ADB commented that, ‘‘octave bands that for an alert signal to be as ‘‘robust’’ requirements are based on the levels are not as great at predicting detection as possible, i.e. for a signal to be heard needed to detect a signal having content as overall sound levels’’ based on by the most diverse range of pedestrians in a single one-third octave band, our research conducted by WMU. WMU across the widest range of ambient requirement that signals must include stated that its research has shown that conditions, specific combinations of multiple one-third octave bands individual octave bands are not as one-third octave bands in different provides an additional margin of safety. useful in determining detection as is the frequencies must be included in the We believe that requiring EVs and HVs overall sound level. WMU stated that requirements of the final rule. The to produce sounds with content in while some regulatory specification in requirements for one-third octave bands multiple one-third octave bands will octave band make-up of alert sounds at various frequencies contribute to the provide an additional safety margin of might be useful, there is limited overall sound pressure level of the time and distance due to the increased justification for such a restrictive sound emitted by the vehicle. overall sound pressure level resulting requirement. WMU also stated that a Conversely, the agency maintains that from the combination of one-third pedestrian with hearing loss would minimum one-third octave band sound octave bands. In addition, the

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00050 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90465

specifications in this final rule are conditions present during the sound which contains frequencies as minimum levels for compliance. recordings,134 135 which allowed low- low as 160 Hz. Vehicle manufacturers are likely to frequency combustion related tones and OICA stated that a tone that is pitched exceed the minimums by some amount wide range broadband content 136 to be would simulate the sound of a machine in order to provide themselves with a audible. and this in combination with the tire/ margin of compliance. We believe these The agency sought comments on the road noise would be enough to factors address concerns that the following topics related to the proposed recognize the sound as coming from a reaction time the agency used was recognition requirements: vehicle. They also stated that broadband insufficient. • Suggestions for the minimum sound band should not be required. level of low frequency content that SAE indicated that the metric used to F. Acoustic Parameters for Recognition should be included in the agency’s define ‘tone’ (ANSI S1.13—1995), in the of Motor Vehicles recognition requirements; proposed regulatory text, is not robust to • In the NPRM, we stated that Information as to whether speakers all possible sound designs and would recognition includes two aspects: that manufacturers may wish to use to explicitly exclude sound characteristics Recognition that the sound is emanating meet the requirements of the proposal identified as contributing to detection from a motor vehicle that may pose a are capable of producing any and recognition in the preamble. safety risk to the pedestrian, and measurable content in the 160 Hz one- Ford stated that it conducted a study third octave band; and to examine recognition of a given sound recognition of the vehicle’s operating • mode (acceleration, deceleration, Information about the cost of a as the sound of a motor vehicle. The constant speed, reverse or stationary but speaker system that is able to reproduce study consisted of a human factors test activated) so that the pedestrian can some measurable content at the 160 Hz in which audio recordings of vehicle take appropriate measures to avoid a one-third octave band versus the cost of sounds were presented to participants collision with the vehicle. The acoustic a speaker system that is only capable of using headphones. Participants were specification in the NPRM contained producing sound above 315 Hz. asked to assess how recognizable the The Agency received comments from acoustic characteristics similar to the sounds were in the presence of Alliance/Global; SAE; OICA; Honda; sounds that pedestrians associate with background noise. The study included Nissan; Porsche; Mercedes; Denso; current ICE vehicles. 24 Ford employees and 4 blind National Federation for the Blind; Based on our initial assessment of individuals. Sounds tested included an Western Michigan University; simulated sounds and engineering ICE vehicle, a vehicle without an alert Accessible Design for the Blind; The judgment, the agency determined in the sound, and three alert sounds. Two tests Seeing Eye, Inc. were completed; recognition of a NPRM that the sound emitted by the According to Alliance/Global, bands vehicle to meet the detection stationary sound and recognition of a 10 below 500 Hz should not be required. km/h pass-by. Additional tests were requirements must contain at least one They stated that these bands are not tone. A component is defined as a tone conducted to examine recognition of the necessary for recognition and will add sound as an object to avoid. Ford if the total sound level in a critical band significant cost to the alert sound concluded that adding motion to the centered about the tone is 6 dB greater system. Alliance/Global also stated that sound (pass-by vs. stationary) increased than the noise level in the band.133 In isolating and measuring low frequency recognition as either a motor vehicle or the NPRM, we proposed requiring the content under outdoor test conditions an object to avoid. They also explained sound emitted by the vehicle to have at would be impracticable. Alliance/Global that it is not necessary to meet all least one tone at a frequency no higher stated that prescribing an objective proposed minimum levels in the 315 than 400 Hz. The agency also proposed definition to recognizability using one- Hz, 400 Hz, and 500 Hz one-third octave that the sound emitted by the vehicle third octave bands is not possible bands for vehicles or alert sounds to be must have content in each one-third because there are many ways to provide recognized as motor vehicles. octave band from 160 Hz to 5000 Hz. sounds that have similar acoustic Honda indicated that the generation Simulated sounds in the initial characteristics. Finally, they do not of low frequency sound is technically assessment were developed for the recommend one-third octave band challenging, creates extra cost, and adds stationary but activated, constant speed requirements in the 160 Hz band weight to the vehicle. Honda explained pass-by, and accelerating pass-by because existing speakers that are that the sound entering into the conditions. Pass-by sounds included practical for alert systems cannot emit passenger compartment could be Doppler shifts (changes in frequency by significant, which could cause a source moving relative to an observer) 134 Garay-Vega, L; Hastings, A.; Pollard, J.K.; annoyance. Honda suggested that this and simulated acceleration (a pitch or Zuschlag, M. & Stearns, M. (2010, April). Quieter would require testing and an iterative frequency shifting tied to a change in Cars and the Safety of Blind. Pedestrians: Phase 1. design process to minimize negative vehicle speed.) The sound pressure DOT HS 811 304. Washington, DC: National Highway Traffic Safety Administration. effects. level changed as a function of speed and 135 Hastings, A., Pollard, J. K., Garay-Vega, L., Nissan stated that low frequency as a function of position relative to the Stearns, M. D., & Guthy, C. (October, 2011). Quieter content alone will not ensure that a microphone receiver during the pass-by Cars and the Safety of Blind Pedestrians, Phase 2: sound is recognized as a motor vehicle. Development of Potential Specifications for Vehicle simulations. During the original Nissan suggested that requiring development of criteria for recognition, Countermeasure Sounds. DOT HS 811 496. Washington, DC: National Highway Traffic Safety frequency content in this region means we stated that an alert signal should Administration. that either broadband or narrowband sound like an ICE in order to be 136 Broadband content is content over a wide content (e.g. tones) could be used, recognizable. In order to identify frequency range that could be spectrally continuous which would sound quite different than qualities of the ICE vehicle, ICE sounds or periodic. Periodic content can be generated by engine combustion related harmonics or by periodic an ICE. were evaluated in the quiet ambient tire/pavement interactions, such as caused by Mercedes indicated that the proposed transversely tined pavement. Continuous content specification is restricting 133 The agency explained that a component is can be generated by turbulence at the engine intake considered to be a tone if the Tone-to-Noise ratio and exhaust ports, by non-periodically tined fan manufacturers flexibility to produce according to ANSI S1.13–199573 is greater than or blades as well as by aerodynamic noise and random alert sounds for EVs and HVs that are equal to 6 dB. tire/pavement interactions. effective yet pleasant to consumers and

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00051 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90466 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

expressed concerns about potential Comments from the Accessible Design pedestrians would need to use other impacts to market penetration. for the Blind (ADB) are consistent with cues to recognize a vehicle (ICE or Mercedes explained that low one-third WMU. ADB indicated that tones are otherwise), such as the location of the octave frequency bands down to 315 Hz masked by the ambient and that most sound source (e.g. on the street at a stop and broadband content down to 160 Hz people find tones to be annoying. ADB light), the frequency and level changes are difficult to isolate inside the vehicle stated that added sound should be the caused by sound source motion (e.g. on cabin and this may result in adding same for all EVs and HVs. ADB the street approaching or passing the vehicle weight due to added insulation. explained that this would help with pedestrian), etc. Mercedes also mentioned that a speaker recognition and prompt interpretation of A recent study by NHTSA examined would need to increase in size in order the sound as the sound of a vehicle. In several alert signals in the presence of to accommodate the proposed lower response to the request for comments a 55 dB(A) ambient for a vehicle frequency requirements. about the minimum levels of low traveling at 10 km/h.139 The signals Porsche mentioned that pitch shifting frequency content that should be included simulations based on recorded is the most important factor to included for recognition, ADB stated vehicles, tones, and noise components characterize motor vehicles. Porsche that they are not aware of any research over a frequency range from 315 to 5000 suggested that the number of that supports the notion that adding low Hz. Some signals had only a single frequencies and the frequency range be frequency content makes sounds more component, e.g. a tone or a noise at 315, kept flexible. Porsche also indicated that recognizable. 630 or 2500 Hz, or multiple broadband sound should not be The Seeing Eye, Inc., stated that, for components, e.g. low frequencies (315 required. Porsche stated that all sounds recognition purposes, it is important to 500 Hz), high frequencies (2000 to emitted by a vehicle are based on tones that all vehicles regardless of 5000 Hz), or components matching the while broadband sound comes from tire manufacturer, emit the same NPRM frequencies. Participants were noise. Porsche also explained that standardized sound. asked to indicate when they heard a sound that would influence their broadband sounds would require Agency Response to Comments different devices and cannot be decision to cross a street. The study generated by the prototype control After reviewing the comments and provides a practical indication of a modules currently used by Porsche. conducting additional research, we have pedestrians ability to recognize sounds Denso requested clarification of the decided to remove the requirements in emitted by HVs and EVs as motor definition of the terms ‘‘tone’’ and paragraph S5.2 of the NPRM requiring vehicle sounds since recognition is required in order to respond to the ‘‘critical band.’’ Denso also mentioned EVs and HVs to produce sound that detected signal in the form of making a that the agency did not identify sound includes broadband content and low decision regarding whether it is safe to pressure levels for the broadband frequency tones. We believe these cross a street. requirement in the NPRM. Denso stated acoustic characteristics are not necessary for pedestrians to recognize All alert signals tested (with the that the broadband requirement may not exception of one signal that had levels be as effective for recognition and artificial sounds produced by EVs and HVs as coming from a motor vehicle in below NPRM values) were detected and localizability because the sound emitted recognized on average by the minimum by the vehicle speaker system may be operation. During the agency’s initial work to safe detection time of 2.0 seconds or masked by ambient sound if no sound greater. These results are consistent level for the broadband content is develop criteria for recognition, the agency assumed that an alert signal with comments by the Alliance/Global specified. and with the study submitted by Ford. NFB stated that recognition should sound like an ICE in order to be recognizable. In order to identify Based on these results, it appears that requirements were included in the vehicle recognition cued by an alert PSEA to prevent excessive qualities of the ICE vehicle, ICE sounds were evaluated in the quiet ambient signal in the presence of a ambient at 55 customization. They stated that the dB(A), which is the target ambient for inclusion of pitch shifting will conditions present during the 137 detection, does not require that the alert potentially be sufficient to insure recordings which allowed low- frequency combustion related tones to signal contain low frequency tones. recognition. Because low frequency tones are not WMU indicated that the inclusion of be audible. These low frequency tones make up part of the sound of a typical necessary for pedestrians to recognize tones is unlikely to enhance recognition sounds as vehicles sounds, could also because tones are readily masked by ICE vehicle at low speeds in quiet ambients. However, these low frequency add cost to the system, and may be sounds in the environment, especially annoying when not masked by the by sound from other vehicles. WMU tones are masked in many ambient conditions, and in particular the 55 ambient, the agency is not including a also indicated that many blind requirement for low frequency tones in pedestrians would not detect sound dB(A) ambient used for determining the minimum sound requirements the final rule. energy above 2000 Hz, especially those Similarly, the agency study showed described in the NPRM.138 In such cases with hearing loss; therefore, this is not that participants detected and a reliable way to enhance recognition. 137 Garay-Vega, L; Hastings, A.; Pollard, J.K.; recognized alert signals with a wide WMU indicated that rhythmic, cyclic Zuschlag, M. & Stearns, M. (2010, April). Quieter aspect of a sound would enhance Cars and the Safety of Blind. Pedestrians: Phase 1. threshold levels described in the NPRM, it can be recognition. In terms of speaker DOT HS 811 304. Washington, DC: National seen that for most vehicles in Table 29 many of the capabilities, they suggested that the cost Highway Traffic Safety Administration.; see also measured vehicle one-third octave band levels are Hastings, A. et al. (2011). Quieter Cars and the below the computed thresholds for the 55 dB(A) of using speakers capable of producing Safety of Blind Pedestrians, Phase 2: Development ambient used in the NPRM. Thus these components sound energy in the 160 Hz range is not of Potential Specifications for Vehicle would not be reliably detectable in such an balanced by additional benefits. They Countermeasure Sounds. DOT HS 811 496. ambient. explained that their studies have not Washington, DC: National Highway Traffic Safety 139 Hastings A.; and McInnis, Catherine. Administration. ‘‘Detectability of Alert Signals for Hybrid and found this low range to be useful for 138 OICA measured stationary but activated levels Electric Vehicles: Acoustic Modeling and Human detection and noted that tones can be are presented in Table 29 of the Phase III report. Subjects Experiment’’ (2015) Washington, DC: DOT/ annoying. Comparing these data with the associated minimum NHTSA.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00052 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90467

range of sound characteristics including frequency from the compliance tests at A tone is an acoustic component with signals that do not include broadband stationary, 10 km/h (6.2 mph), 20 km/ well-defined features that make it content over the entire range from 160 h (12.4 mph), and 30 km/h (18.6 mph). relatively easy to recognize compared to Hz to 5000 Hz. For example, several The NPRM provided a definition for the noise. The pitch, or frequency, of an signals in the study consisted of only a fundamental frequency but did not alert sound could be verified by tracking single pure tone or a single one-third specify how the fundamental this tone as it increases in frequency for octave band of noise and were detected frequencies at each vehicle speed each pass-by test as the vehicle and recognized at a safe distance should be compared. increases speed. In the proposal, we provided the component met minimum As mentioned, the agency did not said it would be difficult to verify a levels as determined by the detection include a separate acoustic sound’s increase in frequency if the model. Based on these results, it appears measurement procedure for frequency sound does not have any strong tones. that vehicle recognition cued by an alert shifting in the NPRM, instead relying on We mentioned our concerns about signal in the presence of a 55 dB(A) other requirements specified and the identifying the tone of a sound and ambient does not require broadband increase in overall sound level as the tracking this tone as the vehicle content in all one-third octave bands vehicle increases speed (or the decrease increases speed. The NPRM mentioned from 160 Hz to 5000 Hz. Given the in sound level as the vehicle that we planned to conduct further potential costs associated with meeting decelerates) to provide enough research on this issue. We explained the low frequency requirements of such information so that pedestrians will be that if it was not possible to identify a broadband content and the fact that able to determine when EVs and HVs tone to track in order to verify the signals meeting the detection criteria are are accelerating and decelerating. One increase in a sound’s frequency, we may safely detectable, the agency is not reason why a separate acoustic have to use a different method to verify including a broadband content measurement procedure was not the increase. The agency sought requirement in the final rule included was due to the concerns about comments on this issue. specification. the feasibility of testing. The agency The agency received comments on Overall, the agency believes that stated that it would be difficult for even frequency shifting from SAE, Alliance/ pedestrians would use other cues to an experienced test driver to repeatedly Global, OICA, and Porsche. The agency recognize a vehicle (ICE or otherwise), achieve and maintain a specific rate of also separately received a joint comment such as the location of the sound source acceleration or deceleration on a test submitted by the Alliance, Global, the (e.g. on the street at a stop light), and the track if such a test was required. Given American Council of the Blind (ACB), frequency and level changes caused by the difficulty of ensuring a repeatable and the National Federation of the Blind sound source motion (e.g. on the street acoustic test for acceleration and the (NFB). approaching or passing the pedestrian), fact that information about changes in Several commenters stated that the etc. (See Section III.G on ‘Frequency vehicle speed could be provided by NPRM did not include a test procedure (Pitch) Shifting and Volume Change’). varying sound pressure levels, NHTSA to measure compliance with the determined that the test procedure did proposed frequency shifting G. Frequency (Pitch) Shifting and not need to include a dynamic test for requirements. These commenters Volume Change acceleration or deceleration. recommended that the agency use the The NPRM contained a requirement The NPRM explained that frequency shift procedures specified in for frequency shifting which gives the manufacturers and their representatives, SAE J2889–1 to measure compliance pedestrian information about the in meetings with NHTSA staff, with the frequency shifting acceleration or deceleration of an expressed concerns that it is difficult to requirements and that the agency allow approaching vehicle. The PSEA measure the change in frequency of a indoor testing or component level required NHTSA to include sounds to sound produced by a vehicle by testing to measure frequency shifting. alert pedestrians to acceleration and measuring a complete vehicle during a SAE commented that use of indoor deceleration. As discussed in the pass-by test. Manufacturers requested facilities for the measurement of the NPRM, this information is important to that the agency measure frequency frequency shift is necessary to obtain the pedestrian in making a decision shifting using a component-level test, accurate results. SAE said that about whether or not to cross in front of meaning that the alert system hardware provisions for indoor measurement a vehicle. The driver of an accelerating is removed from the vehicle and tested either at a component level or a vehicle probably does not intend to stop as a separate unit. simulated full-vehicle level are included and, according to the NPRM, ‘‘the sound In the NPRM, we said that we were in SAEJ2889–1 (May 2012). SAE also of accelerating vehicles in the parallel hesitant to include a component-level mentioned that in a December 2012 street indicates, for example, that the test because we wanted the standard to meeting with NHTSA, an alternative perpendicular traffic does not have the be technology neutral and because we method of analysis was under right of way and thus a crossing do not wish to limit technological investigation to eliminate the need for opportunity is available’’. A innovation. As further explained, the prior knowledge of the signal. decelerating vehicle on a path parallel agency was aware that manufacturers Alliance/Global mentioned that tonal to the pedestrian may be slowing to might use different technologies to tracking for frequency shifting becomes make a turn into the pedestrian’s path comply with the standard, so defining quite difficult at higher speeds (30 km/ if she or he were to cross the street. the hardware components subject to the h) due the tire noise masking, The proposal required that the component-level test could prove particularly when testing outdoors. fundamental frequency of the sound difficult. The agency sought comment Alliance/Global stated they prefer an emitted by the vehicle increase with on including a component-level test to indoor component level test because speed by at least one percent per km/h measure frequency shifting in the test they think that is the best way to ensure between 0 and 30 km/h (18.6 mph). The procedure. that the correct tones are being tracked NPRM did not include a test procedure In the NPRM, the agency said that the and that noise from tires (at higher associated with this requirement but proposed method for measuring speeds), accessory equipment, or other stated that frequency shifting could be frequency shifting depends on the sounds not intended for pedestrian verified by comparing the fundamental presence of a strong tone in the sound. safety, are not incorrectly counted

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00053 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90468 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

toward the sound measurement. with requirements for frequency shift at shifting. The agency believes that it Alliance/Global indicated that they are this time has been deemed unfeasible. would be difficult to reliably and not aware of a procedure that can As proposed in the NPRM and repeatably verify compliance because identify these tones during whole- finalized here, the sound pressure level the frequencies identified for frequency vehicle testing. in each one-third octave band changes shifting by different technicians are OICA suggested that NHTSA change as speed increases, leading to an unlikely to always be exactly the same. the definition of ‘‘fundamental increasing overall sound pressure level Since issuing the NPRM, the agency frequency’’ in S4 to read, ‘‘[Frequency] that corresponds to the behavior of an has conducted additional research in an shift frequency means, for purposes of ICE vehicle. Thus pedestrians will be attempt to develop a cohesive this regulation, any frequency or able to tell if an EV or HV is accelerating methodology for analyzing and verifying frequencies used to comply with or decelerating based on the increase or frequency shifting. NHTSA considers S5.1.6.’’ decrease in sound level emitted from frequency shifting measurement to OICA suggested requiring that the the vehicle, just as they would be able consist of three main steps: (1) frequency of the sound shift frequency to in the case of an ICE vehicle. In this Measurement of the signal to be used in within each individual gear ratio rather final rule, the agency has chosen to use the analysis and its conversion to the than over the entire range of speeds the increase and decrease in sound corresponding frequency domain; (2) between 0 and 30 km/h. OICA stated produced by the vehicle at different identification of the alert sound tonal that this will allow for the simulation of speeds as an alternative to frequency components that meet the definition of an ICE vehicle using different gear ratios shifting. tone and that are expected to shift at We have decided to identify this within the tested speed range. each of the measured operating alternative method by the term ‘‘relative Furthermore, OICA indicated that there conditions (stationary, 10 km/h, 20 km/ volume change.’’ Basically, the method might be various ways to determine the h, and 30 km/h); and (3) calculation of of ‘‘relative volume change’’ involves frequency tone and rate and suggested the actual magnitude of frequency summing and comparing the normalized that NHTSA leave the way to measure shifting that has occurred from the measured one-third octave band levels it to the individual manufacturer. OICA identified tonal components. Of these for each of the operating speeds for each indicated that there is no known steps, step one, recording the test vehicle. For each operating speed, measurements and converting them to method to identify the proper tone in all the normalized sum of the measured the frequency domain, is relatively situations without specifying the tone in one-third octave bands should increase routine as this is a standard signal advance. OICA stated that information by a specified minimum amount at each processing technique. Also, in step about the signal under evaluation will successive speed interval. Further three, once the proper tones and base be necessary. details about the ‘‘relative volume frequencies of the vehicle alert have Porsche made reference to the signal change’’ method and why the agency been identified and have been processing requirements in SAE J2889– believes the original frequency shifting determined to be a continuous result of 1 (7.2.3) and stated ‘‘The fundamental requirement is not feasible are discussed frequency shifting, it is relatively easy to frequency is dependent on the setup of below. mathematically determine the amount the analysis system and is typically less The agency acknowledges comments of frequency shifting that has occurred. than two Hertz.’’ Porsche also suggested regarding the lack of a test procedure to From both a process basis and a that NHTSA change the definition of measure frequency shifting in the calculation basis, steps one and three fundamental frequency in S4 to read NPRM. Many of the commenters appear consistent with the methodology . . . ‘‘S4 Fundamental frequency means, requested that, in lieu of a test specified in SAE J2889–1. for purposes of this regulation, any procedure being included in the rule, Unfortunately, in step two above, prominent frequency of a valid the agency adopt the frequency shifting identification and validation of tonal measurement taken in S7.’’ procedure set forth in SAE J2889–1 components is exceptionally difficult. In the joint comment submitted by Section S7.2. In essence, this procedure The procedure detailed in Section S7.2 Alliance/Global/NFB/ACB, those calls for identification of a frequency of SAE J2889–1 specifically requires commenters agreed that at least one that has changed as a function of vehicle that the person conducting the test frequency emitted by the vehicle must speed, which can be measured and can know in advance what frequencies are vary with speed by at least an average be tracked during the operating shifting to avoid having to subjectively of one percent per mph over the range conditions specified. However, the SAE identify and verify the critical tones from 5 mph to the crossover speed. procedure, as stated in appendix B–5 of produced by the vehicle alert system. To They indicated that this frequency may the SAE standard, requires prior identify and validate tonal components, also contribute to meeting the spectral knowledge of the frequencies to be the test operator first must know and overall sound pressure level tracked (‘‘The persons conducting the precisely how a tone is defined. The requirements. test know what frequencies should be NPRM defined a component as a tone if produced by the device or vehicle under Agency Response to Comments the total sound level in a critical band measurement’’). NHTSA believes that centered about the main tonal frequency After reviewing the comments and the need for prior knowledge of the is 6 dB greater than the noise level in conducting additional research on the frequencies precludes a readily the band; however, the terms ‘‘noise topic of frequency shifting, we have verifiable and practicable test level’’ and ‘‘critical band’’ were left decided not to include a requirement procedure. Also, the procedure set forth undefined, and this omission was cited that a vehicle’s emitted sound must in J2889–1, Section 7.2, requires an by the commenters. As such, the change in frequency as the vehicle acoustics expert to determine both the language in the NPRM was insufficient changes speed. Although this starting frequency (and/or tone) as well to resolve a tone in a way that would characteristic is still considered useful as the shifted frequencies as speed allow frequency shifting determinations. and we encourage its use on hybrid and increases, to verify compliance. The During further research into defining electric vehicles for enhanced agency believes that this contributes to a tone, NHTSA found that there are four detectability and recognizability, a test a lack of objectivity in the SAE test main ways of identifying and verifying procedure to determine compliance procedure for measuring frequency tones: By using predetermined

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00054 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90469

information from manufacturers; subjective and unsuitable for the Relative Volume Change visually, by plotting various sound data purposes of safety standard compliance. Because it is not feasible to include and determining an overall pattern; by The other methods for determining requirements for frequency shifting in utilizing a small amount of tones both require technical data from the final rule for the reasons discussed predetermined information (such as the the manufacturer. Either the above, the agency has decided to base frequencies measured while the manufacturer would have to supply all include in the final rule a requirement vehicle is in a stationary mode) and of the data on frequency shifting, for vehicle-emitted sound level or assuming a rate of frequency shifting to specifying all tones which will be used ‘‘volume’’ rather than in frequency to determine values for 10 km/h, 20 km/h, to calculate compliance, or the increase as the vehicle increases speed. and 30 km/h; or lastly by utilizing a manufacturer would have to provide a The agency has decided to include this computer program to analyze sound smaller amount of information, such as volume change requirement as a means data and search for tonal characteristics. the tonal components at stationary, and for pedestrians to utilize the sounds Identification and verification of tones, the agency then would have to assume emitted by a vehicle to determine if a regardless of method, is further a rate of frequency shifting as a function vehicle is accelerating or decelerating. complicated by the fact that vehicles do of speed and would estimate where the The agency understands that the not generate a simple sound pattern and new tonal components should lie. concept of ‘‘relative volume change’’ is in general have a mixture of many tones, Unfortunately, this process also is not not a direct replacement for frequency coupled with broadband noise as well, objective, as the agency would be shifting, but we believe it is a reasonable which is consistent with what relying on information from the alternative. While frequency shifting commenters said. There are also pre- manufacturers and on acoustics experts would be a more certain method for existing sound sources that have tonal to validate that information. determining vehicle acceleration and and inherent frequency shifting deceleration, volume change will NHTSA also investigated the use of qualities (for example, tires can produce provide useful audible information to automated procedures utilizing ANSI a sound that has specific tonal qualities pedestrians about the operating state of S1.13: 2005, ISO 3745, and SAE J2889– that will shift to a higher frequency that nearby vehicles. We believe that the 1. However, NHTSA has been unable to is proportional to the increasing speed volume change specifications will produce a fully workable automated of the wheel). These sound sources can partially compensate for the absence of work together to make searching for method. More research would be pitch shifting requirements. vehicle alert system tones very difficult needed, but it is uncertain if the agency To better understand the concept, as and subjective. could ultimately develop repeatable, a vehicle approaches a pedestrian at a NHTSA investigated using visual reliable, and objective procedures that constant speed, the pedestrian would methods to identify tones: plotting the do not require verification by an expert. hear the vehicle alert sound increase in frequency levels versus sound levels as In light of the above discussion volume, identifying that the vehicle is a function of both frequency and time as highlighting the impracticality of approaching but maybe not accelerating the vehicle is accelerated at a constant identifying and verifying tones without or decelerating. However, if the vehicle rate (a so-called ‘‘run-up’’ graph, prior knowledge of the expected is approaching a pedestrian and presented as a spectrogram plot) where frequency shift, NHTSA agrees with the accelerating (or decelerating), the alert prominent frequency components can note 2 of Section S7.2.5.1.1 of SAE sound will increase (or decrease) in be tracked as they change due to J2889 Rev DEC2014, ‘‘. . . there is no volume more rapidly as the vehicle frequency shifting; or by graphing sound known identification specification that approaches while transitioning between levels as a function of frequency can clearly identify frequencies which 0 km/h and 10 km/h, between 10 km/ (referred to as the discrete method) for shift with vehicle operating conditions, h and 20 km/h, and between 20 km/h each speed condition (stationary, 10 primarily vehicle speed, when the and 30 km/h. A rapid ramp up in km/h, 20 km/h, and 30 km/h) and frequency content of the desired signal volume as the vehicle approaches will identifying prominent frequency and any background noise is unknown.’’ be indicative of a vehicle accelerating, components which seem to be a Since no practicable test methodology and a rapid reduction in volume as the function of frequency shifting. An consistent with the requirements of an vehicle approaches will be indicative of example of these types of visual plots FMVSS has been developed to date to a vehicle decelerating. can be found in Figure B–1 of SAE objectively determine frequency The minimum detection thresholds J2889. Because the discrete method shifting, the agency is not including a which are contained in this final rule looks at individual test cases, there is no requirement for frequency shifting in increase with speed. Consequently, guarantee that the frequencies identified the final rule. vehicles that meet the minimum will be a result of continuous frequency requirements, without exceeding them, Nevertheless, the agency encourages shifting, and that the frequencies are not will have an innate volume increase manufacturers to include frequency instead merely tonal artifacts present in commensurate with the increase in the individual test case. It would be left shifting in their development of alert speed. The minimum specifications up to the judgment of an acoustics sounds as this shifting does provide incorporate a volume change of expert to make this determination. Also, aural information to pedestrians about approximately 6 dB between stationary utilizing the run-up method would whether they are at risk or not and about and 10 km/h, approximately 6 dB require the judgment of an acoustics the distance, speed, and acceleration of between 10 km/h and 20 km/h, and engineer to determine the characteristics approaching vehicles. These are useful approximately 5 dB between 20 km/h of a potential tone, identifying center cues for pedestrian navigation. and 30 km/h. However, manufacturers frequencies, and determining if In the future, should a practicable, could design alert signals that have only irregularities are present. Although it objective method to quantify frequency a single sound level, such as one that may be more objective than discrete shifting of vehicle alert sounds be meets the highest sound level visualization, this method can yield developed, NHTSA may reconsider its requirements (those required at 30 km/ multiple interpretations of the same decision to exclude a frequency shifting h) across all speeds (thus exceeding the data, which makes it inherently requirement from the safety standard. minimum levels at stationary, 10 km/h

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00055 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90470 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

and 20 km/h). In this case, the alert vehicle could comply with the level of individual one-third octave would have no built-in volume change requirements of the NPRM by meeting bands. Also, the overall sound pressure with increasing or decreasing speed, the minimum acoustic requirements for level is influenced by bands that are and the potential pedestrian cue to the highest test speed for all test speeds outside of the range of one-third octaves increasing or decreasing vehicle speed without any variation in the sound covered by NHTSA’s specifications (i.e., would not exist. The ‘‘relative volume produced by the vehicle. In other words, those greater than 5000 Hz and less than change’’ requirement specified in this a vehicle alert system could be designed 315 Hz). Therefore, in order to evaluate final rule will ensure a minimum sound such that it would emit the loudest changes in perceived volume level, we level increase and decrease as a vehicle required sound level in all test will consider only the one-third octave reaches each successive higher or lower conditions from stationary up to 30 km/ bands that account for sound energy speed operating condition. h. Under this scenario, a pedestrian contained in the range from 315 Hz to In discussing the minimum acoustic would have limited ability to detect 5000 Hz. Normalized one-third octave requirements for the eight one-third changes in vehicle speed without pitch band values are derived by subtracting octave bands in the NPRM, NHTSA said shifting because the sound produced by the minimum one-third octave values the minimum requirements in each one- the vehicle would not change as the specified for the stationary operating third octave band increased as the vehicle changed speed. To eliminate condition from each of the one-third vehicle increased in speed to give this possibility, NHTSA has included octave band alert measurements. This pedestrians more time to detect faster the volume change requirements in the normalization process allows moving vehicles and to allow the final rule to ensure that the alert sound measurements of different one-third pedestrian to determine whether the varies produced as vehicle changes octave bands to be compared by vehicle was accelerating or decelerating. speed. accounting for the differences in the While the minimum acoustic Since an alert signal’s acoustic minimum levels specified for each requirements in the NPRM increased for components can change from one band. The logarithmic sum of the each test speed, the NPRM did not operating condition to the next, changes thirteen normalized one-third octave include maximum sound requirements in the overall SPL level will not band levels is then determined (i.e., the for each test speed. This meant that a necessarily correspond to changes in the ‘‘band sum’’).

Finally, the relative volume change is octave bands outside of the range of third octave bands of the first four valid, calculated as the difference in these interest (315 Hz to 5000 Hz). Another ambient-corrected runs, from the louder band sum values between consecutive key characteristic of this approach is side of the vehicle (left or right), for operating speed conditions. that frequency is not tracked, which each operating condition (Stationary, 10 Evaluating the increase in band sum provides design flexibility because km/h, 20 km/h, and 30 km/h). By values from one speed to the next then different one-third octave bands can be comparing the calculated band sum at a provides a metric for ‘‘relative volume prominent at different speeds. given operating speed with the band change.’’ This approach allows for the The relative volume change procedure sum value for the next lower speed tracking of volume as a function of will utilize the same vehicle condition, a relative volume change can speed, as the volume is characterized by measurement data collected for the be computed. the sound pressure levels above the determination of compliance with the minimum levels required at the baseline minimum detection standards. That is, An example calculation is provided in stationary operating condition. It also the volume change determination uses Figure 9. allows for the rejection of one-third the average values for the thirteen one-

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00056 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 ER14DE16.010 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90471

Figure 9 illustrates the four-step (stationary, 10 km/h (11+/¥ 1km/h), 20 stationary operating condition from procedure used to calculate the relative km/h (21+/¥ 1km/h), and 30 km/h each of the one-third octave band volume change for sample data for the (31+/¥ 1km/h)). averages calculated for each operating 10 km/h to 20 km/h conditions as Step 2: Calculate the normalized scenario (stationary, 10 km/h (11+/¥ follows: values for each of the 13 one-third 1km/h), 20 km/h (21+/¥ 1km/h), and Step 1: Calculate the average octave bands for each of the operating 30 km/h (31+/¥ 1km/h)). measured one-third octave band level scenarios, relative to the minimum SPL Step 3: Calculate the BAND SUM for for each of the 13 one-third octave requirements specified for the stationary each critical operating scenario bands (315 Hz to 5000 Hz) using the operating scenario. The normalized (stationary, 10 km/h (11+/¥ 1km/h), 20 four valid test runs identified for each values are calculated by subtracting the km/h (21+/¥ 1km/h), and 30 km/h of the test operating scenarios minimum SPL values specified for the (31+/¥ 1km/h)) as follows:

Where: subtracting the BAND SUM of the lower standards at each operating condition, i represents each of the 13 one-third octave speed test case from the BAND SUM of its relative volume change would be bands. the next higher speed test case. approximately 6 dB between stationary Normalized Band Leveli is the calculated The performance specifications for and 10 km/h, approximately 6 dB normalized value for each of the 13 one- the relative volume change requirement between 10 km/h and 20 km/h, and third octave bands. were derived based upon the minimum approximately 5 dB between 20 km/h Step 4: Calculate the relative volume detection standards for each operating and 30 km/h. It is the agency’s desire to change between each operating scenario condition. The minimum detection ensure that vehicles equipped with (stationary to 10 km/h; 10 km/h to 20 standards increase with speed such that, compliant alert sounds are only as loud km/h; 20 km/h to 30 km/h) by if a vehicle just meets the minimum as they need to be for detection by

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00057 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 ER14DE16.011 ER14DE16.012 90472 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

pedestrians, and not excessively louder. minimum requirement of 3 dB between similar in construction or design.’’ If a To meet the relative volume change each operating condition ensures the manufacturer calls a group of vehicles requirements, a manufacturer could volume change will be discernable by the same general name as it applies simply increase the sound levels well while providing manufacturers with the to another group, but adds a further beyond the minimum standards to greatest flexibility in the design of their description to that name (e.g., Ford achieve the required separation at each alert systems. Fusion Hybrid, or Toyota Prius Three), speed interval. However, we believe that It is NHTSA’s expectation that the the further description indicates a manufacturers will also want to reduce volume change requirement will unique model within that line. alert sounds to the greatest extent provide pedestrians with the audible Also, the NPRM conveyed that the possible while meeting the minimum cues needed to discern vehicle requirement for vehicles of the same standards in order to maximize acceleration and deceleration. However, make and model to have the same sound customer satisfaction and minimize we reiterate that frequency shifting still or set of sounds does not apply across environmental noise. To accomplish the is a useful characteristic of a vehicle model years. For example, a model year goal of minimizing excessive noise, the alert system, and we encourage system 2020 Prius Two could have a different relative volume change values should designers to incorporate frequency sound than a 2019 Prius Two (same not exceed the already established shifting even though this final rule does model but different model years). A differences of 6 dB, 6 dB, and 5 dB built not include specific requirements for it. 2019 Prius Two could have a different into the minimum operating condition Lastly, in regards to the commenters sound than a 2019 Prius Four (same specifications. The relative volume who requested that the proposed test model year but different models). All change specifications that NHTSA has procedure for frequency shifting be Prius Two’s from the 2019 model year decided to require are provided in Table modified to allow for indoor testing would be required to emit the same 17. and/or testing at the component level, sound or set of sounds (same model and those comments are no longer model year). TABLE 17—MINIMUM RELATIVE applicable since the agency has decided The PSEA includes language that VOLUME CHANGE REQUIREMENTS to exclude a frequency shifting test. In requires ‘‘the same sound or set of regard to comments about indoor and sounds for all vehicles of the same make Minimum component testing in general, we have and model.’’ We interpreted this to Critical operating scenarios relative addressed that issue in Section III.K of mean that a manufacturer may choose to volume today’s final rule, where we have stated equip a vehicle to have different sounds change, dB that NHTSA will conduct compliance for different operating modes such as Between: testing on complete vehicles on outdoor forward, reverse, and stationary [NPRM, Stationary and 10 km/h ..... 3 test tracks. p. 2804]. Each sound would have to 10 km/h and 20 km/h ...... 3 meet the corresponding performance 20 km/h and 30 km/h ...... 3 H. Sameness requirements in each operating mode. The NPRM criterion for sameness was We did not interpret this language in the These performance levels were that the alert sound of two example PSEA to mean that a vehicle can have established using the following criteria. vehicles must have a sound pressure more than one alert sound for a given First, as explained above, to minimize level within 3 dB(A) in every one-third operating mode, such as a suite of alert sound levels, the maximum octave band between 315 Hz and 5000 sounds that a driver can select from volume change between operating Hz. That requirement would limit the according to personal preference. scenarios would be 6 dB, 6 dB, and 5 amount of variation in one-third octave In general, commenters from industry dB, respectively. So, as a starting point, bands over a range of frequencies when stated that speaker tolerances make it the relative volume change measured on a stationary vehicle. We impossible to make all vehicles of the requirements should not exceed these proposed that requirement as an same year/make/model produce the values. Second, a manufacturer might objective way to determine if the alert same sound in accordance with the choose to design an alert signal that sounds produced by two different NPRM criterion, i.e., to have the same exceeds the minimum values at a given vehicles of the same make and model sound level, within ±3.0 dB, in each of speed and just meets the minimum are the same. the thirteen specified one-third octave values at the next higher speed. Such a In the NPRM, the agency interpreted bands. Also, industry commenters favor design would have a decreased relative the PSEA language on sameness as an indoor, component-level test for volume change, i.e., less than 5 dB or 6 applying ‘‘only to sound added to a sameness, rather than an outdoor test dB, between operating conditions. vehicle for the purposes of complying conducted on an ISO pad. Third, as discussed in the NPRM, the with the NHTSA regulation’’ [NPRM, p. Advocacy groups that provided sound level change that can be 2804]. The proposed sameness criteria comments on the proposed sameness discerned by an untrained observer is were not intended to apply to sounds requirement generally supported it, or approximately 3 dB, so the relative generated by a vehicle’s tires or body supported some performance-based volume change between each successive parts or by the mechanical operations of assessment of sameness, but did not operating scenario should be at least 3 the vehicle. suggest specific technical criteria for dB in order to be useful. Considering all In the NPRM, NHTSA stated that we such a performance test. these criteria, we want to target relative interpret a vehicle ‘‘model’’ as a specific Alliance/Global stated on behalf of volume changes within the range of 3 grouping of similar vehicles within a their member companies that the dB to 6 dB. Within this range, we have vehicle line. The Federal Motor Vehicle classification of sounds by an objective decided to specify 3 dB as the minimum Theft Prevention Standard,140 defines metric that would determine sameness volume change requirement for the vehicle line as ‘‘a name which a first needs to have ‘‘sameness’’ defined. transitions between successive manufacturer applies to a group of The NPRM proposal for a three decibel operating conditions. This means that vehicles of the same make that have the limit in each one-third octave band is the manufacturer can incorporate a 3 dB same body or chassis, or otherwise are not sufficient for the measurement volume change or any level above 3 dB uncertainty, let alone production to meet the specified requirements. The 140 49 CFR part 541. variation, according to Alliance/Global.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00058 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90473

Alliance/Global recommended that minimum requirements defined in the comment; the graphic showed that one- sameness be measured at a component safety standard. OICA suggested that the third octave band measurements in level under indoor laboratory language of Section S5.3 should state repeated tests of the same vehicle conditions. They stated that their only that two vehicles of the same make, appeared to vary by up to about 7 dB; practical course of action to assure model, and model year must ‘‘emit the but the results were quite different for sameness between two vehicles is to same sound within a set of sounds,’’ and the various one-third octaves and for the compare the input signals to the that their overall sound level should be different test vehicles Honda tested, speakers (the output from the signal required to be within 6 dB(A). with variability in some instances being generator or the programmed digital Denso stated that this requirement is close to zero.] Honda suggested that sound file). Alliance/Global stated that not feasible for a number of reasons. For NHTSA should specify an overall sound measuring sameness through one, there is inherent variability in level and require that there be two peak microphone recordings of operating vehicle sound characteristics and in frequencies that fall within specified vehicles is not possible as a practical speaker and amplifier characteristics frequency ranges. matter. Furthermore, due to the and performance. When combining this Advocates for Highway and Auto variation in production speakers, it also variability, it is very difficult to limit the Safety stated that, to ensure that is not reasonable to require them to emit sound difference to within 3 dB(A) different vehicles of the same make/ the same sound within the proposed between two vehicles, even for vehicles model have the same sound, the agency three decibel specification. They having nominally identical sound must establish a test procedure for acknowledged that the requirement systems, according to Denso. Denso comparing different vehicles of the same cannot be deleted altogether because it stated that sound pressure levels will make and model to ensure compliance is included in the PSEA. Alliance/ decrease by approximately one decibel and production uniformity along with Global also agreed with OICA that when the ambient temperature increases meeting the FMVSS sound NHTSA should allow manufacturers the from 0 to 40 degrees Celsius. Therefore, requirements. option of demonstrating compliance Denso suggested it is very difficult to Accessible Designs for the Blind with the sameness requirement through measure the sound level within a stated that sameness should be tested at tolerance of ±1.5 dB with good comparisons of elements such as the all speeds from idle up to the crossover repeatability in outdoor conditions. In software sound file, input to the point speed. ADB stated it does not addition, since the perception of sound speakers, etc. believe that testing at idle only is depends on ambient conditions (wind appropriate for establishing the OICA stated that the proposed direction, wind speed, temperature, standard. ADB stated that changing a sameness criterion needs revision, atmospheric pressure, etc.) and vehicle’s tires or body design is likely to pointing out that industry has already surrounding noise, Denso stated that affect the vehicle’s sound profile and shown that even 6 dB may not be a ICE vehicles of the same model have up therefore it is essential that the single sufficient tolerance between vehicles of to a 3 dB and greater sound level sound specified be well documented as the same make and model. OICA stated difference. For these reasons, Denso that the measurement uncertainty is the requested that NHTSA not adopt a detectable and localizable under most significant factor, and that the requirement for sameness. common traffic and ambient sound proposed allowance of 3 dB is not The SAE stated that, although 3 dB conditions by visually-impaired commensurate with the measurement may be an acceptable tolerance on pedestrians who are at least 60 years of uncertainty. OICA suggested that overall SPL, it is not sufficient for one- age. There will be differences in the NHTSA should carefully consider how third octave bands. SAE also stated that perceived sound even if it is generated sameness is defined as that will drive restricting one-third octave band using the same wav file. The nature of the necessary measurement procedures. variation does not guarantee sameness the loudspeaker and where and how it OICA noted that sound-generating in any reasonable sense related to this is mounted will also result in devices that use the same software will regulation. Sounds can be filtered to differences. Perceived sound will, of inherently have the same sound, even meet the same one-third octave course, also vary by road surface. ADB when the sound is altered slightly requirements, yet still could be rejected the notion that a variety of through various factors such as perceived as substantially different by sounds will be consistently and installation into a vehicle. Using the pedestrians. SAE provided an example accurately recognized by pedestrians as same software also means that vehicles of two sound files having the same coming from vehicles. Any added sound will produce the same sound even when overall SPL and very similar average should be the same for all EVs and HVs the hardware is changed somewhat, spectral distribution, but different time in order to be maximally recognized and according to OICA. OICA also noted that signals. Despite their similarities, the quickly interpreted as being a vehicular NHTSA could resolve issues with two sound files were from recordings of sound, according to ADB. ADB stated measurement of Sameness by specifying completely different sounds. SAE stated that having more than one sound is a requirement that applies to the that this demonstrates how sounds can likely to decrease any safety benefit software sound file. Citing the PSEA appear to be similar based on a selected added sound might provide for visually- language, ‘‘The Secretary shall allow measurement criterion when in fact they impaired pedestrians. manufacturers to provide each vehicle might be very different in how they In a February 2014 letter to NHTSA with one or more sounds that comply sound to listeners. co-signed by the Alliance, Global, the with the motor vehicle safety standard Honda stated the criterion for NFB, and the ACB, the co-signers jointly at the time of manufacture,’’ OICA sameness in the NPRM is too stringent submitted their mutually agreed-upon stated that vehicle manufacturers and cannot be complied with due to the position about aspects of the PSEA’s should be allowed to offer vehicles to variability of sound-producing devices. sameness requirement. They stated that customers with more than one alert An attachment to Honda’s comment vehicles with the same overall sound sound and to equip vehicles with graphically represented the variability pressure level, within a reasonable multiple alert sounds for the driver to in repeated testing of the same vehicles. engineering and manufacturing select from during vehicle operation, as [We note there was very little tolerance, should be considered as long as each of the sounds fulfils the explanation of the data in Honda’s having the same sound.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00059 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90474 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

The joint letter said that vehicles of this is that the NPRM method has a sound may not show up as significant different model years should not be relatively low resolution and would not differences in the power spectrum of considered to be the same make and distinguish between tonal signals and two signals. For this metric to be useful model. In other words, only vehicles of noise signals, which are different by for evaluating sameness, it probably the same make, model, and model year definition but can have the same one- would be necessary to evaluate the should be required to emit the same third octave band spectra. statistical correlation (R2 value) of the sound. Consequently, even if two vehicles of power spectra of two sound signals and The joint commenters also expressed the same make and model were to to specify a degree of correlation that their agreement about two other aspects comply with the NPRM criterion, there must be achieved in order for the two of the PSEA Sameness requirement: would be little assurance that they in sounds to be considered the same. For First, OEMs should have flexibility to fact produce identical alert sounds. a variety of reasons including a lack of provide EV/HVs with some number of We also acknowledge the concern any established procedure using this driver-selectable sounds instead of just expressed in comments that speakers method and also repeatability concerns, a single sound; and second, OEMs used in alert systems have some we do not know if it is feasible to should be allowed to install updated inherent manufacturing variation. develop a compliance requirement sounds once per model year to address However, NHTSA has not conducted based on this method. any dissatisfaction that might arise on tests to verify the level of speaker Frequency Response Functions would the part of vehicle owners with the alert variation claimed by commenters. provide a better comparison. For some sounds their HV/EVs are originally Regarding the Alliance/Global alert sounds, the FRF could be used to manufactured with. The latter would be suggestion that overall sound pressure show that certain periodic variations are separate from updates that OEMs might levels produced by two vehicles should highly correlated between two signals. need to make to remedy a be used to determine whether they are However, other signal variations may noncompliance or for conducting a the same, we do not believe that method not be correlated. Additionally, an recall, as provided for in the PSEA. The would provide a meaningful evaluation of the FRF would require a joint commenters believe the language comparison. That approach would standardized method to synchronize the of the PSEA, which uses the terms ‘‘one merely characterize how loudly two phase between the two signals, and the or more sounds’’ and also ‘‘sound or set vehicles’ alert sounds are perceived. agency currently does not have any such of sounds’’ allows for driver-selectable That approach would not evaluate other method. sounds and voluntary updating of acoustic characteristics that make Overall, we have concluded that sounds. sounds alike such as phase or spectral comparisons using Power Spectrum We note that NHTSA did not receive shape, and it normally would not Analysis or Frequency Response comments specifically in response to distinguish between sounds that are Functions might provide a higher degree our request for comment on the extent obviously different to listeners. For of confidence than the NPRM method to which changing a vehicle’s tires or example, music, construction noise, and that two unknown signals are the same, body design would affect the vehicle’s thunder all can have the same overall A- but developing a requirement and test sound profile for the purposes of weighted sound pressure level. procedure based on these metrics for a determining whether two example compliance test application may involve Other Sameness Metrics Considered by vehicles have the same sound. considerable additional agency research NHTSA and testing. Agency Response to Comments Subsequent to concluding that a Furthermore, in order for either of In light of the comments the agency requirement based on one-third octave these metrics to be useful in a received on the NPRM sameness levels is not appropriate for the final compliance test, the measurement requirement, we have reconsidered the rule, the agency considered various variability of the data collected for a proposed requirement and have decided alternatives for objectively determining sameness evaluation would have to be that it is not appropriate for the final that alert sounds among vehicles of the extremely low, such that even small rule. We agree with at least one same make and model are the same. differences in measurements of two shortcoming that was pointed out by To address issues with the NPRM example vehicles could be attributed to several commenters: Even if two approach, we considered two additional actual differences in their alert sounds. vehicles’ alert sounds are within three types of acoustic metrics to evaluate the As discussed in the Repeatability/ dB(A) in each specified one-third octave similarity of the alert sounds on Reproducibility section (Section III.K) of band, the alerts would not necessarily vehicles of the same make and model: this preamble, we have determined that sound the same because sounds that Power Spectrum Analysis and the variability of pedestrian alert sound have identical one-third octave sound Frequency Response Functions (FRF). measurements is on the order of several pressure levels can vary considerably in These are both acoustic metrics that decibels when measured on a vehicle in terms of how they are perceived by a could be used to analyze the actual operation (although stationary tests like listener. In fact, it is possible for output of the alert system speaker to those used for Sameness tend to be completely different types of sounds to quantify the difference between two somewhat less variable.) Although the have similar one-third octave band sound signals. Both of these metrics level of variability of the NHTSA levels, even across a wide range of characterize amplitude and frequency. measurement procedure promulgated in frequency bands. The FRF is sensitive to phase as well. today’s final rule is sufficiently low for We now believe that the NPRM metric Both metrics have higher resolution stationary, reverse, and pass-by tests, we based on A-weighted one-third octave than one-third octave bands. believe it is inadequate for a sameness band sound pressure levels would be Power spectrum analysis generally evaluation using power spectra and suitable only to identify ‘‘defective’’ has resolution sufficient for signals that FRFs. For these metrics to be useful for sounds, i.e., to identify when two do not change over time. However, sameness, we would need to obtain a sounds that are intended by design to temporal differences such as time clean signal prior to its exposure to sound the same are not the same, for reversal (e.g., playing of a signal in external influences like speaker example if a particular test vehicle had reverse) and amplitude modulations tolerances and ambient noise a damaged speaker. The main reason for which change the perceived character of fluctuations.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00060 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90475

Another option would be to evaluate The joint comment of the Alliance, Compliance Evaluation of Sameness the alert signal at the point where it is Global, the NFB, and the ACB addressed After fully considering the NPRM transmitted to the alert system speaker, both the issue of ‘‘selectable’’ sounds comments on sameness and other i.e., at the speaker input. While speaker and the issue of alert sounds being acoustic metrics, we have concluded input would have very high updated or improved after vehicles are that the compliance requirement for repeatability, this approach would delivered to customers. Regarding the sameness in this final rule should not be require that the speaker inputs must be first issue, the joint commenters stated based on acoustic performance physically accessible, which the agency that they believe the PSEA allows measurements, including the one has found is not always the case. For vehicles to be equipped with more than proposed in the NPRM. The difficulties example, speakers might be integrated one sound for a given operating and unknowns with comparing direct into a sealed module that incorporates condition. This comment would mean, measurements of acoustic metrics, as the control electronics, making access for example, that a particular vehicle well as the potential need for more difficult without destructive measures. agency research in this area if we Another option is to evaluate the make/model might have an alert sound decided to use any of the metrics signal at the point where it is generated X, an alert sound Y, and an alert sound discussed above, leads us to conclude internally in the alert system. On typical Z for when the vehicle is in forward that, currently, the most effective and alert systems, this would amount to motion at a given speed, and the driver expedient way for NHTSA to evaluate evaluating the actual digital source of could select X, Y, or Z based on the alert sound, such as a wav file, or personal preference and could switch sameness is to explicitly require that an equivalent digital element of the alert among those choices at any time. specific design aspects of vehicle alert system from which the signal originates. Regarding the second issue, the joint systems must be the same, particularly NHTSA may not have the means to commenters stated the PSEA allows a the software and hardware that extract a digital file for a compliance manufacturer or dealer to provide comprise the systems. Although this approach would not be evaluation of a test vehicle and would vehicle owners with opportunities at based on acoustic measurement, it need the assistance of the vehicle any time during a model year to update would provide assurance that the design manufacturer. At that point, a more the alert sound or sounds with which of alert systems on vehicles of a given practical option might be for NHTSA to their vehicle came equipped from the make and model are consistent from one simply request that information from factory. They contended that this the vehicle manufacturer. However, vehicle to the next because the vehicle allowance exists under the PSEA even manufacturer would be certifying not even if an OEM were to provide NHTSA in cases where the original sound is not with a digital source file from two just that the sounds are the same but defective or out of compliance with the that the hardware and software vehicles of the same make and model, safety standard, and that updates may it is uncertain whether the agency could components that are used to generate be provided for aesthetic purposes the alert sound are the same from verify that they are identical. rather than for remedy of a recalled alert Because alternative acoustic metrics vehicle to vehicle. system (the latter being expressly have these issues, we believe they are This approach is consistent with the provided for in the PSEA.) not viable for a regulatory application, comments NHTSA received in response and we have decided not to adopt Given our understanding of the PSEA, to the NPRM. In response to NHTSA’s acoustic metrics for the sameness we are not including provisions request for comment in the NPRM requirement in the final rule. Instead, as requested by these commenters that regarding its proposed method of detailed later in this section, we have would allow for driver-selectable measuring whether the sound produced concluded that the final rule pedestrian alert sounds and mid-year by two vehicles was the same, the requirement for sameness should be updates of pedestrian alert sounds. As Alliance/Global joint comment stated based on certification by vehicle such, the provision in paragraph S8 of that the only way to verify sameness manufacturers that vehicles of the same the NPRM regulatory text, which was to measure the digital signal output make and model are designed to have specifically prohibits alert sound of the sound generator or to examine the identical alert sounds. That is, they modifications except for recall purposes digital sound file itself. Alliance/Global further referenced statements by OICA must certify that vehicles of the same and also prohibits systems designed so supporting a method of determining make, model, and model year are the as to allow manipulation or sameness based on the examination of same with respect to their alert system modification of the alert sound by the software and hardware making up hardware and software components, the anyone other than the OEM or a service the sound generation system. Alliance/ source of the alert sound (such as a provider, is adopted in this final rule digital file) and vehicle inputs used to Global stated in their comments that without modification. We believe that ‘‘OICA notes that current sound vary the sound, as well as all other this approach is necessary to satisfy the elements of the alert system. generating devices that use the same requirements contained in the PSEA software will inherently have the same Other Sameness Issues—Selectable language and that allowing a means for sound, even when the sound is altered Sounds and Mid-Year Updates owners to select or modify alert sounds, slightly through various factors, such as In the proposed regulatory text in the or to allow vehicle manufacturers, installation into a vehicle. The Alliance NPRM, paragraph S8 was included to dealers, or other vehicle service entities and Global agree with OICA that prevent alert sound modifications, to replace or update alert sounds NHTSA should allow manufacturers the except in case of a vehicle recall. That outside the auspices of a recall action, option of demonstrating compliance section of the regulatory text also would be in conflict with the language with the sameness requirement through prohibited systems from being designed of the PSEA. Furthermore, by not comparisons such as: The software to allow access by anyone other than the allowing driver-selectable sounds, the sound file, input to the speakers, etc.’’ OEM or a service provider, so that final rule adheres more closely to the After reviewing the comments and its individuals would not be able to tamper PSEA requirement that vehicles of a own data, NHTSA agrees that the best with or replace the alert sound in their given make and model must have the method for satisfying the requirement in vehicles. same alert sound. the PSEA to require vehicles of the same

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00061 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90476 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

make and model to make the same logic diagram, and a discussion of explicitly certifying that the following sound is to examine the hardware and processor inputs and calculations specific hardware and software software of the subject vehicles and to relating to vehicle understeer as components of the system are the same require that hardware and software to be examples of evidence that may be used from vehicle to vehicle: The alert system the same. to validate the manufacturer’s hardware components including As stated previously, we believe that certification. speakers, speaker modules, and control the Vehicle Safety Act and PSEA In the case of pedestrian alert systems, modules, as evidenced by specific requirement can be satisfied by this we are taking that approach. In our details such as part numbers and methodology. Aside from being a development of today’s final rule on technical illustrations; the location, requirement in the PSEA, requiring FMVSS No. 141, we have not orientation, and mounting of the vehicles of the same make and model to successfully devised a meaningful, hardware components within the emit the same sound limits the universe accurate and repeatable test for vehicle; the digital sound file or other of sounds produced by EVs and HVs sameness. The reasons for this are digitally encoded source; the software that pedestrians, both blind and sighted, discussed previously in this section. and/or firmware and algorithms which must be able to identify as vehicle Instead, we are including a requirement generate the pedestrian alert sound and/ sounds. This is important because that critical aspects of the alert system or which process the digital source file pedestrians must be able to recognize design must be the same from vehicle to to generate a pedestrian alert sound; the sound produced by an EV or an HV vehicle. vehicle inputs including vehicle speed as a vehicle-emitted sound for this rule We also believe that this approach is and gear selector position utilized by to reduce crashes between pedestrians consistent with the Vehicle Safety Act. the alert system; any other design and EVs and HVs. While Congress intended that NHTSA features necessary for vehicles of the If we can establish that vehicles of the issue performance standards when it same make, model, and model year to same make and model are alike with passed the Vehicle Safety Act, courts have the same pedestrian alert sound at respect to the hardware and software interpreting the Vehicle Safety Act have each given operating condition specified they utilize for their alert systems, that recognized that in some instances it is in this safety standard. information will be sufficient to necessary for NHTSA to issue a design To verify the OEM’s certification of an establish their sameness because the restrictive standard in order to achieve alert system in the agency’s annual sounds they generate would be a desired performance or to ensure compliance evaluations, NHTSA’s effectively the same. That is, if two safety.141 In Chrysler v. Department of Office of Vehicle Safety Compliance vehicles are designed the same in regard Transportation, the Sixth Circuit upheld may request that the manufacturer make to having the same software and a FMVSS issued pursuant to the Vehicle available to the agency specific design hardware to generate alert sounds, then Safety Act restricting the design of documentation relating to the alert any overall differences in the sound headlamps. The court held that the system used on same make, model, and produced would not be perceptible in a design restriction on headlamps in the model year vehicles. The meaningful way to pedestrians. Thus, standard was consistent with the documentation that a manufacturer this approach achieves the intent of the Vehicle Safety Act because it fulfilled could provide to demonstrate that the PSEA sameness requirement. the important safety purpose of ensuring sound produced by two vehicles of the Consistent with the NPRM, we are that replacement headlamps were same make and model is the same may applying the sameness criterion only to readily available to consumers. We include documents such as: A sounds added to vehicles for the believe that the provisions in this final description of the source of the alert purpose of complying with this final rule requiring that certain aspects of the sound, such as the digital sound file; a rule. In that way, tire noise, wind noise, vehicle alert sound system be the same copy of the digital file (if applicable); and any other noise associated with in all vehicles of the same make and any algorithms for processing/ vehicle motion and that is not generated model, in addition to fulfilling a manipulating the digital file to generate by the pedestrian alert system is not requirement in the PSEA, fulfils the an alert sound; vehicle inputs such as subject to the sameness requirement. safety purpose of helping pedestrians to speed signal that are needed to process We note that NHTSA has taken a recognize sounds produced by EVs and and generate the alert sound; and details similar approach in other FMVSS where HVs as vehicle emitted sounds. such as part numbers showing that we have relied on manufacturer’s To implement this approach for the vehicles of the same make, model, and assurance and documentation that a sameness requirement, we are model year are consistently equipped system is designed to comply with the modifying the proposed regulatory text with identical alert system components. safety standard. For example, when in paragraph S5.5 (was NPRM paragraph I. Customer Acceptance NHTSA created the safety standard for S5.3) to state that any two vehicles of Electronic Stability Control, FMVSS No. the same make, model, and model year In the NPRM we discussed 126, S5.6 ‘‘ESC System Technical shall generate their pedestrian alert presentations provided by vehicle Documentation,’’ was included for sound using the same external sound manufacturers regarding consumer compliance of ESC systems with an generation system including the acceptance of adding sound to vehicles understeer requirement. In NHTSA’s software and hardware that are part of to provide pedestrian detection. Nissan development of FMVSS No. 126, the the system. Furthermore, we are adding submitted a presentation stating that agency was unable to devise an a definition of Pedestrian Alert System over 60 percent of Nissan Leaf owners understeer test that was both accurate within the regulatory text of S5.5 which surveyed found that added noise was and repeatable. The agency instead took lists the common components of acceptable if the overall sound pressure the approach of identifying certain pedestrian alert systems. In this way, by level of the sound was 55 dB–A or system design characteristics and certifying that a pedestrian alert system quieter for the forward moving verifying them by requesting meets S5.5, the manufacturer is condition. information from the OEM. Standard The NPRM also discussed the ways in No. 126 lists items such as a system 141 See Washington v. Dep’t of Transp., 84 F.3d which NHTSA crafted the proposal to diagram, a written explanation of the 1222, (10th Cir. 1996); Chrysler Corp. v. Dep’t of account for concerns about the system operational characteristics, a Transp., 515 F.2d 1052, 1058 (6th Cir. 1975). community noise impacts of the

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00062 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90477

proposal so that sounds complying with not contain requirements for mid-range the NPRM requiring that the sound the requirements of the final rule would one-third octave bands from 500 Hz to produced by EVs and HVs contain tones not unnecessarily contribute to noise 2000 Hz, resulting sound would have a would make sounds complying with the pollution. In consideration of shrill unpleasant character. Alliance/ NPRM annoying to vehicle occupants. community noise impacts the NPRM Global stated that, based on past Mercedes expressed concern that omitted the mid-range frequencies from experience with shrill sounds, their including requirements for low one- the proposed acoustic requirements as members fear that costumers may be third octave frequency bands down to these are the frequencies that contribute unwilling to purchase EVs and HVs if 315 Hz and broadband content down to the most to increasing the overall sound they are equipped with sounds meeting 160 Hz may affect consumer acceptance pressure level of sound. the proposed requirements. of sounds meeting the requirements of NHTSA also conducted a draft GM stated that the proposed sound the NPRM because sounds with content Environmental Assessment (EA) to levels and operating conditions are in in this area of the spectrum are difficult analyze the environmental effects of the excess of the safety needs of pedestrians to isolate from the vehicle cabin. proposed rule. The analysis in the EA and further explained that this would Agency Response to Comments most relevant to analyzing the impact of likely result in customer annoyance the rule on consumer acceptance is the leading to customers disabling the alert As discussed in Section III.E of this single car pass-by analysis. This sound and also affecting vehicle notice, the agency made several changes analysis is designed to show what a purchases. Chrysler and Honda also to the acoustic requirements of the person standing near the road way expressed concerns about marketability NPRM in this final rule. In response to would hear when a EV or HV emitting and customer acceptance. comments from manufacturers, the final sound complying with the NPRM Toyota also stated that sounds rule allows compliance with its acoustic passed by. In an urban ambient with an meeting the requirements of the NPRM requirements by placing minimum overall sound pressure level of 55 dB(A) would be too loud and would content in the mid-range one-third a listener standing near the roadway discourage consumers from purchasing octave bands from 500 Hz to 2000 Hz. would not be able to perceive the EVs and HVs. Toyota commented that it We believe that this change will difference between a EV/HV that did not had examined customer acceptance of increase manufacturer’s flexibility to produce added sound and an EV/HV sounds meeting the NPRM create sounds that are pleasing to that complied with the requirements of specifications. Toyota used a prototype motorists and pedestrians. NHTSA does the NPRM.142 In a non-urban ambient speaker and included 56 Prius owners not believe that the overall sound with an overall sound pressure level of (ages 20 to 55 years old). Participants pressure level of sounds meeting the 35 dB(A) the difference between the were asked to drive an alert-equipped requirements of this final rule will single-vehicle pass-by for EVs/HVs vehicle on a specific route and then rate discourage consumers from purchasing meeting the minimum sound the sound. The operating conditions EVs or HVs or effect consumers requirements in the NPRM and those experienced during the study included acceptance of the requirements in the without the added sound would be 3.1 slow acceleration; 40 km/h pass-by; final rule. The overall sound pressure to 6.3 dB, depending on speed, and 10.1 slow deceleration; and 16 km/h pass-by. level of sounds meeting the dB at stationary. In the non-urban Toyota reported that 68 percent of the requirements of the final rule for the 10 ambient a single vehicle pass by of an drivers were somewhat dissatisfied or km/h pass by are between 53–56 dB(A). EV/HV meeting the minimum sound very dissatisfied with their overall According to Nissan’s presentation, 60 percent of consumers would accept requirements of the NPRM would experience with the sound emitted by added sound to their vehicle if the produce less sound than an average ICE the test vehicle. Toyota asked the overall sound pressure level of the vehicle although this difference would participants how the sound might affect their future vehicle purchases, and 54 sound was 55 dB(A) or quieter for the only be noticeable at stationary. We received several comments in percent of the drivers indicated a forward moving condition. NHTSA response to the NPRM that certain somewhat negative or very negative believes that the Nissan study indicates aspects of the proposal would be impact, while 46 percent indicated no that consumers will accept sounds annoying to passengers or drivers or impact or a somewhat positive impact. meeting the requirements of the final rule. would not be accepted by consumers. Toyota also mentioned that a sound While the minimum sound We also received several comments meeting the proposed requirements in requirements in the final rule increase from members of the general public the NPRM resulted in an increase in the above 55 dB(a) for the 20 km/h and 30 stating that the whole concept of adding interior noise relative to the same km/h pass-by tests, sound emitted from any sound to hybrid and electric vehicle with the alert system turned off. WBU commented that allowing the other sources on the vehicle, such as the vehicles would be annoying and would sound to be emitted over fewer one- tires, increases as the vehicle increases lead to decreased sales of EVs and HVs. speed as well. NHTSA believes that the Alliance/Global stated in their joint third octave bands may alleviate increased sound from these other comment that the loudness and manufacturers concerns about consumer sources will limit the extent to which frequency composition of sounds acceptance of alert systems. Several commenters also stated that drivers notice, and are negatively meeting the proposed requirements requiring a sound while the vehicle is affected by, the sound produced in would be unpleasant to vehicle stationary would lead to lower compliance with this final rule at 20 occupants. Specifically sounds with consumer acceptance of EVs and HVs. km/h and 30 km/h. minimum content in eight one-third Nissan submitted with its comment the NHTSA finds that it is difficult to octave bands would be too loud to be result of a customer survey that draw conclusions about consumer accepted by consumers. acceptance of sounds meeting Alliance/Global further stated that indicated that over 60 percent of requirements of the final rule from the because the proposed requirements did costumers would accept an idle sound with an overall sound pressure level of survey submitted by Toyota. The Toyota survey does not breakout the views of 142 NHTSA, Minimum Sound Requirements for 49 dB–A or less. Hybrid and Electric Vehicles; Draft Environmental NHTSA also received comments from the participants in the survey by Assessment (2013), at 39–40. OICA stating that the requirements in operating speed like the survey

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00063 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90478 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

conducted by Nissan. One of the The agency addressed comments 20 °C to improve test repeatability and conditions included by Toyota was a 40 regarding consumer acceptance of a to remove issues with battery cooling km/h pass-by for which the agency did sound at stationary in Section III.I of fans running. Honda also recommended not propose requirements in NPRM. this notice. We note briefly here that we that the ambient weather conditions be Furthermore, the Toyota study did not do not believe that the requirements in measured at the specified microphone state the overall sound pressure level of the final rule for EVs and HVs to emit height in FMVSS No. 141 S6.4 with a the sound to which the participants a sound at stationary will substantially tolerance of ±0.02 meters instead of the were exposed during the test. We affect consumer acceptance of the specified microphone height with a believe that reducing the number of requirements in the final rule. As tolerance of ±0.0254 meters that was required one-third octave bands to indicated by the survey conducted by proposed in the NPRM. either four or two and allowing Nissan, 60 percent of consumers Agency Response to Comments manufacturers to comply with the accepted a sound at stationary with an requirements of the final rule by placing overall sound pressure level similar to After the NPRM was issued, NHTSA minimum content in the mid-range one- the levels required by the final rule. analyzed the sound measurement third octave bands from 500 Hz to 2000 We note that the final rule does not repeatability data that it collected in Hz will allow manufacturers more contain the requirements for broadband 2012 for a Ford Fusion to determine if flexibility to create pleasing sounds. sound, low frequency content, and tones there were systematic effects of the The final EA replicates the findings of proposed in the NPRM. In satisfying the atmospheric conditions, particularly the draft EA indicating that sounds mandate in the PSEA to establish temperature, on measured sound emitted by EVs/HVs in compliance with minimum sound requirements for EVs pressure level for the vehicle’s 10 km/ this final rule will be noticeably louder and HVs, NHTSA has taken several h pass-by. This data consisted of 96 than EVs/HVs without added noise but steps to minimize the impacts of the individual measurements taken over a will produce less sound than the requirements on drivers and pedestrians six-month period from April to average ICE vehicle. For this reason we while also ensuring that these vehicles September of 2012. For each individual do not believe that the requirements in are detectable to pedestrians when measurement the following data was the final rule will lead to sounds that operating at low speed. This includes recorded: will be so loud as to be annoying to reducing the number of required bands • Overall Sound Pressure Level (dBA) drivers and pedestrians or to effect and removing requirements for tones • Temperature (°C) and low frequency content. Given these • Wind Speed (m/s) consumers’ desire to buy these vehicles. • Furthermore, according to the analysis changes from the NPRM to the final Wind Direction (degrees from rule, NHTSA believes manufacturers North) of national annual noise caused by this • final rule in the Final EA, EVs and HVs will be able to design pedestrian alert Atmospheric Pressure (Pa) • Relative Humidity (%) subject to the final rule would only be sounds that will be accepted by drivers and pedestrians. Analysis of variance for each required to emit sound in compliance variable’s effect on overall sound with this rule during 2.3 percent of all J. Test Conditions pressure level showed no statistically travel hours in urban areas.143 Ambient Temperature Range for Testing significant variation (at the a = 0.05 Therefore, the amount of time during level) for any variable over the range of which drivers and pedestrians would be In the NPRM, we proposed that, for the data. Linear modeling of all terms exposed to sounds produced in sound measurement testing, the ambient also showed no statistically significant temperature be in the range 5 to 40 °C. compliance with the final rule is limited effect on overall sound pressure level This proposal is consistent with SAE which also limits the possibility for for any variable.145 annoyance to drivers and pedestrians. J2889–1. However, SAE J 2889–1 Since ambient temperature has no This is not the case for LSVs, contains a note stating that testing of statistically significant effect on however. These vehicles have top some vehicles may not be possible in measured sound data, NHTSA agrees speeds of greater than 20 mph and less warmer weather conditions (above 20 with the commenters that we should not than 25 mph and, because final rule ° C) since such things as battery cooling restrict ambient temperatures to would require sound at speeds of up to fans (if there is one) will always be between 5 °C and 20 °C (however, we 18.6 mph, sound is likely to be nearly running. Since the NPRM proposed that note that the tendency of thermal constant for these vehicles. In addition, measurements that contain sounds management system cooling fans to these vehicles are often open, lacking emitted by any component of a vehicle’s activate at higher temperatures may windows and, sometimes doors. For this battery thermal management system be effectively limit testing to this reason, occupants of these vehicles are considered not valid, the NPRM stated temperature range). Doing so could limit likely to hear the required sounds more that SAE J2889–1 note will also apply compliance testing opportunities while so than occupants of other vehicles. to FMVSS No. 141 sound measurement not providing any test accuracy or However, we did not receive any testing. Therefore, in the NPRM repeatability benefit. We would expect a comments indicating that consumer preamble, NHTSA requested comments vehicle’s thermal management system to acceptance of sounds required by this on narrowing the permitted temperature operate more frequently in tests during final rule would be a greater issue for range to 5 to 20 °C to improve test warmer ambient conditions. As owners of LSVs than other vehicles to repeatability and to remove issues with discussed in Section III.K, the agency which this rule applies.144 battery cooling fans running. We received comments from Alliance/ has clarified when a test can be deemed invalid, including instances when 143 NHTSA, Minimum Sound Requirements for Global and Honda regarding the ambient Hybrid and Electric Vehicles; Final Environmental temperature during testing. Both cooling fans engage intermittently Assessment (2016), at p. 56. (docket NHTSA–2011– commenters were opposed to narrowing 0100). 145 For a complete analysis see, Garrott, W.R., 144 Note that the category of Low Speed Vehicles the permitted temperature range to 5 to Hoover, R.L., Evans, L.R., Gerdus, E., and Harris, is defined in NHTSA regulations as vehicles whose J.R., ‘‘2012 Quieter Vehicle Testing Report: top speed is more than 20 mph and not more than not considered LSVs and, in fact, are not regulated Measured Sound Levels for Electric, Hybrid 25 mph. Electric vehicles with top speed of 20 mph as motor vehicles, and thus are not subject to this Electric, and Low Speed Vehicles’’ Washington, DC, or less, like many electric golf carts for example, are final rule. DOT/NHTSA, November 2016.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00064 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90479

during testing. Therefore, the final rule Agency Response to Comments vehicles, light trucks, and buses with a will permit sound measurements to be The agency has considered EMA’s GVWR of 4,536 kilograms or less, and made when the ambient temperature is low-speed vehicles. The final rule only ° ° comments and agrees that the correct in the range from 5 C and 40 C. inflation pressure should be used for all specifies how NHTSA (not Honda’s other recommendation was applicable vehicles. For passenger cars, manufacturers) will perform compliance that the ambient weather conditions be multipurpose passenger vehicles, light testing and, as with other NHTSA safety measured at the specified microphone trucks, and buses (with GVWR of 4,536 standards, manufacturers may elect not height in FMVSS No. 141, paragraph kg or less) the requirement as proposed to adopt specific portions of a test S6.1, with a tolerance of ±0.02 meters. procedure if they are convinced that ± in the NPRM is appropriate. For low- NHTSA agrees that the 0.02 meters speed vehicles, the required doing so will not affect how their test tolerance instead of the proposed height results compare to the results from ± certification label generally includes tire tolerance of 0.0254 meters that was size and inflation pressure information. NHTSA compliance testing. proposed in the NPRM is more All low-speed vehicles tested to date by Self-Locking Doors consistent with SAE J2889–1. the agency’s Compliance division have In the NPRM, NHTSA proposed that The NPRM used the microphone shown the requisite tire inflation positions of S7.1 of SAE J2889–1 and the test vehicle’s doors are shut and pressure information on the certification locked for all measurements of vehicle also used the microphone height label. tolerance of ±0.02 meters. It seems pedestrian alert sounds. To address EMA’s comments and NHTSA received comments on this logically consistent to use the same ensure that all vehicles subject to the height tolerance of ±0.02 meters for the topic from OICA and Alliance/Global. new safety standard are addressed in the Commenters requested that NHTSA meteorological instrumentation. Making language relating to recommended this change is not expected to have any clarify the vehicle condition section of inflation pressure, paragraph S6.6(e) of the final rule test procedure for self- impact on the stringency of the the regulatory text has been revised. compliance test. It will merely make locking doors by adding a sentence testing slightly easier to perform. Tire Conditioning saying that in the case of self-lockable vehicles, the doors shall be locked Therefore, the final rule will have a In the NPRM, NHTSA proposed that, before starting measurement. meteorological measurement height prior to sound measurement testing, the tolerance of ±0.02 meters (±2.0 vehicle’s tires be conditioned by driving Agency Response to Comments centimeters). it around a circle 30 meters (100 feet) in NHTSA does not think that it is Tire Inflation Pressure diameter at a speed that produces a necessary to add clarification about lateral acceleration of approximately 0.5 vehicles with self-locking doors to the In the NPRM, NHTSA proposed that, to 0.6 g for three clockwise laps, regulatory text. The applicable proposed prior to sound measurement testing, the followed by three counterclockwise regulatory text, as contained in the vehicle’s tires be inflated to the laps. This tire conditioning procedure NPRM, is S6.6(b): ‘‘The vehicle’s doors recommended tire inflation pressure was derived from ISO 362, ‘‘Road Noise are shut and locked and windows are listed on the vehicle’s tire placard. for Passenger Vehicle Tires.’’ shut.’’ This seems quite clear. This text EMA recommended that NHTSA Honda and OICA recommended that requires that all doors, whether self- adopt the tire inflation pressure NHTSA not require tire conditioning locking or not, be locked prior to testing. requirements for medium and heavy prior to testing unless NHTSA can show This text is used in this final rule in re- trucks in FMVSS No. 121, Air Brake differences in measured acoustic data numbered paragraph S6.6(a). Systems. NHTSA’s proposal deviates attributable to conditioning. OICA from the test procedure in FMVSS No. recommended changing the tire Accessory Equipment 121 which states that tires will be conditioning language to state that In the NPRM, NHTSA proposed that, inflated as specified by the vehicle before sound measurements are started, for sound measurement testing, all manufacturer for its GVWR. the tires shall be brought to their normal accessory equipment (air conditioner, EMA cited two factors in support of operating conditions. wipers, heat, HVAC fan, audio/video its suggestion to harmonize the test Agency Response to Comments systems, etc.) be turned off. We also procedures in this final rule with those stated that propulsion battery cooling contained in FMVSS No. 121 for tire NHTSA does not have measured fans and pumps and other components fitment and inflation pressure. First, acoustic data showing differences that of the vehicle’s propulsion battery EMA pointed out that a conflict between are attributable to tire conditioning. thermal management system are not FMVSS No. 121 and FMVSS No. 141 However, NHTSA’s goal for tire considered accessory equipment. would add a burden to manufacturers conditioning matches the OICA NHTSA received comments on this without any safety benefit by imposing recommendation that, before sound topic from OICA and Alliance/Global. a unique tire inflation pressure measurements are started, the tires be Commenters requested that NHTSA specification for the new FMVSS. brought to their normal operating state that accessory equipment that Second, EMA stated that ‘‘the tire conditions. NHTSA also thinks that cannot be shut off need not be shut off. inflation pressures on a heavy-duty sound measurement testing with brand The commenters suggested that the vehicle’s certification label or tire new tires may produce non- compliance test procedure prohibit the information label may lead to inaccurate representative sounds due to mold vents use of any results which include sound tire inflations.’’ EMA stated that a and mold lubricant. The goal of tire from any vehicle systems other than heavy-duty vehicle’s certification label conditioning is to remove sound those which would be constantly or tire inflation pressure label contain anomalies caused by these effects. We engaged under the specified the recommended cold inflation believe that achieving this goal will performance conditions. pressures for the tires identified on require minimal effort during testing. those labels; however, it is possible that Therefore, NHTSA will retain tire Agency Response to Comments the vehicle may be equipped with a tire conditioning in the final rule for NHTSA’s goal during compliance not listed on those two labels. passenger cars, multipurpose passenger testing is to measure the sound

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00065 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90480 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

produced by the vehicle when it is in its the vehicle during testing. Therefore, we This may not matter since we believe quietest state after sale to the general believe it is necessary to specify vehicle that the external sounds generated by a public. It is not to test the vehicle in test weight in the final rule. vehicle are relatively insensitive to some artificially quiet state that will In specifying vehicle test weight in vehicle weight. Second, this vehicle test never be attained by the driving public. other rules, NHTSA has not provided a weight specification is inconsistent with These comments are in accord with weight tolerance. Organizations any other FMVSS. A given NHTSA test NHTSA’s goal for compliance testing. performing a test should make vehicle often is tested by NHTSA and by The point made by commenters, that reasonable efforts to comply with the manufacturers to determine compliance accessory equipment that cannot be shut test specifications exactly as written. with multiple 100-series FMVSS at one off need not be shut off, is sensible, is Therefore, we are choosing not to do so time, with compliance testing for one in the spirit of what NHTSA is trying to here and FMVSS No. 141 will not standard being performed right after that accomplish, and clarifies a point not contain a vehicle test weight tolerance. for another. Adopting the SAE J2889–1 addressed previously. Therefore, in the Since NHTSA agrees with the vehicle test weight specification would final rule we are adding the phrase ‘‘that commenters that the sound produced by require a test vehicle undergoing such a can be shut down’’ to the proposed a vehicle at the relatively low test sequence of compliance tests to be regulatory text of section S6.6(c) in the speeds being used for FMVSS No. 141 reloaded before and after FMVSS No. NPRM that dealt with accessory testing is not sensitive to minor changes 141 testing slightly increasing the costs equipment. The re-worded requirement in vehicle loading, minor deviations in of performing such testing. is in Section S6.6(b) of the final rule vehicle test weight from the exact values 3. Specify a vehicle test weight that is regulatory text. specified in the rule should not have specified by other NHTSA FMVSS. any effect. These test weights are different Vehicle Test Weight As to what the vehicle test weight depending on vehicle class and brake In the NPRM, we proposed that, for specified in final rule should be, system type. For pedestrian alert sound sound measurement testing, the vehicle NHTSA wants to measure sounds testing, a fairly lightly loaded weight test weight will be the curb weight (as produced by lightly loaded vehicles. We would be used, not the heavier loading defined in 571.3) plus 125 kilograms. believe that, all else being equal, the specified in some FMVSS. The vehicle Equipment, driver, and ballast should tires of a heavily loaded vehicle will test weight specifications used by other be evenly distributed between the left produce a louder sound than will the FMVSS are as follows: and right side of the vehicle. The tires of that same vehicle when it is • FMVSS No. 105 is applicable to vehicle test weight should not exceed lightly loaded. vehicles with hydraulic or electric the GVWR or Gross Axle Weight Ratings NHTSA has identified three possible service brake systems and a GVWR (GAWRs) of the vehicle. alternatives for vehicle test weight in greater than 3,500 kg (7,716 pounds). Commenters addressed three issues FMVSS No. 141. These are: FMVSS No. 105 defines Lightly Loaded related to vehicle test weight: the need 1. Retain the NPRM vehicle test Vehicle Weight (LLVW), for vehicles for the final rule to specify vehicle test weight specification. This does not seem with a GVWR of 10,000 pounds or less, weight, the need for a vehicle test to have any particular advantages and as equal to unloaded vehicle weight weight tolerance, and what the specified has multiple disadvantages. Some of the plus 400 pounds including driver and vehicle test weight should be. disadvantages are that this test vehicle instrumentation. FMVSS No. 121 is Both Alliance/Global and OICA weight specification does not match that applicable to vehicles with air brake commented that vehicle test weight has contained in SAE J2889–1; this vehicle systems. FMVSS No. 121 tests at a no effect on measured vehicle sounds. test weight specification is not used by weight equal to unloaded vehicle weight Honda commented that, since FMVSS other FMVSS; and this vehicle test plus 500 pounds including driver and No. 141 testing is being conducted at weight specification imposes weight instrumentation plus not more than an relatively low vehicle speeds (a limits on NHTSA test drivers. To additional 1,000 pounds for a roll bar maximum of 30 km/h), small changes in elaborate on the last point, since the structure on the vehicle (if needed). vehicle test weight would have a proposed NPRM regulatory text would • FMVSS No. 135 is applicable to minimal effect on measured vehicle require the weight above vehicle curb vehicles with a GVWR of 3,500 kg sounds. Alliance/Global and OICA both weight to be evenly balanced from side- (7,716 pounds) or less. FMVSS No. 135 commented that, if the final rule does to-side, the test driver for NPRM-based defines Lightly Loaded Vehicle Weight specify vehicle test weight, then, for compliance tests cannot weigh more (LLVW) as equal to unloaded vehicle practical reasons, a vehicle test weight than 62.5 kg (138 pounds). Since a 50th- weight plus 180 kg (396 pounds) tolerance should be specified. Alliance/ percentile adult male weighs 76 kg (168 including driver and instrumentation. Global and Honda both recommended pounds), the use of this vehicle test • FMVSS No. 500 is applicable to low using the vehicle test weight specified weight specification could create speed vehicles. FMVSS No. 500 defines in SAE J2889–1 (manufacturer-defined difficulties in finding drivers to perform the test weight as equal to unloaded unloaded weight + one person + compliance testing. vehicle weight plus 78 kg (170 pounds) measurement instruments). 2. Specify the SAE J2889–1 vehicle including driver and instrumentation. test weight specification for NHTSA NHTSA does not believe that any one Agency Response to Comments tests. This was the method of these alternatives is better for safety NHTSA believes that a vehicle test recommended by commenters. It would than any other. As was previously weight specification is necessary. While harmonize with SAE J2889–1, and it has stated, NHTSA thinks that the sound we have not conducted research in this the advantage that NHTSA could use produced by a vehicle at the relatively area, we believe it is reasonable to any test drivers. It has two low test speeds being used for FMVSS anticipate that if a large load (relative to disadvantages. First, it would mean that No. 141 testing is not sensitive to minor the curb weight of the vehicle) is placed the weight of the test vehicle will vary changes in vehicle loading. Therefore, in a vehicle (say 1,000 pounds in a with the weight of the test driver (i.e., NHTSA’s goal in selecting a test vehicle passenger car’s trunk or 30,000 pounds the test weight is not a precisely weight specification is to choose one on a heavy truck), there would likely be specified number of pounds above the that will minimize the economic burden some change in the sound produced by manufacturer-defined unloaded weight). of performing compliance testing. We

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00066 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90481

think that this alternative is best paragraph 7.1.2.2. This will accomplish K. Test Procedure achieved through the selection of the our two objectives of (1) having a Indoor Testing third alternative listed above changing battery’s state of charge during testing to the vehicle test weights specified by be high enough that the ICE or other In the NPRM, the agency tentatively other NHTSA FMVSS. Vehicle test vehicle non-essential systems do not concluded that outdoor acoustics testing weights will therefore be specified by automatically activate, and (2) was preferable to indoor testing in hemi- vehicle type and GVWR in the final specifying a practicable, achievable, anechoic chambers. The agency rule. battery state of charge for testing. explained that outdoor testing was more representative of real-world vehicle-to- Battery Charge During Testing Battery Thermal Management Systems pedestrian interactions, and that In the NPRM, NHTSA proposed that, In the NPRM, NHTSA proposed that outdoor tests, especially pass-by tests, for sound measurement testing, the measurements that included sounds transmit to the pedestrian not just vehicle’s electric propulsion batteries, if emitted by any component of a vehicle’s vehicle-generated sounds (e.g., engine- any, be fully charged. propulsion battery thermal management powertrain and pedestrian alert system), NHTSA received comments on this system are not considered valid. In but also sounds from the vehicle body’s topic from Advocates, Alliance/Global, addition, when testing a hybrid vehicle interaction with the atmosphere (wind Honda, Navistar, and OICA. Advocates with an ICE that runs intermittently, noise) and road test surface (tire noise). requested that NHTSA either establish a measurements that contain sounds These complete sound profiles are battery charging procedure or require emitted by the ICE would not be transmitted to the pedestrian over the that the vehicle be charged in considered valid measurements. ‘‘outdoor ambient’’ noise. Outdoor accordance with the manufacturer’s sounds also contain a Doppler shift stated charging procedure as outlined in NHTSA received comments on this topic from OICA and Alliance/Global. when the vehicle is moving relative to vehicle documentation to ensure that the pedestrian. the ICE or other vehicle non-essential Commenters pointed out that the battery’s thermal management system Conversely, the NPRM also explained, systems do not start during sound when a vehicle is tested on an indoor testing procedures. Alliance/Global and might always be running when the vehicle is performing the test scenarios. dynamometer in a hemi-anechoic OICA recommended using the language chamber, the body of the vehicle is from the charging procedure in SAE Therefore, they requested that NHTSA state that a battery thermal management static and does not produce J2889–1. OICA stated that many hybrids aerodynamic noise. The agency said that cannot be charged by external charge system that would normally be operating during the specified test it was unclear how representative the devices and that by driving the vehicle tire noise generated during rotation on a 100-percent charge level will nearly conditions need not be shut down. The commenters suggested that the the curved dynamometer test rollers is never be reached. Honda pointed out of actual tire-road noise. As explained, that controlling the battery condition of compliance test procedure prohibit the use of any results which include sound the vehicle approach and passing of the a hybrid vehicle to attain a specific level microphones could be simulated by of charge can be difficult. Honda from any vehicle systems other than those which would be constantly phasing a row of microphones next to recommended testing with the the vehicle, and interior tire noise could propulsion battery at a normal (as is) engaged under the specified performance conditions. be digitally replaced with exterior tire condition and deleting this requirement noise recordings, however, the agency as being unnecessary. Navistar Agency Response to Comments has not determined the fidelity of such recommended that batteries be charged NHTSA’s goal during compliance methods.146 The agency voiced its to the manufacturer’s recommended full concern about both the availability of state of charge. testing is to measure the sound produced by the vehicle when it is in its repeatable specifications for all aspects Agency Response to Comments quietest state after sale to the general of indoor testing and the availability of NHTSA agrees with Advocates that public. It is not to test the vehicle in hemi-anechoic chambers in which to the battery needs to be sufficiently some artificially quiet state that will conduct compliance testing. charged during sound measurement never be attained by members of the The NPRM mentioned the agency’s testing so that the ICE or other vehicle driving public. These comments are in belief that specifications for outdoor non-essential systems do not accord with NHTSA’s goal for testing have a more detailed history of automatically activate. Provided that compliance testing. The commenters’ objective and repeatable performance this condition is met, the battery’s state statement, that a battery thermal than specifications for indoor testing. of charge during sound measurement management system that would The agency noted that a substantial testing should have no impact on the normally be operating during the amount of development and refinement safety of the vehicle. NHTSA also agrees specified test conditions need not be has gone into the test procedures and with commenters that precisely shut down, is sensible and is consistent facilities used for outdoor vehicle noise controlling the battery condition of a with what NHTSA is trying to testing. hybrid vehicle to attain a specific level accomplish. Clarifying this will address The NPRM explained that SAE J2889– of charge can be difficult. However, an important test factor that was not 1 contains specifications on the cut-off getting the battery’s state of charge covered in the proposed version of the frequency of the indoor hemi-anechoic during testing high enough that the ICE regulatory text. This factor is addressed test facility and requirements. However, or other vehicle non-essential systems in S7.1.2 and S7.3.2 of the regulatory the agency stated that it was not aware do not automatically activate should be text in this final rule. We have modified of specifications for dynamometer drum feasible. both of these subsections by adding surface textures, materials, diameters, Following review of the comments, appropriate wording to include systems road loads coefficients (i.e., to produce NHTSA has decided to accept the OICA which would be constantly engaged 146 see https://www.bksv.com/en/products/ and Alliance/Global recommendations under the specified test performance PULSE-analysis-software/acoustic-application- and use the SAE J2889–1 language for conditions (backing, stationary, forward software/pass-by-noise-testing/indoor-testing-7793 the battery charge specifications in motion at specified speeds). (weblink last accessed 2November2016).

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00067 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90482 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

appropriate engine RPMs), etc. to allow SAE explained that to achieve the chambers on dynamometers requires comparable results between different goals of practical, repeatable, and further development and validation. indoor dynamometers. reproducible test results, the use of SAE J2889–1 contains specifications for Lastly, the NPRM explained that there indoor and component level test indoor testing but does not appear to are some advantages to testing indoors. facilities are necessary. Furthermore, provide the specifications for Testing in an indoor hemi-anechoic SAE stated that for measuring the dynamometer drum surface textures, chamber would not be influenced by acoustic one-third octaves at any speed materials, diameters, road loads weather conditions or high ambient greater than zero, the use of indoor coefficients (i.e., to produce appropriate noise levels that can affect outdoor facilities will be necessary to reduce engine RPMs), etc. to allow comparable testing. Indoor testing could be more measurement uncertainty. results between different indoor dynamometers and outdoor ISO 10844 predictable and time efficient than Agency Response to Comments outdoor pass-by testing because testing noise pads. time would not be limited by weather In this final rule, the agency is The agency continues to be concerned and noise conditions at the test site. The specifying performance requirements for that hemi-anechoic chambers that have agency sought comment on the vehicle-emitted sounds that are four-wheel dynamometer drive availability of hemi-anechoic facilities detectable and recognizable to a capabilities are not widely available for that could accommodate indoor pass-by pedestrian as a motor vehicle in commercial testing. The agency was testing and the desirability of including operation. All components of the able to locate a large number of outdoor a test procedure for indoor pass-by vehicles’ sound profile that convey the 10844 noise pads in the United States, testing in this standard. signature of a motor vehicle in operation most of which were available for paid Auto manufacturers and groups that (including aerodynamic and tire noise) use by outside parties. As mentioned in represent them, along with SAE, stated up to the crossover speed are important the NPRM, one vehicle manufacturer facets of the vehicle’s sound in their comments that the agency stated that it has nine noise pads performance. Upon consideration of the should allow indoor testing in the throughout its global operations and we above comments, and as explained compliance test procedure. According to believe the standardized outdoor noise further below, the agency has decided to Alliance/Global, OEMs would prefer pads have widespread commercial only specify requirements for outdoor and support the use of indoor availability. testing as proposed in the NPRM. While indoor testing is appealing measurement facilities meeting Vehicle manufacturers may choose to because it eliminates inclement weather specifications contained in SAE J2889– test their vehicles indoors but the final and seasonal downtimes, which may 1and ISO 16254. Alliance/Global 147 rule has not added that option to the provide more flexibility for explained that in consideration of the regulatory text. manufacturers, we believe this is practicability and repeatability of the As previously mentioned, the agency outweighed by the fact that outdoor required tests, they believe that the test believes that outdoor testing is more testing will provide a more conditions specified in the final rule representative of real-world vehicle-to- representative real-world condition should allow both the outdoor testing pedestrian interactions, and that including realistic interaction of the and indoor hemi-anechoic testing which outdoor tests, especially pass-by tests, vehicle and vehicle alert system with are specified in SAE J2889–1. The reproduce not just vehicle sounds that the outdoor environment. The NHTSA Alliance/Global mentioned that some of are internally generated (e.g., engine- acoustic measurement procedures its members have indoor hemi-anechoic powertrain and pedestrian alert system), incorporate strategies such as the chambers for pass-by testing and some but also sounds from the vehicle body’s rejection of test runs having extraneous do not, but all can gain access to them. interaction with the atmosphere (wind background noise to ensure that Honda stated it is necessary to noise) and road test surface (tire noise). interaction with the outdoor include indoor test procedures in the When a vehicle is tested on an indoor environment does not affect test results. final rule and requested the agency dynamometer in a hemi-anechoic Several of the commenters explained allow use of an anechoic chamber as an chamber, the body of the vehicle is that we should allow indoor testing as option for system testing. Honda stated static and does not produce specified in SAE J2889–1. In addition to that this option will be more practical aerodynamic noise. Additionally, the conducting indoor testing in a hemi- for automakers and can yield more agency does not know how anechoic chamber using a dynamometer consistent and repeatable results representative the tire noise generated to simulate vehicle motion, it is possible without compromising the quality of the during rotation on the curved to conduct pass-by testing in an indoor sound measurements. Honda explained dynamometer test rollers is of actual hemi-anechoic chamber, provided that indoor chamber tests are necessary tire-road noise. sufficient space is available to allow not only for pass-by tests, but for To date, the agency has had limited testing of all test conditions. SAE J2889– stationary vehicle tests using an experience and access to testing for and 1 seems to allow for both methods of artificial speed signal and component- measuring acoustic sound levels on indoor testing. Full vehicle indoor pass- based pitch shifting tests. dynamometers in hemi-anechoic test by testing in a hemi-anechoic chamber OICA stated that indoor test facilities chambers. As we stated in the NPRM, without a dynamometer (i.e., an indoor meeting the specifications in SAE J– the test setup and test execution track) would capture elements of the 2889–1 are an acceptable alternative to procedures for outdoor testing have long vehicle sound profile (including outdoor testing. According to OICA, been established.148 As mentioned aerodynamic and tire noise) that hemi-anechoic test facilities are widely previously, a substantial amount of contribute to the detectability of the available for testing and should be development and refinement has gone vehicle’s sound signature until the allowed but not required. OICA into the test procedures and facilities vehicle reaches the crossover speed. mentioned that some specifications for used for outdoor vehicle noise testing. Therefore, indoor pass-by testing in a the facilities will be needed but did not Establishment of corresponding indoor hemi-anechoic chamber is able to record elaborate further. procedures to be used in hemi-anechoic all aspects of the vehicle’s sound profile while still achieving the convenience 147 NHTSA–2011–0148–0251. 148 78 FR 2836 (Jan. 14, 2013). and efficiency advantages of indoor

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00068 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90483

testing. In this case, an indoor pass-by mean with 95% confidence depending variability inherent in the procedures procedure, without a dynamometer, on the one-third octave band being utilized indoors and outdoors. For the would be the same as the outdoor pass- analyzed. Reproducibility of estimated one-third octave bands, higher levels of by procedure contained in Section means between the two indoor tests variability were noted between several 7.1.5.4 of SAE J2889–1 DEC 2014 except sites was about 2 dB on average; indoor facilities and between indoor that the 50-meter radius free of however, individual measurements had and outdoor facilities. The variability reflecting objects around the test track significant variation resulting in a 95% noted may be associated with different would not apply. The provision in SAE confidence interval range of +/¥2.5 dB dynamometers used and the fact that the J2889–1 DEC 2014 that the hemi- to +/¥6.7 dB depending on the one- comparison vehicles were not in all anechoic chamber used for indoor pass- third octave band. cases the exact same vehicles. The by testing comply with ISO 3745 or ISO In addition to comparing the two agency believes that further research 26101 would ensure that reflection from indoor test facilities to one another, both and specification refinements are the test enclosure would not interfere facilities were also compared with required to establish and properly with the vehicle’s sound measurement. outdoor measurements made at TRC. validate indoor testing utilizing The Alliance/Global 149 mentioned Measurement reproducibility between dynamometers. Further discussion on that some OEMs have indoor facilities each indoor test facility and TRC was test repeatability and reproducibility is large enough to execute full vehicle evaluated by comparing the average provided in Section III.K of this pass-by tests at required test speeds but values of each vehicle at each one-third document. octave band for each speed at the did not provide corresponding details. In conclusion, after considering recent The agency is not aware of the respective sites. Results indicate that the indoor facilities tend to have higher agency research and the comments availability of hemi-anechoic chambers received on the NPRM, the agency that are large enough to accommodate acoustic sound levels, especially at 20 and 30 km/h. Because the differences continues to believe outdoor testing on indoor pass-by tests and continues to an ISO test pad is preferable to indoor believe that the existence of such are smaller at 10 km/h, it is not likely that the differences in acoustic testing in hemi-anechoic chambers with facilities is limited, which would be an dynamometers. Section S7 of the final issue if NHTSA favored this approach as reflections from the indoor floor and the outdoor pavement are causing the rule specifies the test procedures for an option and wanted to conduct its outdoor testing. own compliance testing in such an difference. Rather, it is likely that the environment. tire/dynamometer interaction is We again note that vehicle SAE stated that when measuring the producing the higher sound pressure manufacturers’ testing can deviate from acoustic one-third octaves at any speed levels. We believe that these results the procedures in an FMVSS, which in excess of zero, the use of indoor show that it may be necessary to communicate the method the agency facilities is necessary to reduce conduct further studies about the tire/ will use to determine whether a vehicle measurement uncertainty. SAE also dynamometer interaction before any complies with the requirements of that explained that to achieve the goals of level of confidence can be established standard. Vehicle manufacturers may practical, repeatable, and reproducible with the procedures utilizing a choose to test their vehicles indoors for test results, the use of indoor and dynamometer. Because our research the purpose of demonstrating component level test facilities are shows that the tire/dynamometer compliance with the standard, but the necessary. NHTSA has issued a interaction could influence the final rule has not added that option to technical report presenting an analysis repeatability of the test and because the regulatory text. The agency believes of its indoor test data for hybrid and there are no specifications for that further developments, refinements electric vehicles.150 This report includes dynamometer drums or other aspects of and validation are required before the the analysis of acoustic measurements indoor testing that would increase indoor hemi-anechoic chambers in hemi-anechoic chambers equipped repeatability, we believe that the equipped with chassis dynamometers with chassis dynamometers. The procedures for indoor testing are not can be specified by the agency. If further analysis includes data for electric, currently sufficient to be used by the developments, data and information hybrid, and internal combustion engine agency for compliance testing. become available in the future the vehicles and examines ambient noise, Considering confidence intervals of agency may decide at that time to revisit estimated mean values for individual repeatability and reproducibility of the possibility of adding the indoor vehicle/speed/frequency pairs, the vehicle acoustic signals (measurements). testing option. standard deviation between TRC and The analysis includes a limited MPG was as high as 5 dB and the Test Surface for Compliance Testing comparison of indoor and outdoor test standard deviation between TRC and data provided by Transport Canada and In the NPRM, NHTSA proposed that IAC was as high as 4.7 dB. Thus 95% NHTSA in conjunction with confidence intervals would be as large the test surface used during compliance Transportation Research Center (TRC). ¥ ¥ testing meet the requirements of ISO Test results between two indoor test as +/ 9.8 and +/ 9.2 dB respectively. It is important to keep in mind that 10844:2011. sites (General Motors Milford Proving these confidence intervals included not NHTSA received comments on this Grounds (MPG) and International only site-to-site differences, tire/ topic from OICA, Alliance/Global, and Automotive Components (IAC)) and one dynamometer differences, and EMA. OICA and Alliance/Global outdoor test site (TRC) were compared. differences as a result of using different recommended that NHTSA allow Repeatability, as measured by standard vehicles and in some cases different compliance testing on a test surface errors for each indoor site was good. model years, therefore, these confidence meeting the requirements of either ISO The estimated mean value was found to intervals can be considered a worst case. 10844:2011 or ISO 10844:1994. They be within 0.5 to 0.75 dB of the true It is expected that confidence intervals supported this recommendation by for the same vehicles would be smaller. stating that they believe that surfaces 149 NHTSA–2011–0148–0251. 150 See Hastings et al. ‘‘Analysis of Indoor Test In response to the SAE comment, we meeting the requirements of ISO Data for Hybrid and Electric Vehicles.’’ (2015) U.S. note the limited data available seem to 10844:1994 and ISO 10844:2011 are Dept. of Transportation, Washington, DC. demonstrate that there is measurement technically equivalent.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00069 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90484 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

Agency Response to Comments vehicle’s starting system.’’ The NPRM support the 500 milliseconds NHTSA agrees with OICA and test procedure to measure compliance requirement also should consider Alliance/Global that surfaces meeting with the proposed stationary-but-active whether a lower sound level is the requirements of ISO 10844:1994 and condition included a separate appropriate for the parked condition. Honda stated that NHTSA should ISO 10844:2011 seem to be technically microphone two meters in front of the 151 clarify the definition and the equivalent. Our understanding is that vehicle on the vehicle centerline. We measurement procedure of ‘‘after the the major impetus for the 2011 update stated in the NPRM that this other vehicle’s starting system is engaged’’ in of the ISO 10844 standard was to microphone is needed in addition to the two specified in SAE J2889–1 to the NPRM. If the definition of incorporate laser profilometry measure the sound that a pedestrian ‘‘activation is the instant when the technology that has recently become standing directly in front of a vehicle driver operates the vehicle’s starting available which allows more precise would hear. We wanted to ensure that system, then it may be possible to measurements of the porosity of the there was no drop off in sound level engage the alert sound within 500 surface. NHTSA’s understanding is that from the side of the vehicle where the milliseconds. However, it may be the majority of surfaces that are within measurement is taken to the front of the difficult to consistently achieve the the 1994 standard should pass the 2011 vehicle, where the sound would be specified one-third octave levels in each standard without change. We know that beneficial in warning pedestrians of the eight bands as specified by this was the case for the Transportation standing in front of the vehicle of its NHTSA in the proposed rule. Research Center, Inc.’s (TRC’s) ISO presence. Mitsubishi stated that the alert sound sound pad that has been used for much There were a number of comments on should start when a vehicle is shifted of NHTSA’s testing. Prior to NHTSA’s the proposed stationary-but-active out of Park, and the 500 milliseconds testing, TRC’s ISO sound pad was requirement, focusing on two aspects of interval should start at that point. certified under ISO 10844:1994. At the regulatory language: (1) The start-up Mitsubishi stated that it would be NHTSA’s request, TRC recertified their delay of 500 milliseconds for the alert technically impracticable to meet the sound pad under ISO 10844:2011; this to begin, and (2) the meaning of 500 milliseconds requirement from the required certification testing but no ‘‘activation of the vehicle’s starting moment a driver first activates the structural changes to the sound pad. system’’ for HVs and EVs. propulsion system. Mitsubishi also Thus a 1994 certified sound pad is We note here that these two issues are pointed out the need for NHTSA to likely to generate a sound profile directly related to the sound-at- define ‘‘activation of the vehicle’s equivalent to that generated on a 2011 stationary requirement which is starting system.’’ certified surface. During the NHTSA’s discussed in Section III.C, ‘‘Critical Denso commented that 500 2011 testing, a Ford Fusion vehicle was Operating Scenarios,’’ in today’s final milliseconds is not enough time to tested on both ISO 10844–1994 and ISO rule. Many of the NPRM comments initiate the alert sound, and that only 10844–2011 surfaces and no significant addressed start-up delay and definition individual vehicle manufacturers can difference in sound profile levels were of ‘activation’ to the extent that they determine how much of a delay is found. opposed any requirement for an alert necessary for a given vehicle. Denso also For light vehicle sound measurement, sound in the ‘‘Stationary-but-Active’’ said that the safety risk to pedestrians NHTSA has had no difficulties in operating condition. Because comments can be avoided if the alert sound is finding sound pads certified to ISO on the ‘‘Stationary-but-Active’’ emitted beginning at the moment that a 10844–2011 for its testing. operating condition were summarized vehicle commences motion. In that NHTSA prefers to harmonize FMVSS in that previous section of this final regard, Denso suggested introducing No. 141 with SAE J2889–1 absent rule, and we wish to avoid duplication, minimum SPL requirements for a rationale for departing from that we are not repeating all of those vehicle commencing-motion sound in standard. The updated version of SAE comments here. Rather, we focus here place of the minimum SPL requirements J2889–1 that was released in December on aspects of the Stationary-but-Active for a vehicle at ‘‘start-up and stationary 2014 specifies performing outdoor comments that directly relate to Start- but activated.’’ sound testing on a surface that meets the up, the definition of Activation, and the WMU stated that 500 milliseconds requirements of ISO 10844:1994, ISO associated measurement procedure. should provide enough time from a 10844:2011, or ISO 10844:2014. Since Commenters, mainly OEMs, said that safety standpoint because, in most NHTSA believes these three surfaces to 500 milliseconds is too rapid to emit cases, a driver does not initiate be technically equivalent, we are sound in a controlled fashion, and that movement for several seconds after first expanding the list of test surfaces it is technically unfeasible to achieve starting up a vehicle. This would give specified for FMVSS No. 141 the one-third octave band levels in that any nearby pedestrian several seconds compliance testing to include those short an interval. of acoustic warning. certified to any of the above three Advocates stated that NHTSA should We also received comments from versions of ISO 10844. provide data to support the requirement Alliance/Global stating that, for testing Based on the preceding discussion, all that the alert sound must initiate and in the stationary condition, we should types of vehicles to which this rule meet the acoustic specifications within amend the test procedure to eliminate applies will be tested on surfaces that 500 milliseconds of activation to justify the additional measurement at a point meet either ISO 10844:1994, ISO that this is an appropriate amount of two meters in front of the vehicle on the 10844:2011, or ISO 10844:2014 time to warn pedestrians. Advocates vehicle centerline since that would have specifications. also suggested the agency should applied only to the stationary test which they were in favor of excluding from the Vehicle Start-Up/Activation investigate the delay times of typical vehicles, i.e., the delay between when a final rule. The NPRM proposed in Section S5.1.1 vehicle is started and when it is able to A number of commenters challenged that a vehicle must emit sound meeting begin moving. NHTSA’s analysis to the proposed requirement on the basis the specifications for the stationary-but- that 500 milliseconds is too short an active operating condition ‘‘within 500 151 The vehicle centerline is referred to as the CC’ interval for an alert system to become milliseconds of activation of the line in the test setup diagram in J2889–1. active upon vehicle start-up because

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00070 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90485

vehicle manufacturers cannot ensure traffic, and which therefore are unable speed measurement instrumentation). that an alert system is fully engaged and to drive off until they are put into gear, Alliance/Global also requested that the operating at the required sound level in would not have to emit sound. For accuracy specification for independent such a short amount of time. example, vehicles that are parked but speed measurement equipment match Commenters stated that one reason for idling so that occupants can use the heat that contained in SAE J2889–1. this is speaker transients, i.e., once or air-conditioning would not have to NHTSA received comments on the sound production begins it takes a while emit sound. We recognize that this will speed tolerance for valid test runs while for it to stabilize. Therefore, while a distinguish EVs/HVs from ICE vehicles the vehicle is traveling forward from vehicle’s alert system may be capable of since the latter emit sound whenever Alliance/Global. They recommended emitting some level of sound within 500 their engines are running, even in Park changing the speed tolerance to ¥0.0/ milliseconds, it may not achieve the (although this may not be the case for +2.0 km/h. Their justification for specified sound pressure levels in each ICE vehicles with stop-start capability.) recommending this is to correct the one-third octave band until a On the other hand, an ICE vehicle could inconsistency between the standard’s considerably longer time has elapsed be parked with its ignition in the ‘ON’ performance requirement and after start-up. position but with its engine not running. compliance test procedure while still Commenters also questioned how We have decided to maintain the use maintaining an overall tolerance of 2.0 NHTSA intends to measure the lag time of the additional front-center km/h. between starting system activation and microphone for determining compliance Agency Response to Comments the initiation of the alert sound. OEMs with the stationary-but-active and industry groups commented that requirement. We believe this is NHTSA wants to harmonize FMVSS the NPRM did not define what important to ensure that pedestrians No. 141 with SAE J2889–1 when ‘‘activation of a vehicle’s starting standing or passing in front of EVs and feasible and consistent with the system’’ means exactly. Without an HVs are able to detect them. If the agency’s focus on safety. For the exact definition, any attempt to measure agency did not ensure that sounds instrumentation used to measure the lag time would be subject to produced by EVs and HVs met the vehicle speed during compliance arbitrary selection of a starting point minimum sound requirements in testing, we see no reason not to which could result in inconsistent today’s final rule two meters in front of harmonize with SAE J2889–1. Allowing independent speed measurements. the vehicle it would be possible that a measurement will not affect compliance pedestrian standing in front of an EV or Agency Response to Comments test severity (or the safety benefits HV would not be able to hear it within As a consequence of our decision provided by this standard) because the the vehicle’s safe detection distance. discussed in Section III.C of this final 10 meters between the AA’ Line and the rule to require sound at stationary only Vehicle Speed During Compliance PP’ Line is not enough distance to when a vehicle’s gear selector is not in Testing permit the vehicle to vary more than ‘‘Park,’’ and also due to the fact that In the NPRM, NHTSA proposed that minimally from the target speed. In the most recent versions of SAE vehicles are designed so that they must the instrumentation used to measure J2889–1, the accuracy specification for be in ‘‘Park’’ in order to be started, the vehicle speed during compliance testing the continuous speed measurement proposed requirement for an alert to be capable of continuous speed instrumentation (±0.5 km/h) is tighter initiate within 500 milliseconds of measurement over the entire zone from than the earlier SAE J2889 (Sept 2011) vehicle activation is no longer the ‘AA’ Line to the ‘BB’ Line with an applicable. Therefore, that proposed ± version and the NHTSA’s proposal of accuracy of 1.0 km/h. ± requirement is not included in this final NHTSA’s proposal also set a speed 1.0 km/h. The SAE J2889–1 continuous rule. tolerance for valid test runs. For a test speed measurement accuracy In addition, our decision on sound-at- run to be valid, the vehicle speed must specification is known to be both stationary obviates the need for NHTSA be within ±1.0 km/h of the target speed feasible and practical since NHTSA’s to define the term ‘‘activation of the for that run as the vehicle travels commercially-purchased sound vehicle’s starting system’’ as it appeared through the measurement zone from the measurement equipment package in the proposed S5.1.1 regulatory text. AA’ Line to the PP’ Line. includes speed measurement Because alert system engagement will instrumentation with an accuracy NHTSA received comments on the ± not depend on when a vehicle is started, instrumentation used to measure specification of 0.1 km/h. The SAE no definition of ‘‘activation’’ is vehicle speed during compliance testing J2889–1 independent speed necessary. measurement accuracy specification from Honda and Alliance/Global. ± We note that this decision does not Commenters requested that NHTSA ( 0.2 km/h) is tighter than the SAE mean that vehicles would have to be in allow independent,152 as well as J2889–1 continuous speed measurement motion before they are required to emit continuous, speed measurement during accuracy specification. While NHTSA an alert sound. Vehicles that are not compliance testing. Honda requested does not have first-hand knowledge of moving must emit an alert sound unless that the accuracy specification for speed independent speed measurement, we they are in a condition typical of a measurement equipment match that believe that the SAE J2889–1 accuracy vehicle that may remain parked for contained in SAE J2889–1 (± 0.5 km/h specification should be both feasible some time. Vehicles that are stationary for continuous speed measurement and practical. Therefore, NHTSA still would have to emit sound if they devices or ± 0.2 km/h for independent accepts Honda’s recommendation and are, for example, waiting at a red traffic will make the FMVSS No. 141 speed light (assuming the drivers do not shift 152 SAE J2889–1 defines independent speed measurement instrumentation accuracy to Park, in the case of automatic measurement as being when two or more separate specification identical to that contained transmission vehicle, or apply the devices are used to measure the vehicle’s speed as in the most recent version of SAE parking brake in the case of manual it crosses the AA’, BB’, and PP’ Lines. In J2889–1. comparison, continuous speed measurement uses transmission vehicles). This means that one device to measure the vehicle’s speed as it Alliance/Global made a good point vehicles that are in Park with an travels through the entire zone from the AA’’ Line regarding the speed tolerance for valid activated ignition and which are not in to the BB’ Line. test runs while the vehicle is traveling

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00071 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90486 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

forward. NHTSA’s proposal required the Repeatability/Reproducibility fan by requiring that any vehicle sound vehicle to emit sounds having a NHTSA is addressing measurement measurements taken while the cooling specified level that varied with the variability in the final rule as a result of fan is operating be discarded. At the speed of the vehicle. The required level comments that were received on the time, the agency believed that this varied in a stepwise manner with the NPRM, coupled with additional testing helped address repeatability issues steps occurring at multiples of 10 km/ and analysis conducted by the agency caused by battery cooling fans. The h, i.e., at 10, 20, and 30 km/h. In other which indicate that measurement NPRM required that for all operating words, NHTSA proposed that the repeatability and reproducibility (the conditions, four consecutive valid vehicle emit sound with one sound latter across test facilities), may impact measurements be within 2 dB(A). As explained, this repetition and decibel pressure level at, for example, 9.9 km/ compliance testing results if not level restriction would ensure h and with a different sound pressure properly accounted for. The NPRM repeatability of vehicle sounds without level at 10.0 km/h. NHTSA also discussed how the agency would the presence of unwanted ambient proposed that compliance testing be attempt to minimize test variability. spikes, other non-vehicle sounds, or performed at multiples of 10 km/h, i.e., However, adequate treatment was not intermittent sounds the vehicle may at 10, 20, and 30 km/h. The problem is given to the potential effect happen to make that are not associated that, when testing at, for example, 10 measurement tolerance may have on ± with its normal operating sound. km/h, due to the 1.0 km/h speed compliance testing. The agency received individual tolerance, valid tests could be A critical component of every Federal comments from Honda, Alliance/Global, performed at any speed from 9.0 motor vehicle safety standard is a Toyota, SAE, Nissan, and Denso. These through 11.0 km/h, inclusive. Therefore, compliance test procedure that is comments generally fell into two a test performed at 9.9 km/h would be objective, repeatable and reproducible. categories: The expected variance in a valid test as would a test performed at The test procedure must be objective recorded measurements in terms of size 10 km/h. However, as previously such that differing parties, including and sources of variability; and the discussed, these two tests would have OEMs and test laboratories will consequences of manufacturers taking different required sound pressure levels. interpret and execute the procedures the steps to address repeatability in same way. The test procedure must be The Alliance/Global suggestion would compliance testing. repeatable and reproducible such that avoid this problem by changing the Honda offered two comments the results obtained are the same results speed tolerance to ¥0/+2 km/h. This regarding measurement variability. The from test-to-test at the same test facility would mean that a valid 10 km/h test first dealt with outdoor testing stating and across different test facilities. ‘‘The Notice of Proposed Rulemaking would have to have a speed in the range In the NPRM, the agency discussed its from 10.0 to 12.0 km/h, inclusive. (NPRM) requires testing of the one-third approach for minimizing test variability. octave requirement at an outdoor site, Alternatively, the proposed 10 km/h The test procedure specified in the pass-by compliance test would become but we are concerned that this poses NPRM requires that all tests be practical concerns due to the low an 11 km/h pass-by test with a ±1.0 km/ conducted on a track with a surface that repeatability of test results which will h speed tolerance. meets the requirements of ISO be influenced by the presence of The Alliance/Global suggestion is a 10844:2011 which specifies, among background noise.’’ Honda also departure from SAE J2889–1 (which has other things, a very particular type of explained that it believes the ‘‘like a 10 km/h pass-by test with a ±1.0 km/ pavement to be used so as to minimize vehicle requirements’’ are too stringent, h speed tolerance). However, this idea the contribution of tire noise to the and practically cannot be met due to the allows NHTSA to vary the required sound measured. As mentioned in the variability of sound producing devices. level of the sounds emitted by the NPRM, using a specified test track Honda provided an attachment with vehicle in a stepwise manner with the surface would minimize test variability. plots that indicate the differences in steps occurring at multiples of 10 km/ The NPRM also contained provisions four tests by the same vehicle is more h, i.e., at 10, 20, and 30 km/h. Adopting for specific environmental conditions than 3dB. this suggestion will have only a very (temperature and wind specifications), Alliance/Global stated, ‘‘The loudness minor effect on the severity of FMVSS vehicle conditions (tire set-up and in NHTSA’s proposal is created by No. 141 compliance tests making them conditioning, door and window opening summing required broadband content in a little easier to pass since each test will adjustments, vehicle accessory settings eight one-third octave bands when the now, on the average, be performed at a and vehicle loading), and track/ sound in each band is already loud 1.0 km/h faster speed. Therefore, tires, instrumentation layout restrictions. enough for detection purposes. The aerodynamics, etc., will contribute These provisions are also important for resultant sum is a sound that is, at a slightly more sound thereby reducing minimizing test variability. The NPRM minimum, 6 dB louder than necessary. the sound that needs to be generated by explained that the instruments used to When a compliance margin (for the vehicle’s external sound generation make the acoustical measurements repeatability and reproducibility) and system. However, the differences in required under our proposal must meet production variation is added on, this the requirements of paragraph 5.1 of sounds due to this 1.0 km/h speed up proposed alert sound becomes 9–12 dB SAE J2889–1. This SAE paragraph are expected to be minor. louder than necessary. The decibel describes procedures for calibration of sound scale is logarithmic, so this Considering all of the preceding the acoustical equipment. Use of such represents a doubling in the perceived discussion, NHTSA has decided to instruments and calibration procedures sound levels.’’ adopt the Alliance/Global suggestion will ensure that test measurements can Alliance/Global further said that they and change the compliance test speed be duplicated repeatedly on the same were concerned that the run-to-run tolerance to ¥0/+2 km/h. NHTSA will vehicle at one facility, or at different test variability is greater than the levels make this revised tolerance applicable facilities. proposed in the NPRM. They stated, to all three moving vehicle compliance In the NPRM, the agency addressed ‘‘Given the uncertainties noted by SAE tests, including the 10, 20, and 30km/h the issue of intermittent vehicle sound for the measurement of one-third pass-by tests. caused by the vehicle’s battery cooling octaves proposed in the NPRM, we

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00072 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90487

suggest that the tolerance should be In general, comments received stated third octave band measurements.153 To increased to 9 dB. This applies to all that the variability present in the do this, NHTSA analyzed data from a measures of performance for compliance vehicles sound measurement is higher 2010 Ford Focus, combining over 100 purposes.’’ than the agency accounted for in the individual test runs recorded at the 10 SAE discussed measurement NPRM, and that variability could be km/h test condition, including right and uncertainties in its comments. SAE said substantial even when using the left side microphone recordings, that that for the measurements of overall measurement procedures set forth in were measured at three facilities (71 test Sound Pressure Levels (SPL) the SAE J2889–1. There was also concern runs at Transportation Research Center identified site-to-site variation at 80% expressed by the commenters that if ± in Marysville Ohio, 17 test runs at the confidence interval is 1.4 dB. SAE said manufacturers increase vehicle alert Ford Motor Company Proving Ground that the uncertainty for the sound pressure levels above the in Romeo, Michigan, and 16 test runs at measurements of one-third octave minimum standards to ensure a the Navistar test track in Fort Wayne, results ‘‘has not yet been determined,’’ reasonable compliance margin, the Indiana) over a period of 6 months. Test but will be larger than the uncertainty vehicle alert sound may become data were considered valid if there were for the overall SPL. According to SAE, excessively loud. for indoor measurements, the site-to-site no anomalies apparent in the sound variation of one-third octave levels at Agency Response to Comments recordings. The recorded files were 95% confidence interval is expected to Upon review and further analyzed using NHTSA’s sound analysis be in excess of ±2 dB. For outdoor consideration of the comments received code. measurements, the site-to-site variation it appears that the provisions for The data from the test runs were at 95% confidence interval is expected addressing variability included in the further processed using a bootstrap to be in excess of ±6.0 dB. According to NPRM and discussed above are not method 154 into three datasets, SAE, these estimated uncertainties sufficient to properly address all the test consisting of 10,000 155 samples of eight should be considered when specifying variability inherent in measuring randomly selected individual test runs, tolerances for regulatory compliance. vehicle acoustic alert sounds. To further for each facility. These samples were SAE also mentioned that any variation address the issue of variability, the then processed into the one-third octave in sound output due to vehicle agency has decided to reduce the bands utilizing the compliance minimum standards required in this component production variability will procedure (the average of the first four final rule by 4 dB in each one-third be in addition to the measurements valid test runs within 2 dB), generating octave band as further discussed below. variation noted. 10,000 sets of the 13 one-third octave Denso commented on the variability We expect sounds produced by EVs and bands between 315 Hz and 5000 Hz. of the speaker unit itself, stating ‘‘There HVs will exceed the minimum one-third is inherent variability in vehicle sound octave band values in the final rule Analyzing the datasets for the characteristics and in speaker and because manufacturers will design alert individual test sites, the maximum 95% amplifier characteristics and systems in order to ensure a margin of confidence interval for the individual performance. When combining this compliance. For this reason, we believe one-third octave bands recorded on the ± variability, it is very difficult to limit the that vehicles complying with the final TRC ISO sound pad was 1.6 dB at 800 sound difference within 3 dB(A) rule, the requirements of which have Hz and 1000 Hz. For the Ford MPG ISO between the two vehicles, even for been reduced by 4 dB in each one-third test pad, the maximum value for the vehicles having nominally identical octave band from the values provided 95% CI of the individual one-third sound systems.’’ Denso also went on to by our revised detection model, will octave bands was ±2.0 dB at 315 Hz, and comment that for a 40 degree rise in still emit alert sounds that are loud at the Navistar ISO pad it was ±1.2 dB temperature (0 °C to 40 °C) the overall enough for pedestrians to safely detect at 400 Hz. Looking at all three sites, the sound level would decrease by 1 dB. EVs and HVs. overall effective maximum variation Nissan, similar to Denso, suggested in During its research, NHTSA occurs in the 315 Hz one-third octave its comments that sound levels must be conducted a series of tests to determine band with a 95% CI of ±2.5 dB. A increased by the variation of speakers. the actual level of variability in the one- summary of the results is in Table 18.

TABLE 18—COMPARISON OF MEAN AND 95% CONFIDENCE LIMIT FOR THE ONE-THIRD OCTAVE FREQUENCIES FOR THE THREE TEST SITES

TRC Ford MPG Navistar Overall Frequency Mean level 95% Mean level 95% Mean level 95% effective 95% recorded Confidence recorded Confidence recorded Confidence confidence dB(A) limit dB(A) limit dB(A) limit limit

315 ...... 41.6 1.3 40.4 2.0 41.8 0.6 2.5 400 ...... 42.5 1.1 41.1 1.1 42.7 1.2 2.0 500 ...... 44.1 1.0 44.3 0.9 44.4 1.0 1.7

153 NHTSA Technical Report ’’ Repeatability, series, each consisting of eight runs, for a total of series were drawn in this manner which made it Reproducibility, and Sameness of Quiet Vehicle 96 runs. To improve our estimate of the variability easy to directly determine the 95% confidence Test Data’’ (2016) Gerdus, E., Hoover, R.L., and in these 96 tests, we used a bootstrap method in interval for these vehicle tests. We used a similar Garrott, W.R. which all of the 96 runs were consolidated into one procedure to evaluate vehicle measurements from 154 ‘‘Bootstrap method’’ is a statistical procedure set. Single runs then were drawn randomly from the Navistar and Ford MPG test facilities, to make wherein a data set consisting of a relatively small this set and the measurement values including one- up three data sets (one from each of the three test set of measurements is resampled many times over third octave band levels were recorded. The run to obtain a much larger data set. This can improve drawn was then returned to the set. This process facilities). statistical estimates and confidence intervals. For was repeated thousands of times using the 155 The dataset size of 10,000 was selected to example, for the Ford Fusion tests on the TRC ISO computational capability of a computer. For the maximize the overall accuracy of the analysis while sound pad at 10 km/h, NHTSA ran twelve test Fusion data, 80,000 runs comprising 10,000 test maintaining a reasonable total computation time.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00073 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90488 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

TABLE 18—COMPARISON OF MEAN AND 95% CONFIDENCE LIMIT FOR THE ONE-THIRD OCTAVE FREQUENCIES FOR THE THREE TEST SITES—Continued

TRC Ford MPG Navistar Overall Frequency Mean level 95% Mean level 95% Mean level 95% effective 95% recorded Confidence recorded Confidence recorded Confidence confidence dB(A) limit dB(A) limit dB(A) limit limit

630 ...... 46.1 1.2 45.6 1.6 46.5 0.8 2.2 800 ...... 48.4 1.6 50.4 1.3 48.3 1.1 2.3 1000 ...... 49.0 1.6 50.7 1.0 49.1 0.7 2.0 1250 ...... 48.8 1.4 50.1 1.1 48.9 0.6 1.9 1600 ...... 49.7 1.5 51.0 1.1 49.3 0.9 2.1 2000 ...... 48.6 1.5 48.7 1.0 48.0 0.5 1.9 2500 ...... 46.6 1.2 46.7 1.1 46.2 0.7 1.8 3150 ...... 45.2 1.2 45.1 1.0 44.9 0.9 1.8 4000 ...... 44.0 0.9 43.9 0.8 43.4 0.9 1.5 5000 ...... 41.9 0.8 42.0 1.2 41.5 0.8 1.6

Furthermore, NHTSA conducted NHTSA agrees with Alliance/Global, concerns about the safety standard research into the effects of speaker as well as the other commenters that requiring alert sounds that are variability on one-third octave band manufacturers will take into account excessively loud. repeatability using a limited sample of measurement variability when Ambient Noise Correction vehicles. Testing was performed on a designing alert systems to ensure group of four model-year 2014 Toyota compliance with the specified In the NPRM, NHTSA proposed that Prius V vehicles under stationary performance requirements. It is possible the ambient noise be measured for at conditions, in a hemi-anechoic that with this margin added, the alert least 30 seconds before and after a series chamber, with only the alert sound sound would significantly exceed the of vehicle tests. A 10-second sample generator active to minimize potential minimum sound requirements. As such, was then to be taken from these variability from other sources. This NHTSA has decided in this final rule to measurements and used to determine testing found that when a single vehicle reduce the minimum levels that were both the overall ambient noise SPL and was tested in the chamber, run-to-run indicated by our detectability modeling the ambient noise level for each one- variability had a 95 CI of ±0.2 dB, effort. We are implementing a reduction third octave band. The 10-second operating with only the speaker active. of 4 dB in each one-third octave band sample selected was to include ambient Overall speaker variability consists of for all test conditions to offset the levels that were representative of the more than just the repeatability of any margin of compliance that we ambient levels that occurred during the one individual speaker, as acknowledge is needed to address test actual vehicle measurement. As manufacturing tolerances will add variability and that we believe OEMs explained in the NPRM, it is important variability when multiple speakers are will build into their alert systems. As to know the background noise level tested. To estimate overall speaker discussed above, our repeatability during the test to get an accurate variability, the agency analyzed the data analysis has shown that a 4 dB measurement of the sound made by the across all four Prius vehicles tested. adjustment will be adequate for this vehicle alone. Because NHTSA’s When all four vehicles were tested in purpose. proposed requirements were established the chamber, run-to-run variability It must be made clear that the reduced using a one-third octave band basis, we increased to ±0.8 dB.156 minimum levels specified in this final stated that ambient corrections should Based upon the limited test data from rule, which include the 4-dB adjustment also be calculated on a one-third octave this analysis, NHTSA estimates an described above, are the absolute band basis. overall test variability of ±3.3 dB, minimums allowed for safety purposes. The NPRM explained that SAE J2889– including both the effective test Testing variability is not a justification 1 contains a procedure for correcting procedure variability (±2.5 dB) and the for failing to meet these minimums vehicle measurements at the overall measured speaker variability (±0.8 dB). which have been adjusted specifically to sound pressure level to account for The commenters indicated that the true address concerns about test ambient influence. In the NPRM, we variability is unknown and repeatability. The agency intends to also acknowledged that the variance of recommended that a 3 to 9 dB increase pursue potential enforcement actions on a signal is greater on a one-third octave is appropriate. To account for other, measured levels below these minimum band basis than at the overall level, and unknown sources of variability, the standards. The agency believes that by thus it may be difficult to apply the agency has decided to add an additional virtue of this 4-dB reduction in the level ambient correction procedure in SAE small tolerance to the variability specified in each one-third octave band, J2889–1 to one-third octave bands. The identified during its research. manufacturers can build a reasonable NPRM further stated that SAE J2889–1 Considering both the measured and the margin of compliance into their alert requires a peak-to-peak variation of less unknown variability, we have systems while maintaining acceptable than 2 dB in order to do a valid concluded that a tolerance of 4 dB overall sound levels. We also believe correction. We also pointed out that, adequately accounts for actual test this reduction, along with other changes even if the fluctuation of the overall variability. in the final rule compared to the NPRM sound pressure level of the ambient is such as the reduction in the number of less than 2 dB, the fluctuation in some 156 See NHTSA Technical Report ’’ Repeatability, required one-third octave bands, further individual one-third octave bands Reproducibility, and Sameness of Quiet Vehicle Test Data’’ (2016) Gerdus, E., Hoover, R.L., and addresses concerns about customer would likely be higher. To address this Garrott, W.R. acceptance, noise intrusion, and other concern, we proposed a procedure that

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00074 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90489

allowed one-third octave band disentangle the signal from the ambient during the stationary and reverse correction within certain limits on both noise in the recorded signal. scenarios. the peak-to-peak ambient fluctuation Finally, OICA questioned which 10 None of the Ford Fusion sound data and the level difference between the seconds should be analyzed out of each collected during the 10 km/h pass-by vehicle measurement and the ambient. 30-second-long ambient noise test, 20 km/h pass-by test, or 30 km/h These criteria were provided in Table 6 measurement since NHTSA did not pass-by test were within 10 dB of in the regulatory text contained in the specify which 10 seconds. ambient levels. Therefore, no ambient NPRM. They were chosen in order to noise correction was performed for any provide a high degree of confidence that Agency Response to Comments of these tests at the overall SPL and one- contamination due to an unobserved, NHTSA believes, based upon data third octave band level. random fluctuation would not impact collected and testing experience gained For the stationary scenario, 82.3 the final reported level by more than over the past several years, that percent of tests were more than 10 dB about one half of one decibel. In the measured one-third octave band sound above ambient noise levels and did not NPRM, we explained that increasing the levels generated by a vehicle can be require correction. The remaining 17.7 acceptable peak-to-peak variability in corrected for ambient noise while percent of tests needed to have either the ambient correction procedure will maintaining adequate repeatability. the overall SPL or one or more allow for testing to be conducted in NHTSA conducted a substantial measured one-third octave band levels ambient sound environments in which amount of vehicle sound measurement corrected. However, none of these tests the agency would expect to be able to repeatability testing using a 2010 Ford had measured signal levels that were make accurate measurements. NHTSA Fusion (with an internal combustion less than 3 dB above ambient noise conveyed its position that this approach engine) to develop this rule.157 That levels (the differential below which tests would increase flexibility in the testing included a large number of are considered invalid). locations and times when outdoor ambient noise measurements. Testing Electric or hybrid vehicles with an testing can be conducted without was performed on the ISO sound pad of alert meeting the requirements of this significantly compromising the accuracy the Transportation Research Center, Inc. rule may be quieter than is the 2010 of measurements. We sought comment in East Liberty, Ohio, and was analyzed Ford Fusion. This may result in more on this topic. to examine ambient noise variability. electric and hybrid vehicle sound tests NHTSA received comments on All of this testing was performed at not giving results that are 10 dB or more above ambient. Nevertheless, NHTSA ambient noise correction from Alliance/ night to minimize the ambient noise. believes that the effects of ambient level Global, Honda, OICA and SAE. The Analyses of NHTSA’s measured variability on vehicle sound comments from these organizations on ambient sound data found substantial measurement repeatability will be this topic have been divided into three variability. The overall ambient SPL issues: Validity of applying ambient limited. varied over a 15.9 dB range from a low The purpose of ambient noise correction to one-third octave bands; a of 29.5 dB to a high of 45.4 dB. The conflict in the correction procedure; and correction is to reduce variability in ambient one-third octave band levels vehicle sound measurements due to ambient measurement time interval. varied over a 24.4 dB range with a low All commenters stated that measured variations in the ambient noise level. of 13.6 dB and a high of 38.0 dB.158 This one-third octave band sound levels NHTSA uses the minimum ambient ambient sound data was measured over generated by the vehicle could not be noise levels, collected before and after a a six month period from April to corrected for ambient noise while test series, for ambient correction. By September of 2012. maintaining adequate repeatability. As doing so, the ambient noise levels are NHTSA’s calculations indicate that stated by Honda ‘‘[t]he time-to-time expected to vary little with time during these large variations in ambient noise variance of the one-third octave level of a test session. Distinct, transient loud levels had only a minimal effect on the ambient noise is large and the ambient sounds such as chirping birds, overhead noise measurement and vehicle noise measured one-third octave band sound planes, car doors being slammed, etc., measurement are not simultaneous so levels generated by the vehicle will affect the maximum ambient noise that compensating by one-third octave following ambient noise correction. levels but not the minimum ambient level is not realistic for achieving As per the procedure proposed in the noise levels. The minimum ambient repeatability.’’ All four organizations NPRM, any sound generated by the noise levels are expected to be primarily therefore recommended only performing vehicle at the one-third octave band the result of more slowly varying ambient noise correction for the level (and per SAE J2889–1 for the environmental factors such as steady measured overall SPL generated by the overall SPL) will not be corrected at all state wind speed, the test site geometry, vehicle using the procedures contained if it is more than 10 dB above the and the foliage on nearby vegetation. in SAE J2889–1. ambient noise level. NHTSA examined Therefore, NHTSA believes that the OICA questioned the proposed its vehicle sound measurement minimum ambient noise levels used for procedure to correct the measured one- repeatability testing to see how correction will typically be similar third octave band sound levels frequently this situation occurred. before, during, and after a test series. generated by the vehicle for ambient NHTSA analyzed MY2010 Ford The ambient noise correction is noise. They pointed out that the Fusion sound data measurement expected to eliminate the effects of this proposed procedure contains a repeatability for five scenarios: slowly varying ambient noise from the contradiction. It requires measurement Stationary, reverse, 10 km/h pass-by measured sound levels for a vehicle. of both the sounds generated by the test test, 20 km/h pass-by test, and 30 km/ NHTSA also recognizes that distinct, vehicle during a test and of the ambient h pass-by test. The vehicle was quietest louder events such as passing vehicles noise at the same time and using the or wind gusts could, if they were to same equipment. The problem is that 157 Garrott, W.R., Hoover, R.L., Evans, L.R., occur at certain times during a vehicle’s sound measurement during testing Gerdus, E., and Harris, J.R., ‘‘2012 Quieter Vehicle operational sound measurement, Testing Report: Measured Sound Levels for Electric, records both sounds generated by the Hybrid Electric, and Low Speed Vehicles’’ increase both the measured vehicle vehicle (signal) and ambient noise. Washington DC, DOT/NHTSA, November 2016. sound and sound measurement There is no objective method to 158 Ibid. variability. Therefore, NHTSA has

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00075 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90490 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

added regulatory text in the final rule 3 dB of the ambient, and thus it is not and unlikely to occur. If such a low stating that measurements containing necessary to consider whether bands at ambient did occur, then the overall any distinct, transient, loud sounds that difference level should be corrected vehicle SPL would require correction (e.g., chirping birds, overhead planes, or not. only if it was within 10 dB of the passing trains, car doors being slammed, Accordingly, we have decided to ambient noise, i.e., if the overall SPL of etc.) are considered invalid. Further revise the required difference between the vehicle test was quieter than 35 dB. discussion about determining the the vehicle and ambient at the one-third However, any vehicle that produces an validity of vehicle measurements can be octave band level from 10dB as overall SPL of less than 35 dB is very found in Section III.K. proposed in the NPRM to 6 dB, the same quiet and most likely would not comply In September 2014, the agency as in the draft ISO and revised SAE with the requirements of this final rule received a copy of the latest draft of ISO standards, as the threshold difference or be heard by pedestrians. SAE J2889– 16254, Acoustics—Measurement of between when one-third octave bands 1 states that in this situation, no overall sound emitted by road vehicles,159 and should or should not be corrected for SPL correction should be made. Instead, in December 2014 SAE issued a revised ambient conditions. Additionally, for the technician conducting the test version of SAE J2889–1.160 Both the one-third octave bands having 3 dB should report that the corrected overall standards are of interest to the agency to 6 dB separation between the vehicle SPL will be less than the measured because, unlike the May 2012 version of and ambient measurements, the agency signal overall SPL. NHTSA desires to SAE J2889–1, they both attempt to has decided to continue to correct as correct both overall SPL and one-third address measurements at the one-third proposed in the NPRM. The draft ISO octave band levels when necessary. octave band level as well as overall SPL and SAE standards reject all the one- Since overall SPL is the antilog of the level. These standards appear to agree third octave bands with separation less logarithmic sum of all one-third octave with the various comments, including than 6dB whereas now the agency’s band levels, the one-third octave band the comments received from SAE, procedure considers them usable in an levels will, for any wide-band sound, be advising against ambient corrections at attempt to reduce possible test burden substantially lower than overall SPL. the one-third octave band level. Both by rejecting fewer sound measurements. During NHTSA’s outdoor testing, we standards specifically state, Finally, as proposed in the NPRM, any have never seen an ambient overall SPL ‘‘Background compensation is not bands found to have a separation of less that is below 25 dB. However, we permitted for one-third octave band than 3 dB would be considered routinely have seen ambient one-third measurements.’’ Both standards also unusable. These revisions have been octave band levels below 25 dB, with specify that when analyzing the one- incorporated into the respective tables some being as low as 14 dB. third octave band measurements the in the final rule. Furthermore, for some scenarios and Finally, based upon further level of background noise in each one- one-third octave bands, NHTSA’s consideration of the comments received, third octave band of interest shall be at minimum safety standard criteria are set evaluation of the ambient data collected, least 6dB below the measurement of the at a level below 35 dB. NHTSA needs and review of the latest ISO and SAE vehicle under test in each respective a robust correction procedure that is documents received, we have decided to one-third octave band. In effect, both applicable when one-third octave band make a few additional revisions to the standards state that the one-third octave ambient levels are below 35 dB. If ambient correction paragraph S6.7 in bands cannot be corrected for ambient ambient is less than 25 dB in one or the final rule. These additional revisions noise and that the only one-third octave more one-third octave bands and the to S6.7 are as follows: bands useful for evaluation are those • Ambient corrections may be difference between ambient and vehicle bands found to have at least a 6 dB required at the overall sound pressure measurements in those bands is less difference between the vehicle level when considering which four valid than 6 dB, we still need a way to make measured value and the ambient test runs can be used for performance corrections. Therefore, NHTSA has measured value. evaulation during each operating decided to use the ambient noise The NPRM proposed that no scenario. Ambient corrections at the correction procedure regardless of the corrections are needed at the one-third one-third octave band level may also be level of ambient noise present. To octave band level when there is at least required during the one-third octave accomplish this, we have removed the a 10 dB difference between the vehicle band evaluations for each operating 25 dB limitation by deleting the first measured value and the ambient scenario. For clarification purposes column and the last two rows from both measured value. The ISO and SAE Table 6 as proposed in the NPRM will tables. standards reduce this cut-off point for be replaced with two new tables, Tables • The second column in Table 6 of one-third octave band levels to a 6 dB 6 and 7, one for overall SPL corrections the NPRM and Table 2 of SAE J2889– difference. Based upon the earlier and one for one-third octave band 1 sets peak-to-peak limits on the discussion of test data, our experience corrections when required. As in the variability of measured ambient has been that very few ambient NPRM, both of these tables are derived conditions relative to the corresponding corrections are required at the 10 dB from Table 2 in SAE J2889–1. differences measured between the difference level. Even fewer would be • The first column in Table 2 of SAE vehicle alert sound profile and the required at the 6dB difference level, J2889–1 and Table 6 in the NPRM measured ambient sound levels. which has the potential to reduce the differentiate between ambient noise According to the tables, the larger that number of test runs needed for a vehicle levels greater than or less than 25 dB. difference, the larger the acceptable compliance evaluation. We agree with We do not believe this differentiation is ambient peak-to-peak variation. OICA the commenters that one-third octave required. Table 2 in SAE J2889–1 mentioned that the proposed procedure bands are not viable if they are within applies to overall SPL correction. for ambient noise correction was NHTSA understands that SAE J2889–1 confusing and contained a 159 NHTSA–2011–0148–0334. included the 25 dB breakpoint to contradiction. According to OICA, the 160 In December 2014, SAE issued a revised SAE notes to NPRM Table 6 indicated that in J2889–1. That version of J2889–1 contains the same separate overall SPL correction because proscription on background correction at the one- an ambient noise of less than 25 dB in some test scenarios the ambient noise third octave band level as does ISO 16254. an outdoor setting is extremely quiet levels must be measured at the same

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00076 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90491

time as the actual vehicle, i.e., during correct, but report OBLtestcorr,j < of a small minority of tests, and they do the vehicle pass-by run, and using the OBLtest,j’’ appears in only one cell of the not constitute any greater test stringency same microphones. The NPRM did not table. The row containing this cell will or an increase in the required sound state how this should be done. We have only be used when the separation levels over those proposed in the considered OICA’s comment and agree between the measured vehicle sound NPRM. that the notes in conjunction with the (signal) and the ambient (either overall In response to OICA’s question as to proposed Table 6 are confusing and SPL or one-third octave band level as which 10 seconds should be analyzed contradictory. Ambient measurements appropriate) is less than or equal to 3 out of each 30 seconds (or more), during actual vehicle tests are not dB. NHTSA believes that a signal- to- NHTSA has decided that the entire possible without subjective ambient difference of 3 dB or less is too ambient noise measurement (including determination as to what sounds are small to ensure the ambient is not an interval of 30 seconds or more taken ambient noise versus what are generated influencing the measurement. before a test series and another interval by the test vehicle. NHTSA does not Therefore, test runs performed for of 30 seconds or more taken after a test intend to measure ambient and vehicle which the overall measured SPL does series) should be analyzed. Since sounds at the same time through the not exceed the ambient measured SPL ambient noise correction is based upon same microphones. The purpose of by more than 3.0 dB should be the minimum ambient noise collected column 2 is to ensure the validity and considered not valid and should not be before and after a test series, analyzing minimum variability of ambient sound used. For test runs for which the overall the entire period collected instead of files collected just prior to and after measured SPL exceeds the ambient two 10-second periods may result in a vehicle tests. The objective is to avoid measured SPL by more than 3.0 dB, it lower minimum ambient noise. Having ambient sound measurements that is possible that the measured sound a lower minimum ambient noise makes contain any distinct, transient, sounds level may not exceed the ambient sound it less likely that ambient noise (e.g., chirping birds, overhead planes, level in one or more one-third octaves. correction of the measured vehicle car doors being slammed, etc.) for When this happens, it is acceptable to sound will be necessary. In the event correcting vehicle sound files. We use the data from the one-third octave that ambient noise correction is understand that column 2 is intended to bands for which the measured sound necessary, having a lower minimum provide a quantitative method for levels exceeded the ambient sound ambient noise reduces the magnitude of determining when distinct, transient, levels by more than 3.0 dB. However, the resulting correction resulting in a sounds are too loud and risk causing the data for those particular one-third slightly easier compliance pass/fail excessive variability in ambient sound octave bands for which the measured criterion. measurements. Clearly, a high sound level was too close to the ambient It is NHTSA’s belief that making this variability in ambient sound can have a sound are considered not valid and change to the ambient noise correction procedures will have no effect on safety compounding effect on vehicle sound cannot be used. because NHTSA intends to perform pressure variability. Such variability Appropriate modifications also have compliance testing on ISO sound pads could have a major impact on been made to paragraph S6.7 of the during times with as low an ambient measurement repeatability. Due to regulatory text, describing how to noise as is reasonably achievable. This ambient differences, test results from perform ambient noise corrections. These decisions are clarifications and will minimize the need for ambient one day to another for the same vehicle refinements that are needed for noise corrections during NHTSA might not be the same. To minimize the consistent compliance testing. Because compliance testing. likelihood of ambient variability, the they address practical issues that arise agency has decided, as originally Conditions for Discarding Results from application of the ambient proposed in the NPRM, to use the correction procedures of the NPRM, The NPRM discussed the agency’s minimum ambient level instead of the which in turn are based as closely as approach for measuring the sound maximum ambient level. Use of the possible on SAE J2889–1, we believe produced by hybrid vehicles (HVs) minimum ambient was discussed in these changes are within the scope of without their associated internal more detail previously in this section. the NPRM. In one case, we deleted a combustion engines (ICEs) operating Furthermore, variability of the ambient specification that doesn’t apply to because of the need to measure the sounds measured during any vehicle NHTSA testing and thus is not relevant sound of those vehicles’ in their quietest test may also cause difficulties in for this final rule. Another change clears state. As explained, the proposal was capturing the true vehicle alert profile. up confusion arising from a designed to ensure that HVs and EVs To address OICA’s issue we have contradiction in the ambient correction emit a minimum level of sound in deleted the entire second column and table as it appeared in the NPRM. situations in which the vehicle is the associated notes from NPRM Table Another arises from the agency’s operating in electric mode because in 6. We have also added regulatory text decision to do ambient corrections at that mode these vehicles do not provide stating that measurements containing the one-third octave band level which sufficient sound cues for pedestrians. any distinct, transient, loud sounds the agency explicitly proposed in the Therefore, we proposed to control the (e.g., chirping birds, overhead planes, NPRM (some commenters disagreed situation in which an ICE engine does car doors being slammed, etc.) are with that approach, and we have start operating during a test by considered invalid. addressed those comments in this invalidating test measurements that are • The entries in some cells in Column preamble.) taken when a vehicle’s ICE is operating. 4 of NPRM Table 6 and Table 2 of SAE Overall, these technical changes are The proposed test procedure stated that J2889–1 are confusing. It is not clear consistent with the SEA/ISO standard when testing an HV with an ICE that what an entry of ‘‘Do not correct, but which the agency has referenced in the runs intermittently, measurements that report OBLtestcorr,j < OBLtest,j’’ means in NPRM and which commenters urged contain sounds emitted by the ICE are the context of a NHTSA compliance NHTSA to adhere to. Furthermore, as not considered valid. test. Since, as previously discussed, the we’ve noted, these refinements in the The NPRM also discussed that tests last two rows of NPRM Table 6 have ambient correction procedure will have occurring within the temperature range been deleted, the entry of ‘‘Do not a very minimal impact on the outcomes specified in SAE J2889–1 can produce

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00077 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90492 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

divergent results when a vehicle is intermittently, then sounds produced by whose thermal management system is tested at different temperatures. In high these vehicle systems cannot be relied always operational. ambient temperatures, the battery upon to provide sound to pedestrians The comments received from cooling fan, part of the thermal for safety purposes under all conditions. Alliance/Global were similar to those management system on electric While the proposed specifications provided by OICA. These commenters vehicles, can activate intermittently requiring four valid measurements recommended that the agency clarify for while the vehicle is operating. As within 2 dB(A) would to some extent testing purposes that all auxiliary discussed, the agency decided to address repeatability issues caused by equipment capable of being shut off address the issue of intermittent vehicle intermittent vehicle noise, the agency actually is shut off as part of the test sound caused by the vehicle’s battery explained that it wanted to guard procedure. Alliance/Global along with cooling fan by requiring that any vehicle against a situation in which OICA provided several suggested sound measurements taken while the measurements are accepted with the regulatory text edits to address their cooling fan is operating be discarded. battery cooling fans active on an EV or related concerns. While the agency believed that this the ICE engaged on a hybrid-electric if Nissan stated that given the would address repeatability issues those noise sources are intermittently complexity of EV and HEV technology caused by battery cooling fans, we noted engaged. and the expectation for future system that there may be other vehicle The agency also acknowledged, as innovation, it believes that OEMs would functions that produce inconsistent discussed in the NPRM, that it may be need to identify potential vehicle sound levels as a result of the ambient possible that not all the HVs to which systems and components which could temperature. The agency tentatively this proposal would apply are designed contribute to the overall noise concluded that we had sufficiently to be operated in EV-only mode for measurement on a model-by-model controlled this situation in the test every operating condition for which the basis. SAE explained that the 2dB criteria procedure by invalidating safety standard would specify was included in the SAE and ISO measurements in which any component requirements. Because the agency standards as a data quality check and of the vehicle’s thermal management would be testing HVs in their quietest was designed to provide some objective system (i.e. a cooling pump or fan) is state, the test procedure and engaged. We solicited comments on criteria to assist the user of the standard requirements as proposed were not to know when unrelated transient other vehicle functions that produce designed to test a vehicle that produces varying noise levels at different ambient sounds are likely to have occurred. SAE added sound while its ICE is operating. said that engineering judgment by an temperatures. Therefore, the agency stated it would Furthermore, to ensure the goal of experienced test engineer is still not require that HVs meet the required to determine when other testing the vehicle in its quietest state, requirements of the proposal for a given the NPRM specified the vehicle test unrelated sounds have occurred, and a operating condition if they are not condition that all accessory equipment decision to invalidate a measurement capable of operating in electric-only on the vehicle should be turned off. must be made. SAE noted that there mode in that operating condition. For This step was included because the may be certain accessories that cannot example, if a vehicle is not designed to vehicle’s air conditioning system, be turned off. When tested, those operate in electric-only mode above 25 heating system, and windshield wipers, accessories should be in the lowest km/h, it would not be required to meet for example, can all produce sound noise emission mode. SAE referred to the requirements in the proposal at any when activated which can introduce paragraphs 7.1.2.3 and 7.1.2.4 in SAE speed above that (e.g. at 30 km/h). The inconsistency into the acoustic J2889–1 May 2012 which further defines NPRM also included a provision to measurements. accessory loads and multi-mode The NPRM went on to explain that for exclude an HV from meeting the operation. all operating conditions, the proposed minimum sound requirement for a given Advocates for Highway Safety test procedure (and that of SAE J2889– operating condition after ten commented that the requirements 1) specified that four consecutive valid consecutive tests during which the should prohibit use of any test results measurements be within 2 dB(A). This vehicle’s ICE is operating during the which include sounds from any vehicle repetition and decibel level restriction entire test. systems other than those which would are to ensure repeatability of vehicle In response to the NPRM and the be constantly engaged under the sounds without the presence of issue of invalid test results, OICA, specified test conditions (backing, active unwanted ambient spikes, other non- Alliance/Global, Nissan, SAE and but stationary, forward motion). vehicle sounds, or intermittent sounds Advocates provided comments. Agency Response to Comments the vehicle may happen to make that are OICA recommended discarding any not associated with its quiet operating measurements that are influenced by the The agency has considered the state. presence of vehicle functions that comments received and the suggested As explained in the NPRM, the produce intermittent sounds. According changes to the regulatory text. Based on agency has no preference in how to OICA, intermittent sound sources review of the comments, NHTSA finds manufacturers choose to comply with include cooling fans and pumps, and air general agreement with the agency’s the minimum sound level requirements conditioning components. OICA said overall approach for identification of in this standard. If the agency could rely that turning off the A/C and minimizing valid and invalid test runs. The goal is on battery cooling fans on electric powertrain operation before executing a to identify and utilize those test runs vehicles or the ICEs on hybrid vehicles test will reduce the incursion of these that exhibit a vehicle’s quietest to be activated whenever the vehicle is sounds. OICA explained that operating mode. In consideration of turned on or is moving, this may be a ‘‘experienced engineers must know Nissan’s comments about the satisfactory manner for a manufacturer what is truly an intermittent sound for complexity of EV/HV technology, the to comply with the minimum sound a specific vehicle, and what is part of agency anticipates that there will be a level requirements. However, if the the normal vehicle emitted sound.’’ need to inquire about specific noise- battery cooling fans and the ICEs on OICA also asked the question about how generating technologies and systems hybrid-electrics are only running the regulation will handle a vehicle utilized on test vehicles prior to

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00078 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90493

conducting FMVSS No. 141compliance shut off during testing. That text has ANSI S1.11. ANSI S1.11 specifies a testing. We note that NHTSA uses this been added to the final rule. wide range for filter roll-off rates, and approach to enforce other safety these rates, if selected at the upper and Calculation of Results standards. For example, in FMVSS No. lower extremes of the range, could 126; Electronic Stability Control The NPRM explained that the produce different results. The agency Systems, there is a requirement for the proposed compliance test procedure sought comment on whether the test vehicle manufacturer to make available was consistent with the Society of procedure should specify a maximum technical documentation about the ESC Automotive Engineers Surface Vehicle roll-off rate that is finite. understeer countermeasures. Similarly, Standard J2889–1, ‘‘Measurement of The agency also considered in the in FMVSS No. 226, Ejection Mitigation, Minimum Noise Emitted by Road NPRM whether the procedures for there is a requirement for the vehicle Vehicles,’’ September 2011,161 and that analyzing the frequency spectrum in manufacturer to make technical several sections of the SAE standard SAE J2889–1 were sufficient to ensure information about rollover sensing were incorporated by reference into the that the results of the acoustic systems available to NHTSA. With this proposed FMVSS regulatory text. The measurements were recorded in a information, the agency can identify agency further discussed that for all consistent manner. The agency asked which systems produce noise pass-by operating conditions, the additional questions about which filter continuously rather than intermittently. proposed test procedure (and that of roll-off rates have been used, if the one- Once this is established, test runs that SAE J2889–1) specified that at least four third octave band analysis should be include sounds from intermittent ICE valid test trials must be completed done in the frequency domain or in the operations and/or intermittent thermal while recording corresponding acoustic time domain, and if an exponential management system activations can and sound measurements for each operating window should be used when will be deemed invalid. condition, and upon completion of conducting the frequency analysis. Advocates recommended modifying testing the first four valid trials with an Several organizations including the language to ‘‘prohibit use of any test overall SPL within 2 dB(A) of each other Alliance/Global (combined comment), results which include sounds from any would be chosen for analysis. We SAE, OICA, NFB, Honda, and Toyota vehicle systems other than those which explained that this repetition and submitted comments regarding the need would be constantly engaged under the decibel level restriction were to ensure to clarify the procedures for processing specified performance conditions repeatability of vehicle sound the acoustic measurements used to (backing, active but stationary, forward measurements without unwanted determine vehicle compliance. motion up to 18 mph).’’ During testing, ambient disturbances, other non-vehicle Alliance/Global stated that the NPRM all accessory equipment that can be sounds, or intermittent sounds the was ambiguous as to what SPLs should physically turned off will be turned off. vehicle may happen to make that are not be reported when four sets of OICA asked about a thermal associated with its operating mode. measurements are made with two management system that is operational The proposed rule required that for microphones. They suggested that the at all times. To address that, systems each pass-by test, the sound emitted by agency proposal was not clear if side-to- and accessories that cannot be turned the vehicle at the specified speed be side measurements are to be averaged off will be operated in their quietest recorded throughout the measurement with the lower of the four measurements mode. As mentioned by SAE, the agency zone specified in S6.4. The regulatory reported or if each side’s four agrees that engineering judgment by an text specifically stated in S7.3(a), ‘‘The measurements are to be averaged and experienced test engineer will be test result shall be the lowest value the lower measurement reported. required to determine when other (average of the two microphones) of the Alliance/Global also stated that they do unrelated sounds have occurred, and a four valid pass-bys. The test result shall not agree with the use of the SAE J2889– decision to invalidate a measurement be reported to the first significant digit 1 ambient background correction must be made. after the decimal place.’’ The proposed procedures when applied to one-third In consideration of the comments regulatory text also stated in S7.3(b), octave band measurements as proposed received and associated changes to the ‘‘The test result shall be corrected for because it differs from the ISO/SAE regulatory text that were suggested, the the ambient sound level in each one- procedures which recommends agency has decided to revise the third octave band according to the correcting for ambient background only regulatory text in the final rule procedure in S6.7 and the correction at the overall SPL level, not at the one- accordingly. criteria given in Table 6 and reported to third octave band level. According to The NPRM regulatory text addressed the first significant digit after the the Alliance/Global, its members said situations where the ICE ‘‘remains decimal place.’’ that they support the test procedures as active for the entire duration of the The NPRM also explained that to specified in SAE J2889–1. test,’’ but we also need to be concerned ensure measurements can be duplicated SAE commented that, ‘‘Section with an ICE or thermal management repeatably on the same vehicle at one S7.3(a) proposed text is unclear.’’ SAE system that operates intermittently. If facility or at different facilities, the explained that the four measurement any of these three conditions occur instruments used to make the acoustical runs are to be averaged independently during ten consecutive tests the vehicle measurements should meet the per side, and then the lower of the two is not required to meet the applicable requirements of paragraph 5.1 of SAE sides is chosen to be the intermediate or requirements. The agency has J2889–1. Since the filter roll-off rates final result, as applicable, in accordance considered the total number of tests that used affect the results of the acoustic with SAE J2889–1. The NFB supported may have to be executed to acquire the measurements at the one-third octave the SAE comments on the proper necessary four valid tests and has band level, the NPRM explained that measurement procedure. OICA said that decided to include an absolute number SAE J2889–1 requires conformance with the overall SPL values should be of tests that must be attempted before averaged per side and that the reported the test sequence can be terminated. 161 The agency recognized that SAE had final result is from the vehicle side with published an updated version of J2889–1 in May The NPRM regulatory text did not 2012. At that time we had not yet evaluated the new the lower average overall SPL level. specifically state that all accessories that version, but said we intended to do so before Toyota stated, as mentioned in the can be physically shut off should be publishing a final rule. Alliance/Global joint comment, that the

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00079 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90494 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

measurement procedure in the NPRM the overall sound pressure level. To do by will be positioned on either the left introduces significant variability within this, the agency will follow the or right side of the vehicle. Since the the results and that a more appropriate procedures specified in SAE J2889–1 pedestrian will be on one side of the measurement procedure would be that for: (1) Obtaining the ambient sound vehicle or the other as it passes, the SAE which is specified by SAE J2889–1. files both before and after execution of J2889–1 procedures appropriately select Honda stated that it supports the a series of test trials; (2) measuring the the side of the vehicle that is found to principle of taking four measurements, sound profiles for each of the first four be the quietest during the test runs. averaging the lower values from each valid test trials as appropriate for each Taking an average that includes sound side, and reporting the calculated value, test condition; and (3) determining from both the left and right per SAE J2889–1. which recorded sound files to use for microphones as specified in the NPRM In regards to roll-off filter selection for the one-third octave band evaluation. It would not provide an accurate post processing acoustic files, Alliance/ should be noted that the agency’s final representation of what any pedestrian Global supported the use of ANSI rule test procedure augments SAE would hear. Therefore, the regulatory S1.11–2004 Class 1 one-third octave J2889–1 by specifying how exactly the text has been revised to agree with the filters as specified in SAE J2889–1. selected acoustic measurements will be SAE standard. While they acknowledged the agency’s corrected for ambient conditions and As mentioned previously, Alliance/ concern regarding filter roll-off rates, evaluated at the one-third octave band Global suggested that the proposed they stated that the roll-off rate has a level, which is a critical step in the regulatory text was ambiguous in very small impact on the one-third compliance test procedure and one that regards to the steps involved in octave results (approximately 0.15 dB). is not fully detailed in SAE J2889–1. analyzing vehicle acoustic Honda also voiced concerns regarding All of the commenters indicated that measurements. Upon closer filter roll-off rates, in that specifying a the agency’s proposed ambient examination of our proposed text, we maximum and sub-infinite roll-off rate correction and test procedure, S6.7 and believe the text should be revised to add in this test procedure would represent a S7, do not exactly follow the procedures some clarification and additional detail. change to the general standard of one- in SAE J2889–1. SAE specifically said To that end, we are providing here a third octave analysis already commonly that our proposed regulatory text was detailed, step-by-step explanation in used by automakers. Honda stated that unclear, and the Alliance/Global stated conjunction with several figures to this change would create an extra our proposed text was ambiguous. More further illustrate the process. The testing burden and would require specifically, the commenters noted that corresponding regulatory text in this additional time for development of the the proposed regulatory text specified final rule has been revised accordingly appropriate test instruments and test that, for each of the four consecutive to make the procedures as unambiguous procedures. valid test runs collected during the pass- as possible. by tests, the left and right microphone Agency Response to Comments The process of executing vehicle files are averaged together and then the measurements in each test condition It has been the agency’s intention to one run with the lowest overall SPL (stationary, reverse, pass-bys), collecting 162 follow the SAE J2889–1 test value was used to evaluate the one-third necessary sound files, determining test procedures, when feasible and octaves to determine compliance. On run validity, and processing sound files consistent with the agency’s focus on the other hand, the commenters noted to verify vehicle compliance can be safety. As discussed in the NPRM and that SAE J2889–1 clearly requires that broken down into five main steps, in this final rule, the agency has decided the four data files recorded on the left which are discussed in detail later in to evaluate HVs and EVs for side of the test vehicle are averaged, and this section, and which can be briefly detectability and recognition at the one- the four data files recorded on the right summarized as follows: third octave band level rather than at side of the vehicle are averaged, and 1. For a given test condition, execute then the side of the vehicle with the test runs and collect acoustic sound 162 In the NPRM the agency officially referenced lowest average overall SPL value should SAE Standard J2889–1, dated September 2011, and files; noted that SAE had published an updated version be selected to evaluate the one-third 2. Eliminate invalid test runs and of J2889–1 in May 2012 but that we had not octave bands for compliance. discard the corresponding sound files; evaluated that later version and intended to do so The agency has evaluated these 3. Identify the first four valid vehicle before publishing the final rule. In the May 2012 comments and has further scrutinized test runs that have overall SPLs within version, SAE added testing protocols for vehicle the proposed text and the procedure commencing motion sound and for frequency shift 2dBA of each other; measurements, neither of which the agency has specified in SAE J2889–1. We have 4. Take an average of the four overall decided to utilize as discussed in this final rule. decided that the regulatory text as SPLs from the left side of the test The May 2012 version also included paragraph proposed did not match that in SAE vehicle; separately, take an average of updates and re-numbering. In December 2014, SAE J2889–1 and agree that the text should issued another revision to J2889–1. In the final rule the four overall SPLs from the right side we have decided to update the official reference for be unambiguous. We note that the of the test vehicle; the lesser of these the SAE J2889–1 standard from the September 2011 agency’s intent has been to follow SAE two averages will determine whether version to the December 2014 version and have J2889–1 as closely as possible but to the left side or right side sound data are updated references throughout the FMVSS No. 141 expand and add the necessary details standard accordingly. A number of OEMs, to be used for one-third octave band including some of those that commented on the not currently specified in SAE J2889–1 analysis. FMVSS No. 141 NPRM, are parties to the SAE for the final evaluation of the one-third 5. Evaluate either the left side or right committee that created J2889–1, and they octave bands. side sound data (whichever had the presumably had a hand in subsequent updates. The We further considered how the lower average in Step 4) at the one-third agency has decided to use the Dec. 2014 version since that is the most up-to-date and since the older recorded acoustic data files should be octave band level to determine versions seemed to leave open some important evaluated, and we have concluded that compliance. technical details that are addressed to some extent averaging the data files on each side of Each of these five steps is discussed in the latest version. Safety groups and other non- the test vehicle separately as required in in more detail below. industry commenters did not address SAE For a given test condition, execute test recommended practices, so we assume they are J2889–1 provides the most realistic indifferent about which version of the SAE standard results. During a pass-by scenario, a runs and collect acoustic sound files: To is referenced in this final rule. pedestrian listening to a vehicle driving begin the process, multiple test runs (at

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00080 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90495

least four, but generally five to seven Eliminate invalid test run acoustic identified with, and shall retain, their based on NHTSA’s experience) must be sound files: The sound files collected test run sequence number and their completed for each test condition from each microphone during each test association with left side and right side (stationary, reverse, pass-by) as run are evaluated for validity. The microphone locations. specified in the regulatory test specifics for determining validity of Identify first four valid test run sound procedures. Immediately before and each test run sound file are discussed in files within 2dBA: After a group of test after each test condition, at least 30 Section III.K, Conditions for discarding seconds of ambient noise must be measurements. Each test run deemed run sound files have been determined as recorded. During each test run, a left valid must be numbered sequentially valid, further evaluation is required to (driver’s side) and right (passenger side) based upon the chronological order in identify the ‘‘first four valid test run acoustic sound data file must be which it was executed on the test track, sound files with overall SPLs within recorded. For the stationary tests, data and each must include a left (driver’s 2dBA.’’ Figure 10 identifies a flow from a third microphone located side), right (passenger side), and for the diagram that depicts this process which directly ahead of the test vehicle is also stationary test condition a front center is derived directly from SAE J2889–1. recorded. acoustic sound file. Sound files shall be

For each test run, a valid left (driver’s of four test runs. This test run selection Step 2: Determine the maximum side) and a valid right (passenger side) process as depicted in Figure 10 is as overall SPL value for the left and right sound file must exist. For each sound follows: side sound files from each of the first 4 file the maximum overall SPL must be Step 1: Number each valid sound test runs. determined. Ultimately, the four test Step 3: Compare the four left side measurement test run sequentially in runs to be used for the compliance (driver’s side) maximum overall SPL the chronological order it was evaluation must be sequentially the first values. Calculate the difference between four valid test runs that have four left completed on the test track– e.g., Run 1, the largest and smallest of the four side files within 2.0dB(A) overall SPL Run 2, Run 3, . . . Run N. Each test run values. Use the same process to and four right side files within 2.0 must have a corresponding left (driver’s determine the difference between the dB(A) overall SPL. The left and right side) and right (passenger side) acoustic largest and smallest of the four right side files must come from the same set sound file. side (passenger side) maximum overall

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00081 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 ER14DE16.013 90496 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

SPL values. If the difference is less than 5, then the sum of its run numbers used when evaluating compliance with or equal to 2.0 dB(A) on both the left would be 12 (1+2+4+5). Select the entry the directivity requirement. The front and right sides, then these four test runs which has the lowest sum of run microphone data to be used for will be used for the compliance numbers. This set of runs is the set for directivity are the data recorded during evaluation, and the test run selection which final data will be analyzed for the ‘‘first four valid test runs within process for the given operating compliance. At this point, terminate the 2dB(A)’’ determined according to the condition is complete. The selected runs run selection procedure. This set of runs procedure above. will be considered the ‘‘first four valid will be considered the ‘‘first four valid Average sound files on test vehicle left test runs within 2dBA.’’ Otherwise, test run sound files within 2.0dBA.’’ and right sides to determine final files continue to Step 4. [Note: When there are five runs being for one-third octave band processing: Step 4: Add data from a fifth test run considered, it is mathematically After the ‘‘first four valid test runs to the analysis. impossible for the sums of the run within 2.0dBA’’ have been identified, Step 5: For the driver‘s side numbers for the two entries in the Step the four acoustic sound files from each microphone, list all possible 7 list to be exactly the same. One entry combinations of four runs for which the will always have a lower value. side of the vehicle recorded during largest overall SPL from any of the four However, in NHTSA’s experience there those four runs are analyzed to runs minus the smallest overall SPL have been cases in which six or seven determine which side of the vehicle was from any of the four runs is less than or test runs are needed to find a set of four the quietest during test execution. equal to 2.0 dB(A). shared by the driver’s and passenger’s Figure 11 is a flow diagram that depicts Step 6: For the passenger side sides that have Overall SPLs within 2.0 the process used to further identify the microphone, list all possible dB(A). It might be possible (although the acoustic data files on a particular side combinations of four runs for which the agency has not yet had it happen) in of the test vehicle that will be used to largest overall SPL from any of the four these situations for the sums of the run evaluate vehicle compliance at the one- runs minus the smallest overall SPL numbers for the two entries in the Step third octave band level. For each of the from any of the four runs is less than or 7 list to be exactly the same. If this eight acoustic sound data files (four left equal to 2.0 dB(A). occurs, our procedure will be to side files and four right side files) the Step 7: Examine the list of run eliminate the combination of four runs maximum overall SPL value must be combinations developed in both Step 5 containing the highest run number. If identified. Each of the eight acoustic and Step 6. If a set of four runs (e.g., Run data file maximum overall SPL values 1, Run 2, Run 4, and Run 5) appears in the highest run number is the same in both four-run combinations, we then are then corrected for the recorded both the Step 5 and Step 6 lists, enter ambient conditions as specified in the it into a new list (the Step 7 list). will eliminate the combination of four Step 8: The Step 7 list can possibly runs containing the second highest run final rule. Finally, the four ambient- contain zero, one, or more entries. If the number, and so on.] corrected maximum overall SPL values Step 7 list has zero entries, skip to Step Step 10: Case for which the Step 7 list on each side of the vehicle are averaged 10. If the Step 7 list contains exactly one contains zero entries. In this situation, together for one comprehensive entry, then that entry is the set of runs add data from another test run to the ambient-corrected value for each side of for which final data will be analyzed. analysis and return to Step 5. [Note: In the vehicle. The side of the vehicle with For this case, terminate the run NHTSA’s experience, there have been the lowest average ambient-corrected selection procedure. This set of runs instances in which it was necessary to maximum overall SPL value is the side will be considered the ‘‘first four valid examine data from as many as seven of the vehicle that is further evaluated test run sound files within 2.0dBA.’’ If runs to find a set of four that are shared for compliance at the one-third octave the Step 7 list contains more than one by the driver’s and passenger’s sides band level. Each of the four acoustic entry, go to Step 9. that have Overall SPLs within 2.0 data files on the side of the vehicle with Step 9: Case for which the Step 7 list dB(A).] the lowest average ambient-corrected contains more than one entry. Sum the Note that, although data recorded by maximum overall SPL value are then run numbers for each set of runs in the the front microphone are not considered used for the one-third octave band Step 7 list. For example, if an entry when determining the ‘‘first four valid evaluation as depicted in the flow contains Run 1, Run 2, Run 4, and Run test runs within 2dB(A),’’ those data are diagram in Figure 12.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00082 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90497

In the event that the average corrected corrected for the measured ambient four values calculated in each one-third maximum overall SPL values for the levels as specified in the final rule. The octave band are then averaged together driver’s and passenger’s sides are four corrected values in each one-third to get the average ambient-corrected exactly equal, then the sound from the octave band are then averaged together sound pressure level in each one-third passenger’s side will be analyzed. to get the average corrected sound octave band. The averaged, corrected Evaluate final sound files at one-third pressure level in each one-third octave values in each one-third octave band are octave band level for compliance band. The averaged corrected values in then compared directly to the minimum verification: Figure 12 indicates the flow each one-third octave band are then standards specified in this final rule to process for analyzing the selected four compared directly to the minimum determine compliance. acoustic data files for the one-third standards specified in this final rule to octave band analysis. As shown in determine compliance. As explained previously, the process Figure 11, the side of the vehicle found The stationary test condition, ‘‘first established in this final rule augments to have the lowest overall average and four valid test runs within 2dB(A)’’ also the process specified in the SAE corrected SPL value is the side of the has front microphone acoustic data. standard by clarifying the steps depicted vehicle that is further evaluated for Each sound file for the front in Figure 12 for processing the selected compliance verification. The side microphone is broken down into its sound files for the one-third octave band selected has four individual acoustic one-third octave band levels. The analysis. The current version of SAE data files. Each file is broken down into identified one-third octave band levels J2889–1 does not correct one-third its one-third octave band levels. The in each of the four files are then octave band data, as required in this identified one-third octave band levels corrected for the measured ambient final rule. in each of the four files are then levels as specified in the final rule. The

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00083 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 ER14DE16.014 90498 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

To address commenter issues use in the data post-processing routines was written using MatlabTM. While discussed above and to add that can impact the final results. All of Matlab is a proprietary engineering clarification, the final rule test these critical attributes must be based technical programming language, procedure (paragraph S7) replaces in its evaluated and defined to ensure an the source code developed for acoustic entirety the proposed regulatory text of objective test procedure is specified that data processing is the property of the the corresponding section of the NPRM. provides reproducible and repeatable United States Department of test results. Transportation and can be made Data Post-Processing Over the past few years, the agency publically available. This code uses a In the NPRM, the agency sought has used two different sound analysis more traditional, language based, comment on data post-processing topics codes for processing acoustic sound programing structure. including filter roll-off rates, files. The first code, which NHTSA The agency is aware of other acoustic measurement domains and type licensed from Bruel and Kjaer, is the measurement instrumentation and windows used for frequency analyses. B&K Pulse ReflexTM Code (the B&K associated codes that can also be used Few comments were received, but the Code), and is an integral part of a to collect and process acoustic sound one topic that was commented on was commercial off-the-shelf acoustic sound files but none of these other systems/ filter roll-off rates. The commenters measurement system. NHTSA has codes have been evaluated. It is our strongly supported using the ANSI utilized this system and software code understanding that among these codes, S1.11–2004 Class 1 one-third octave for much of its early research testing. the two used by NHTSA and some of filters as specified in SAE J2889–1. The B&K Code is a data analysis the other available codes function We agree that the ANSI S1.11 filters software that uses preprogrammed similarly. Figure 13 depicts the general should be used for processing the building blocks, known as elements, to process used by these various codes to acoustic sound files. However, as form processing chains. For the purpose derive the overall and one-third octave mentioned in the NPRM, the selected of processing sound recordings two band sound values. filter roll-off rates could affect the processing chains were used, one for The general process involves loading results of the acoustic measurements at determining the overall sound pressure the sound data file, applying the defined the one-third octave band level. levels and one for determining the 13 acoustic sound weighting, and then Furthermore, there are other attributes one-third octave sound levels. performing the necessary respective (i.e., sound analysis code window size, The second analysis code that has processing to arrive at both the overall time used for exponential averaging, been used by the agency is one sound pressure level and one-third and the precise details of the developed by the Volpe National octave band values. The respective implementation of the sound analysis Transportation Systems Center (the processing routines will be further code) that should also be considered for Volpe Code). This sound analysis code outlined in the following sections.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00084 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 ER14DE16.015 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90499

For evaluation purposes, the sound believed that it needed to understand cases for which the correct result is data recorded during some test runs the source of the differences before known. were analyzed using both the B&K Pulse either code could be used in a • The code must meet all of the filter code and the Volpe code. Some test runs compliance test. Therefore, NHTSA requirements for one-third octave band were also analyzed using two different undertook further research work after filters that are contained in the ANSI sets of user-specified parameters. publication of the NPRM to evaluate S1.11–2004 Class 1 standard. and resolve this issue. Analysts looking at the results from • The code can be made publically these runs noted that there were slightly The objective of this research was to select one sound analysis code and one available so all individuals and different overall sound pressure levels set of user-selectable parameters for use organizations know the exact methods, and one-third octave band levels for the in compliance testing of measured specified parameters, and filtering being exact same sound data depending upon vehicle sound data. Our criteria for used by NHTSA. the sound analysis code and the user- choosing an appropriate sound analysis Table 19 shows the standard settings selectable parameters used. While the code were: for the user definable parameters that differences that were seen were not • The code must generate correct can be set in each of the code packages large (less than 2.0 dBA), NHTSA results for mathematically-generated test that were evaluated.

TABLE 19—ANALYSIS CODE USER-SELECTABLE PARAMETERS

Acoustic test data analysis settings Parameter B&K Pulse Volpe Matlab

General Settings: Sampling Frequency ...... 65536 Hz ...... 65536 Hz. Processing Window ...... Test Scenario Dependent .. Test Scenario Dependent. Acoustic Weighting ...... A or Linear Weighting ...... A or Z Weighting. Overall Sound Pressure Level Settings:

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00085 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 ER14DE16.016 90500 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

TABLE 19—ANALYSIS CODE USER-SELECTABLE PARAMETERS—Continued

Acoustic test data analysis settings Parameter B&K Pulse Volpe Matlab

Frequency span ...... 25600 Hz ...... 24000 Hz. Overall Averaging ...... Linear ...... None. Averaging time ...... 0.05 ...... None. One-Third Octave Band Analysis Settings: Bandwidth (Fractional Octave) ...... 1⁄3—Base 10 Exact ...... 1⁄3—Base 10 Exact. Upper Nominal Center Frequency ...... 5000 Hz ...... 5000 Hz. Lower Nominal Center Frequency ...... 315 Hz ...... 315 Hz. Type of Octave Band Averaging ...... Exponential ...... Exponential. Type of Time Weighting ...... Fast ...... Fast. Averaging Time ...... 1⁄4 seconds ...... 1⁄4 seconds. Tau (Time Constant) ...... 1⁄8 seconds (Fast) ...... 1⁄8 seconds (Fast).

NHTSA began evaluating both codes which they span is, nominally, 70Hz– for this test run of 51.1394 dB). The by running the same vehicle sound data 22,300Hz. This range encompasses six second used thirteen pure tones at 60 file through both code packages, looking full one-third octave bands both above, dB (giving an Overall SPL for this test to see how consistent the codes were and six full one-third octave bands run of 71.1394 dB). This test case was relative to each other. The outcome was below, the 13 one-third octave bands of used to check the correctness of the that each code gave slightly different interest to NHTSA. This range was calculated amplitudes when no results, even while using consistent chosen to ensure a full profile of how frequency weighting (Z-weighting) was parameter settings. each code responds to known inputs applied to a complex sound data To systematically determine the was generated and understood. waveform. differences between the two packages, The following aspects of sound In general, in comparing the two both the B&K and the Volpe sound analysis code were checked using Test analysis codes using Test Case 2, analysis codes were checked to ensure Case 1 data files: NHTSA found very little or no that they provided known output results • The correctness of the calculated difference between the calculated for known input values. This was done amplitude, when no frequency amplitudes regardless of weighting type through the development of test cases weighting (Z-weighting) was applied, (A- or Z-weighting) for the individual that were processed using each of the for a pure tone at a frequency pure tones located at the center sound analysis codes. The test cases corresponding to the center of each of frequencies of each of the 13 one-third consist of simple pure tones which are the one-third octave bands of interest. octave bands. Each code set gave either computer-generated rather than taken • The correctness of the calculated 40 or 60 dB at each center frequency, as from actual sound recordings, and thus amplitude, when A-weighting was expected. The results from the two they have none of the complexity of applied, for a pure tone at a frequency analysis codes were also consistent actual acoustic measurements. The test corresponding to the center of each of when the overall SPL for the 13 center cases provide elemental inputs for the one-third octave bands of interest. which the correct outputs are known in • The correctness of the band-pass frequencies were combined, and both advance. The test cases were used to filters that split frequency-weighted the Volpe Matlab code and the B&K evaluate the accuracy of a given code’s sound pressure level data into 13 one- Pulse code produced the correct results analysis routine and to compare the third octave bands. NHTSA and of 51.1 dB and 71.1 dB for the 40 dB and outputs of the two different analysis commenters want these band-pass filters 60 dB inputs, respectively. methods. to meet all of the Type 1 filter However, in looking at the test results Test Case 1 was a series of pure tones. requirements for one-third octave band from Test Case 1, the two analysis codes The sound pressure of each tone as a filters that are contained in the standard were not consistent regarding their function of time is given by a constant- ‘‘ANSI S1.11–2004’’. The Test Case 1 band-pass filter function that splits amplitude, constant-frequency, single frequencies include all of the frequency-weighted sound pressure data sine wave. Multiple pure tones were frequencies listed in Table B1,’’ of ANSI into the 13 one-third octave bands. generated, each at a different constant- S1.11–2004 for the 13 one-third octave When comparing the output of each of frequency. For this research, two bands of interest to NHTSA. the 201 frequencies described in Test constant-amplitudes corresponding to For the second test case, Test Case 2, Case 1 to the requirements specified in 40 and 60 dB sounds were used. To be thirteen pure tones were superimposed ANSI S1.11–2004, NHTSA found that certain of capturing all important effects to form one sound-pressure signal. the B&K software tended to for each of the 13 one-third octave These thirteen pure tones were at the insufficiently attenuate the frequency bands of interest to NHTSA (which have frequencies corresponding to the center bands away from the nominal one-third nominal center frequencies ranging from of each of the one-third octave bands of octave band. An example of this is 315 Hz to 5,000 Hz), the pure tones for interest. No frequency weighting (i.e., Z- shown below in Figure 14 which plots Test Case 1, developed using MatlabTM, weighting) was applied. the minimum and maximum ANSI filter were generated at 201 individual Two test runs were made using Test requirements, the output of the B&K frequencies each corresponding to 1⁄8th Case 2. The first had a 40 dB pure tone Pulse code, and the output of the Volpe of a one-third octave band (1⁄24th of a centered at each of the one-third octave Matlab code, for the one-third octave full octave). The frequency range over bands of interest (giving an Overall SPL band centered at 1000 Hz.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00086 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90501

While some bands displayed better code appears to be in full agreement To address these requirements in the adherence to the ANSI S1.11 with ANSI S1.11–2004 specifications PSEA, the NPRM proposed a phase-in specifications, all of the 13 one-third and requirements. Also, the source code over three model years for new hybrid octave bands displayed similar results is not proprietary, and it can be made and electric vehicles produced for sale as the 1000 Hz band shown above for publically available. To resolve any in the U.S., and full compliance of all the B&K software. On the other hand, potential problems with post-processing new hybrid and electric vehicles by the Volpe Matlab code processed data code conflicts, the agency will make the September 1, 2018. The three-year fell well within the filter attenuation Matlab code to be used publically phase-in was based on a ‘30/60/90’ limits specified in ANSI S1.11–2004 available, either as part of the agency’s phase-in schedule. Given that the NPRM Class 1 across all bands. Complete compliance test procedure, or posted on assumed publication of a final rule in results for all the individual one-third the agency’s Web site. This approach calendar year 2014, the phase-in octave bands can be found in the will help the agency with its recent requirements proposed in the NPRM corresponding NHTSA research efforts to increase public were: 30 percent of each OEM’s HV and report.163 communications and transparency. In EV production in compliance by The results of our research indicate reference to the other parameters that September 1, 2015; 60 percent by that the two codes analyzed have the agency inquired about in the NPRM, September 1, 2016; 90 percent by different filter algorithms. This results measurement domains and type September 1, 2017; and 100 percent by in the two codes calculating slightly windows used for frequency analyses, September 1, 2018. The proposed phase- different one-third octave band levels. no direct comments were received so in schedule was intended to be The exact filtering algorithm used in the the agency has made decisions applicable to all manufacturers of HVs B&K code is unknown because the code according to what it believes are and EVs, except small volume and final is proprietary. The filtering algorithm technically correct. All the parameters stage manufacturers. The latter were used in the Volpe code is known and that will be used for post processing the allowed to postpone compliance until can be made public. Given the results of acoustic files will be specified in the the date on which other manufacturers our examination of the two post- publically available Matlab code. were required to have all their vehicles processing methods, NHTSA has L. Phase-In of Requirements brought into compliance, i.e., September 1, 2018. decided to use the Volpe Matlab code The PSEA directed NHTSA to for the agency’s future compliance establish a phase-in period to set forth The NPRM also included testing programs. As explained above, the dates by which production vehicles amendments to Part 585 Reporting one reason for this is that the Matlab must comply with the new FMVSS No. Requirements to allow for OVSC 141. The PSEA also stated that NHTSA verification of each manufacturer’s 163 Dr. W. Riley Garrott, Richard, L. Hoover, Eric must require full compliance ‘‘on or phase-in of pedestrian alert systems. Gerdus, and Sughosh J. Rao, ‘‘Selecting a Sound Analysis Code for Use With NHTSA Test Procedure after September 1st of the calendar year With the exception of two advocacy to Characterize Vehicle Sound’’ NHTSA Technical that begins three years after the date on groups, all commenters opposed the Report.DOT HS 812 284. which the final rule is issued.’’ phase-in requirements as proposed in

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00087 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 ER14DE16.017 90502 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

the NPRM. The NFB and NCSAB from the date that the final rule is issued assistance from carry-forward credits supported the phase-in schedule as before the phase-in requirements begin. (including early carry-forward credits). proposed. The NCSAB stated that the As mentioned above, Honda also They recommended full credits for EVs rule should be completed by January requested that a credit system be and HVs that comply with their 2014, according to the PSEA. Neither established as part of the phase-in. suggested sound specifications commenter suggested an alternative Toyota stated that it is committed to (assuming those were implemented in phase-in schedule. pedestrian safety, and as such, has the NHTSA final rule) and half-credit All other commenters requested that already equipped every hybrid and (i.e., two vehicles equal one credit) for NHTSA provide more lead time for electric vehicle it produced since model EVs and HVs that are equipped with compliance with the new safety year 2012 under the Toyota and Lexus pedestrian alert systems that do not standard. Some favored eliminating the brands (currently, there is no Scion HV meet the Alliance/Global suggested phase-in altogether and establishing a or EV) with a pedestrian alert sound requirements, but that nevertheless single date for full compliance for all meeting the existing Japanese comply with the spirit and purpose of production hybrid and electric vehicles. guidelines. However, Toyota noted that the PSEA. If NHTSA specifies a phase- Alternatively, commenters requested the proposed requirements of the NPRM in, Alliance/Global stated that carry- that NHTSA begin the phase-in at a later would require significant redesign of forward credits are necessary for their date, unless changes were made in the Toyota’s current production alert member companies to avoid needless final rule to adopt performance system, which will in turn require compliance expenditure on vehicle requirements much less stringent than substantial development and test time. models imminently due to be phased those in the NPRM. Honda and Therefore, Toyota recommended out of production. Alliance/Global requested that NHTSA elimination of the phase-in Alliance/Global commented that allow for carry-forward credits which requirements and suggested that small manufacturers should not be would give a manufacturer credit for NHTSA consolidate the schedule by required to comply until the end of the meeting one of the phase-in stages prior simply requiring full compliance for all phase-in period. Because no current EV to the deadline for that stage, and the HVs and EVs by September 1, 2018 or HV pedestrian alert sound voluntarily manufacturer could use that credit if it (assuming the final rule is published in implemented by vehicle manufacturers did not fully meet a deadline of a later calendar year 2014 or earlier). meets NHTSA’s proposed requirements, stage. Alliance/Global commented that it if the agency proceeds to a final rule A heavy vehicle OEM commented would not be possible for manufacturers that is substantially similar to the that the proposed Part 585 phase-in to meet a phase-in beginning September NPRM, Alliance/Global would prefer reporting should not apply to a 1, 2014. If the requirements of the final that NHTSA does not specify a phase- manufacturer that achieves 100 percent rule were to be substantially similar to in, and instead allows all manufacturers early compliance, and also stated that the NPRM, they recommend foregoing the maximum amount of time to comply paragraph S9.5 of the NPRM, regarding the phase-in and going directly to full with the requirements of the new safety phase-in for multi-stage vehicles, is implementation on September 1, 2018. standard. unnecessary because only a final stage However, if the final rule instead were Finally, Alliance/Global stated that manufacturer would be responsible for to approximate the Alliance/Global phase-in language needs to clarify that meeting the phase-in requirements. recommendations, then a phase-in requirements pertain only to vehicles Porsche, a light vehicle manufacturer period is feasible beginning with described in the Applicability section of that produces only one hybrid model, vehicles built on or after September 1, the regulation and not to every type of provided proprietary production 2015, and ending with vehicles built on vehicle that a full-line manufacturer estimates through September 2018 or after September 1, 2018 (those dates produces. indicating that they would not meet the would need to be adjusted should the The MIC commented that, if NHTSA 90 percent level by the third year of the final rule be significantly delayed does decide to establish minimum proposed phase-in. beyond the original January 2014 sound requirements for motorcycles, it The EDTA commented that, due to deadline). should extend the phase-in exemption the complexity of the proposal, as well Alliance/Global also commented that for small manufacturers, including as the technology needed to implement currently there are no EVs or HVs motorcycle manufacturers, indefinitely. it, substantial lead time will be needed produced by their member companies Nissan requested that the phase-in to design, develop, test and certify new that are capable of meeting the begin at least two years following the alert systems. EDTA stated that it joined requirements proposed by NHTSA. issuance of a final rule. Nissan also with Alliance/Global in recommending They stated that several strategies had requested that NHTSA provide for the that, if the final rule is substantially the been considered, including use of advanced credits for vehicles that same as the proposal, the phase-in reprogramming an existing alert sound comply before the final date for specified in the final rule should be control module. They also stated they compliance. limited to a single 100-percent had interviewed suppliers who Denso commented that vehicle compliance date that is set in currently manufacture alert systems in manufacturers, as well as equipment accordance with the PSEA (i.e., an effort to explore all possible suppliers, need three years of lead time September 1st of the calendar year that solutions for meeting the NPRM. They before beginning phase-in of complying begins three years after the date on concluded that considerably more time vehicles. which the final rule is issued). would be needed than a September 1, Navistar questioned how the Honda commented that, if the final 2014 start of phase-in would allow to proposed phase-in meshes with Parts rule must be complied with starting in package/repackage components, 567 and 568 regarding certification of September 2015, it would need more develop new systems, source the multistage vehicles. time to meet all the requirements components, and certify the new OICA commented that the Phase-in proposed in the NPRM (modification of systems. should include only those vehicles to speakers, control unit, vehicle structural However, Alliance/Global commented which the performance requirements are modifications, etc.). Therefore, Honda that such a phase-in schedule as the one meant to apply, i.e., certain hybrid and requested at least two or more years they suggested still would need electric vehicles.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00088 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90503

Agency Response to Comments the initial date on which the phase-in conservation and theft prevention of Given that this final rule is being begins. wheeled vehicles, equipment and parts. By establishing GTRs under the 1998 published in calendar year 2016 and, IV. International Harmonization and Agreement, the Contracting Parties seek furthermore, given that the PSEA Stakeholder Consultation to pursue harmonization in motor stipulates full compliance on and after NHTSA is required by the PSEA to September 1st of the calendar year that vehicle regulations not only at the consult with the following organizations national and regional levels, but begins three years after the date on as part of this rulemaking: The which the final rule is issued, NHTSA worldwide as well. Environmental Protection Agency (EPA) As a general matter, governments, is requiring compliance for 100 percent to assure that any alert sound required vehicle manufacturers, and ultimately, of HVs and EVs produced for sale in the by the rulemaking is consistent with consumers, both here and abroad, can U.S. by all manufacturers by no later noise regulations issued by that agency; expect to achieve cost savings through than September 1, 2019. This consumer groups representing visually- the formal harmonization of differing compliance date is set forth in the impaired individuals; automobile sets of standards when the contracting Applicability section of the regulatory manufacturers and trade associations parties to the 1998 Global Agreement text of this final rule. representing them; technical implement new GTRs. Formal In addition, after review of the standardization organizations harmonization also improves safety by comments submitted, NHTSA is responsible for measurement methods assisting us in identifying and adopting adopting a one-year, 50 percent phase- such as the Society of Automotive best safety practices from around the in. Under this phase-in, 50 percent of Engineers, the International world and reducing diverging and the total production volume of each Organization for Standardization (ISO), unwarranted regulatory requirements. manufacturer’s hybrid and electric and the UNECE World Forum for The harmonization process also allows vehicles to which the safety standard Harmonization of Vehicle Regulations manufacturers to focus their compliance applies, and which are produced by the (WP.29). and safety resources on regulatory manufacturer for sale in the United The agency has established three requirements whose differences States., must comply by no later than dockets to enhance and facilitate government experts have worked to September 1, 2018. cooperation with outside entities converge as narrowly as possible. This phase-in does not apply to multi- including international organizations. Compliance with a single standard will stage and small volume manufacturers. The first docket (No. NHTSA–2008– enhance design flexibility and allow Those manufacturers would have until 0108) was created after the 2008 public manufacturers to design vehicles that September 1, 2019, to comply. This meeting was held; it contains a copy of better meet safety standards, resulting in should not have any significant effect on the notice of public meeting in the safer vehicles. Further, we support the traffic safety because of the relatively Federal Register, a transcript of the harmonization process because it allows small number of vehicles they produce. meeting, presentations prepared for the the agency to leverage scarce resources Because the phase-in period will have meeting and comment submissions. It by consulting with other governing a duration of only one year, carry- also includes NHTSA’s research plan, bodies and international experts to forward credits would not be of any our ‘‘Notice of Intent to Prepare an share data and knowledge in developing benefit. Therefore, NHTSA is not Environmental Assessment for the modernized testing and performance making any provisions in this rule for Pedestrian Safety Enhancement Act of standards that enhance safety. carry-forward credits. 2010’’ published on July 12th 2011 in Under the 1998 Agreement, countries The agency’s decision on the phase-in the Federal Register, and the agency’s voting in favor of establishing a GTR, issues is a compromise that responds to Phase 1 and 2 research reports. (The agree in principle to begin their internal comments about reducing the phase-in Notice of Intent [NOI] and the agency’s implementation processes for adopting or eliminating it altogether. The one research are discussed more fully in the provisions of the GTR, e.g., in the year phase-in addresses the mandatory other parts of this document.) The U.S., to issue an NPRM or Advanced PSEA requirements and ensures that second docket (No. NHTSA–2011–0100) NPRM, within one year. The ultimate any delay in getting complying vehicles was created to collect comments on the decision whether or not to adopt the to market will be minimized. At the NOI; it also includes a copy of that GTR is at each contracting party’s same time, it responds to commenters’ notice. The third docket (No. NHTSA– discretion, however, based on its requests for additional lead time to 2011–0148) was created in September determination that the GTR meets or comply and to their suggestions that the 2011 to include materials related to the does not meet its safety needs. The NPRM phase-in should be consolidated rulemaking process (‘‘The Pedestrian UNECE World Forum for Harmonization and simplified. A one-year phase-in Safety Enhancement Act of 2010,’’ of Vehicle Regulations (WP.29) provides additional flexibility for Phase 1 and 2 research reports, administers the 1998 Agreement. manufacturers as to when they bring statistical reports, meeting In 2009, the Ministry of Land, their model lines into compliance. presentations, etc.), and outside Infrastructure, Transport and Tourism Furthermore, NHTSA has reviewed comments. (MLIT) of Japan assembled a committee current model lines of vehicle On June 25, 1998, the United States to study the issue of the quietness of manufacturers using OVSC annual signed the 1998 Global Agreement, HVs. The committee concluded that an compliance information and has which entered into force on August 25, Approaching Vehicle Audible System determined that several of the OEMs 2000. This agreement was negotiated (AVAS) was a realistic alternative to that produce HVs and/or EVs have only under the auspices of the United allow pedestrians who are blind or one or two such models among their Nations Economic Commission for visually-impaired to detect quiet vehicle lines. This is one factor that we Europe (UNECE) under the leadership of vehicles. In 2010, MLIT announced have considered in choosing an the U.S., the European Community (EC) guidelines for AVAS based on the appropriate phase-in period. These and Japan. The 1998 Agreement recommendations of the study manufacturers will benefit from a provides for the establishment of Global committee. Although several vehicles shortened phase-in schedule that Technical Regulations (GTRs) regarding were considered in the initial scope, provides additional lead time prior to the safety, emissions, energy MLIT concluded that AVAS should be

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00089 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90504 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

installed only on HVs that can run on people who are visually-impaired for Emitted by Road Vehicles. The agency electric motors, EVs and fuel-cell the purpose of consulting with these had been sending a liaison to VSP vehicles. In terms of the activation groups on topics related to this meetings starting in 2008. SAE is the condition, the MLIT recommended that rulemaking. Participating in these U.S. technical advisory group to the AVAS automatically generate sound at meetings were representatives from the International Organization for least in a speed range from the start of Alliance of Automotive Manufacturers, Standardization (ISO), and they both a vehicle until reaching 20 km/h (12 the Global Automakers (formerly have cooperated in the development of mph) and when moving in reverse. The Association of International Automobile the industry safety standard. The ISO AVAS would not be required when a Manufacturers (AIAM)), American document (ISO/NP 16254, Measurement vehicle is stopped. The system may Council of the Blind, The American of Minimum Noise Emitted by Road include a switch to temporarily halt the Foundation of the Blind (AFB), the Vehicles) and the SAE document are operation of the AVAS. The reason for National Federation of the Blind (NFB), technically identical. The agency used including this switch is because the The International Organization for SAE J2889–1 and ISO 16254 as committee believes that the system is Standardizations (ISO), The Society of references in the NHTSA test procedure not needed on expressways where there Automotive Engineers (SAE), the development. Other international are no pedestrians and to reduce other International Organization of Motor organizations, such as the International issues such as drivers deliberately Vehicles Manufacturers (OICA), The Organization of Motor Vehicle increasing vehicle speed in order to stop Environmental Protection Agency (EPA) Manufacturers (OICA) and Japan the AVAS. and Japan Automobile Manufacturers Automobile Manufacturers Association In its March 2011 session, WP.29 Association (JAMA). (JAMA) have provided NHTSA with determined that vehicles propelled in Representatives of the EPA have also research findings and also have whole or in part by electric means, been included in our activities with attended various quiet vehicle meetings. present a danger to pedestrians and outside organizations. They have been In the NPRM, the agency concluded consequently adopted Guidelines kept updated on our research activities that the voluntary guidelines adopted by covering alert sounds for electric and and have actively participated in our the Japanese government, and hybrid vehicles that are closely based on outreach efforts. NHTSA has also kept subsequently by the UNECE WP.29 the Japanese Government’s guidelines. up to date on EPA activities on the Committee, did not have the level of The Guidelines were published as an international front through the activities detail necessary for NHTSA to establish annex to the UNECE Consolidated of the UNECE Working Party of Noise objective minimum performance Resolution on the Construction of (GRB). requirements for creation of an FMVSS. Vehicles (R.E.3). Considering the The American Foundation of the We did not believe that the agency international interest and work in this Blind, the American Council of the would be able to tell if a sound fell new area of safety, the U.S. decided to Blind and the National Federation of the within one of the exclusions by means lead the efforts on the new GTR, with Blind have provided NHTSA with of an objective measurement, nor would Japan as co-sponsor, and develop invaluable information about visually- we be able to adequately ensure that harmonized pedestrian alert sound impaired pedestrian safety needs since sound levels would be detectable by requirements for electric and hybrid- the 2008 Public Meeting was held. pedestrians or provide manufacturers electric vehicles under the 1998 Global The Alliance of Automobile with a set of requirements that they Agreement. Development of the GTR for Manufacturers and Global Automakers would be able to meet. The NPRM noted pedestrian alert sound has been have met separately with the agency to that the WP.29 QRTV work was assigned to the Group of Experts on discuss our research findings and their scheduled to be completed in 2014, and Noise (GRB), the group most ideas regarding this rulemaking. a draft GTR adopted in November 2014. experienced with vehicle sound Members of both organizations have OICA, EU, Chrysler, EDTA, VW, and emissions. GRB is in the process of also met separately with the agency to Alliance/Global all suggested delaying assessing the safety, environmental and discuss their own research findings and the development of a U.S. regulation on technological concerns to develop a ideas for a potential regulatory approach minimum noise levels until WP.29 has GTR that leverages expertise and to address the safety issues of interest to had sufficient time to develop a globally research from around the world and the agency. harmonized set of regulations via the feedback from consumer groups. The Automotive manufacturers that GTR process. They stated that U.S. is the co-chair (with Japan) of the produce EVs for the U.S. market have establishment of separate requirements informal working group on Quiet Road developed various pedestrian alert that may or may not be harmonized Transport Vehicles (QRTV) assigned to sounds, recognizing that these vehicles, with the recommendations under develop the GTR and, therefore, will when operating at low speeds, may pose negotiation through WP.29 would harm guide the informal working group’s an elevated safety risk to pedestrians. development of electric drive vehicles development of the GTR. GRB will meet They have made vehicles with sound globally and constrain the growth of the regularly and report to WP.29 until the alert systems available for lease by market as a whole. establishment of the new GTR. NHTSA NHTSA for research purposes. This OICA, EU, VW, and Alliance/Global has been participating in the QRTV’s information has been helpful in the commented that the PSEA statute does meetings since its foundation and has agency decision making process. not provide enough time for WP.29 to kept the group informed about ongoing The Society of Automotive Engineers address all remaining technical issues in agency research activities as well as the (SAE) established the Vehicle Sound for development of a globally harmonized results from completed research studies. Pedestrians (VSP) subcommittee in standard that the U.S. could then adopt. At the time the NPRM was issued, the November 2007 with the purpose of EU commented that if the agency is QRTV informal group had held five developing a recommended practice to unable to delay publication of a final sessions to discuss development of a measure sounds emitted by ICE vehicles rule that would harmonize with the GTR on quiet vehicles. and alert sounds for use on EVs and international community, it should at a NHTSA has also hosted roundtable HVs. Their efforts resulted in minimum ensure that any U.S. meetings with industry, technical recommended practice SAE J2889–1, regulations are consistent with the organizations and groups representing Measurement of Minimum Noise recommendations of the WP.29 Informal

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00090 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90505

Working Group on Quiet Road sound levels, or, at a minimum, work V. Analysis of Costs, Benefits, and Transport Vehicles. closely with the QRTV in development Environmental Effects The EU questioned to what extent of requirements that could be A. Benefits NHTSA had taken into consideration recognized globally. The agency, the conclusions and results of the through its leadership role in the QRTV As stated above in the discussion of QRTV–IWG. They believed a delay in informal group, continues to work with the statistical analysis of safety need done for this rulemaking (see Section the NPRM process and the finalization the international community in II.B), the data from 16 states cannot be of the new FMVSS until the new GTR development of criteria that are has been drafted would contribute used to directly estimate the national technically sound and objective. We towards a common approach and an problem size. Also, an analysis of note that the WP.29 QRTV work has overall consensus at the international pedestrian fatalities rather than injuries level with respect to EVs and HEVs. been extended until late 2015, at the is not appropriate for this rulemaking. VW and Alliance/Global commented earliest, with expected eventual The target population analysis will that if NHTSA is unable to delay the adoption of a GTR on minimum noise therefore focus on injuries only. enactment based on statutes within the requirements for electrically driven The PSEA directs NHTSA to establish PSEA, NHTSA should inform the vehicles. Adoption of the GTR is only minimum sound requirements for EVs United States Congress that additional the beginning of the process of and HVs as a means of addressing the time to complete this rulemaking is regulating minimum noise levels by increased rate of pedestrian crashes for required in order to allow for signatories of the 1998 UN agreement. these vehicles. In calculating the completion of the GTR so that a After a GTR on minimum noise benefits of this rulemaking we have harmonized regulation can be achieved. requirements is adopted, NHTSA would assumed that adding sound to EVs and Alliance/Global commented that in still need to issue an NPRM or an HVs will bring the pedestrian crash accordance with the QRTV Terms of SNPRM (Supplemental Notice of rates for these vehicles in line with the Reference, the development of the GTR Proposed Rulemaking) to begin the pedestrian crash rates for ICE vehicles should be concluded in the fall of 2014, process of adopting the GTR as an because the minimum sound requirements in the proposed rule with status reports provided along the FMVSS. This could result in several would ensure that EVs and HVs are at way so that the public can monitor the additional years of delay before an status of the activity. Alliance/Global least as detectable to pedestrians as ICE FMVSS mandating sound for EVs and vehicles. This approach assumes that explained that the benefits of having HVs could be issued. We do not believe consensus on a global technical EVs and HVs have higher pedestrian that a delay of this length is justified regulation are enormous and any crash rates than ICE vehicles because of from a safety perspective. We believe potential downside related to allowing the differences in sound levels an accelerated GTR process to conclude the agency’s approach in development produced by these vehicles. Therefore, prior to finalizing the NHTSA regulation of this final rule to be consistent with the target population for this rulemaking will be negligible given that a majority both the mission and safety goals of the is the number of crashes that would be of current production EVs and HVs are agency and with the PSEA and Safety avoided if the crash rates for hybrid and already voluntarily equipped with Act. electric vehicles were the same as the audible pedestrian alert systems. We agree with WBU and MB that crash rate for ICE vehicles. EU, VW, Chrysler, and Alliance/ development of U.S. regulations for No quantifiable benefits are estimated for EVs because we assume that EV Global all supported using the GTR minimum noise levels might aid WP.29 manufacturers would have added alert process to finalize any remaining in addressing some of the technical technical issues towards a globally sounds to their cars in the absence of issues that hinder development of a this proposed rule and the PSEA. harmonized standard. global regulation that is both measurable WBU and MB supported using the NHTSA was not able to directly and enforceable. We note that the NPRM as a basis for development of the measure the safety differences between leadership role of the U.S. delegation in WP.29 GTR. hybrids with and without sound. development of a global regulation for Although there are now some hybrids in Agency Response to Comments minimum noise levels is consistent with the market that produce sounds to alert The NPRM stated that the the comments regarding using the GTR pedestrians and pedalcyclists, the recommendations of the QRTV informal process to refine a harmonized agency is unable to directly measure the working group do not include objective regulation. In that light, we believe that effectiveness rate of sound by using data criteria with which the agency could development of a U.S. regulation would from these new hybrid vehicles because ensure vehicles comply with an FMVSS. aid WP.29 in drafting a global regulatory there is not sufficient crash data on new The agency maintains that this is still framework that is both measureable and model hybrid vehicles with sound to be the case. Further, as discussed above, enforceable. able to make a statistically significant comparison to hybrids without sound. the agency has determined that a The agency has also continued to The agency’s data base for low speed crossover speed of 30 km/h is necessary actively monitor the work that has been because our conclusion from the data injuries is a sample, and data on crashes done internationally by SAE and ISO. we have acquired to date from all involving hybrid vehicles that emit The SAE recently issued an updated sources (i.e., from commenters and from sound is limited. Furthermore, the data version of J2889–1 dated December our own vehicle evaluations) is that set used to analyze differences in crash some hybrid and electric vehicles 2014. The ISO recently submitted the rates for this rulemaking consists of continue to need sound enhancement at latest draft of ISO 16254 to the agency’s crash data from 16 states. At this time, 164 speeds above 20 km/h in order to ensure docket. The agency has taken into only half of the states have submitted that they are adequately detectable. consideration these documents to the data for the 2012 or later calendar years. Most of the commenters extent possible for the development of Since we believe that most hybrid recommended that the agency wait until this final rule. vehicles have been equipped with some the WP.29 World Forum can complete type of alert sound only since 2012, any development of a GTR for minimum 164 NHTSA–2011–0148–0334. effect that voluntary adoption would

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00091 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90506 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

have on pedestrian crash rates would analyzing crash risk, it does raise a relevant not-in-traffic crashes. In order not be captured by this data set. In problem in this case because NHTSA to combine the GES and NiTS data in a addition, none of the recently was not able to control for VMT. meaningful way, it was assumed that introduced hybrids with sounds were NHTSA assumes that any difference in the ratio of GES to NiTS will be constant designed to meet all of the requirements these ratios is attributable to the lack of for all years 2006 to 2012. in this rule. Therefore, any change in sound in HVs. However, it is possible Because both the GES and NiTS crash rate between original quiet HVs that there are other explanations for databases rely on police-reported and these voluntarily-equipped HVs differences. For example, there may be crashes, these databases do not would not necessarily be indicative of reasons other than sound for why HVs accurately reflect all vehicle crashes the full safety benefits of compliant have higher numbers of pedestrian and involving pedestrians because many of sounds. pedalcyclist accidents. Or there may be these crashes are not reported to the NHTSA has also been unable to reasons why ICEs have higher numbers police. The agency estimates that the directly measure the pedestrian and of other types of accidents.165 This number of unreported crashes for pedalcyclist crash rates per mile could result in a lower ratio for ICEs pedestrians is equal to 100.8 percent of travelled for HVs and EVs to the rates even if the two types of vehicles had the reported crashes. That is to say, for for ICEs because the agency does not similar pedestrian and pedalcyclist every 100 police-reported pedestrian have data on VMT for HVs and EVs. To crash rates. calculate the difference in crash rates The first step in NHTSA’s analysis crashes, there exist 100.8 additional between HVs and ICEs NHTSA was to use injury estimates from the unreported pedestrian crashes. computes the ratio of the number of 2006–2012 National Automotive Table 20 shows the reported and pedestrian and pedalcyclist crashes Sampling System—General Estimates unreported crashes by injury severity. involving HVs to the number of other System (NASS–GES) and both 2007 and Only injury counts will be examined for types of accidents involving HVs and 2008–2011 Not in Traffic Surveillance the purpose of benefits calculations and, compares it to a similar ratio for ICEs. (NiTS) database to provide an average as such, fatalities and uninjured (MAIS While this is a standard technique in estimate for combined in-traffic and 0) counts are not included.

TABLE 20—ESTIMATED ANNUAL QUIET CARS TARGET POPULATION INJURIES REPORTED (GES 2006–2012, NITS 2007, 2008–2011) AND UNREPORTED PEDESTRIANS AND PEDALCYCLISTS, BY VEHICLE

MAIS level 1 2 3 4 5 TOTAL 1–5

Reported (GES+NiTS) and Unreported Injured Pedestrians

Passenger Car (PC) ...... 69,453 11,093 2,249 529 214 83,538 Light Trucks & Vans (LTV) ...... 47,604 7,852 1,629 387 156 57,626

Total Light Vehicles (PC+LTV) ...... 117,056 18,945 3,877 916 370 141,164

Reported (GES+NiTS) and Unreported Injured Pedalcyclists

MAIS level ...... 1 2 3 4 5 TOTAL 1–5 Passenger Car (PC) ...... 42,943 6,148 1,082 239 84 50,495 Light Trucks & Vans (LTV) ...... 26,932 3,957 715 160 56 31,820

Total Light Vehicles (PC+LTV) ...... 69,875 10,105 1,796 400 140 82,315

The estimates in Table 20 are based The estimated injuries in Table 21 estimated lifetime injuries from the on the current make-up of the fleet for and Table 22 are created by combining MY2020 fleet (141,567). Given the all propulsion types. Next, we make the the estimated percentage of annual sales effectiveness assumption of 97 percent, assumption that because the hybrid and of hybrid and electric vehicles for the rulemaking addresses 850 of those electric vehicles pose a higher risk of MY2020 from Table 23 with the odds 877 injuries. When considering pedestrian collisions, each hybrid and ratio of 1.18, representing the increased pedalcyclists injured by MY2020 electric vehicle is producing more risk of an HV being involved in a vehicles, the rulemaking is applied to injuries per year than their ICE pedestrian crash, and the odds ratio of the 1,514 injury difference between that counterparts. Thus, while the 2006– 1.51, representing the increased risk of theoretical fleet (81,455 injuries) and the 2012 time period resulted in 141,164 an HV being involved in a pedalcyclist estimated lifetime injuries from the pedestrian injuries annually, this injury crash.166 Thus, when considering MY2020 fleet (83,015). Given our count is the result of the mixed hybrid/ pedestrians injured by MY2020 vehicles assumption that the pedestrian and electric/ICE fleet during that period. and assuming these pedestrian crashes pedalcyclists crash rates for LSVs Based on the odds ratios from our crash occurred because the pedestrians failed without sound is similar to that for analysis, we can calculate what size of to detect these vehicles by hearing, the other types of light vehicles without theoretical ICE-only fleet would have rulemaking applies to the 877 injury sound, the rule would also reduce been needed to generate as many difference between that theoretical ICE- pedestrian injuries by 4 over the lifetime injuries during that same time period. only fleet (140,663 injuries) and the of the MY2020 fleet of LSVs and

165 For example, HLDI compared overall rates of overall crash rate for hybrids. Highway Loss Data 166 Wu, J. (2015). Updated Analysis of Pedestrian injury for hybrid vehicles and their ICE non-hybrid Institute. ‘‘Injury Odds and Vehicle Weight and Pedalcyclist Crashes by Hybrid Vehicles with twins and found that crash rates are lower for Comparison of Hybrids and Conventional Larger Samples and Multiple Risk Factors. hybrids. HLDI concluded that the heavier weight of Counterparts.’’ HLDI Bulletin 28(10). Arlington, VA, Washington, DC: National Highway Traffic Safety hybrids was an important factor in this lower 2011. Administration.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00092 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90507

pedalcyclist injuries by 7 over the lifetime of the MY2020 fleet of LSVs.

TABLE 21—ENHANCED INJURY RATE (EIR) FOR PEDESTRIANS FOR 2020 MODEL YEAR 167

Injuries Mild Strong EVs + assuming Injuries hybrids hybrids fuel cell ICEs Total 100% assuming Injury Benefits (%) (%) (%) (%) (%) ICE predicted difference fleet fleet

Passenger Car...... 6.94 6.86 0.21 87.02 101.03 83,101 83,953 853 827 Light Trucks & Vans ...... 7.97 0.59 0.08 91.45 100.09 57,563 57,614 51 50

Total ...... 140,663 141,567 904 877

TABLE 22—ENHANCED INJURY RATE (EIR) FOR PEDALCYCLISTS FOR 2020 MODEL YEAR 168

Injuries Mild Strong EVs + assuming Injuries hybrids hybrids fuel cell ICEs Total 100% assuming Injury Benefits (%) (%) (%) (%) (%) ICE predicted difference fleet fleet

Passenger Car ...... 6.94 8.80 0.21 87.02 102.97 49,737 51,215 1,479 1,434 Light Trucks & Vans ...... 7.97 0.76 0.08 91.45 100.26 31,719 31,800 81 79

Total ...... 81,455 83,015 1,560 1,514

As discussed in the Final Regulatory which would mean that adding sound to to the ability to travel independently Impact Analysis (FRIA), MAIS injury hybrid and electric vehicles would not and will, therefore, also lead to levels are converted to dollar amounts. reduce pedestrian and pedalcyclist increased employment and the ability to The benefits across passenger cars, crash rates for hybrids to that of ICE live independently. LTVs, and LSVs of reducing 2,401 vehicles. Based on research conducted pedestrian and pedalcyclist injuries, or by NHTSA’s VOPLE Center,169 NHTSA B. Costs 32 undiscounted equivalent lives saved also assumes the sound added to hybrid Based on Ward’s Automotive (19.80 equivalent lives at the 7-percent and electric vehicles will be 97-percent Yearbook 2014,170 there were 597,035 discount rate and 25.64 at the 3-percent effective in providing warning to discount rate), is estimated to be $320 pedestrians as the sound produced by a hybrid engine installations in light million at the 3-percent discount rate vehicle’s ICE. vehicles (96 percent were in passenger and $247.5 million at the 7-percent In addition to the benefits in injury cars and 4 percent were in light trucks) discount rate. reduction due to this rule, there is also sold in MY2013, which accounts for 3.5 The agency calculated the benefits of the benefit to blind and visually percent of the total 17.2 million this rule by calculating the ‘‘injury impaired individuals of continued MY2013 light vehicles. There were a differences’’ between ICE vehicles and independent mobility. The increase in smaller number of MY2013 electric HVs. The ‘‘injury differences’’ assume navigational ability resulting from this vehicles: 17,480 passenger cars and that the difference between crash rates rule is hard to quantify and thus this 1,046 LTVs, representing 0.1 percent of for ICEs and non-ICEs is explained benefit is mentioned but not assigned a the overall sales. The Annual Energy wholly by the difference in sounds specific productivity or quality of life Outlook (AEO) for 2014 provides future produced by these two vehicle types of monetization. By requiring alert sounds estimates of the fleet broken down into vehicles and the failure of pedestrians on hybrid and electric vehicles, blind hybrid and electric vehicles.171 The and pedalcyclists to detect these pedestrians will be able to navigate number of vehicles that the agency vehicles by hearing. It is possible that roads as safely and effectively as if the projects will be required to meet the there are other factors responsible for fleet were entirely ICE vehicles. The standard is shown in Table 23. some of the difference in crash rates, benefit of independent navigation leads

TABLE 23—ESTIMATED/PREDICTED HYBRID AND ELECTRIC VEHICLE SALES PROPOSED TO BE REQUIRED TO PROVIDE AN ALERT SOUND

Predicted Estimated 2020 sales 2013 sales source: AEO & source: Ward’s NHTSA

Low-Speed Vehicles ...... 1,500 2,500

167 Table values may not add up to the correct 170 Ward’s Automotive Yearbook CD. Path: \2014 via the vehicle’s electric motor. The agency did not total due to rounding. YB CDROM\5. North America\c. U.S. Auto included ‘‘micro hybrids’’ whose ICE is always 168 Table values may not add up to the correct Industry\3. Engines\Engines by Type.xls running when the vehicle is motion when value due to rounding. 171 In calculating the costs of this rule the agency calculating the costs of this rule. 169 See ‘‘Robustness’’ discussion in Section III.E. only included those vehicles that can operate solely

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00093 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90508 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

TABLE 23—ESTIMATED/PREDICTED HYBRID AND ELECTRIC VEHICLE SALES PROPOSED TO BE REQUIRED TO PROVIDE AN ALERT SOUND—Continued

Predicted Estimated 2020 sales 2013 sales source: AEO & source: Ward’s NHTSA

Light Vehicles Electric ...... 18,526 15,020 Light Vehicles Fuel Cells ...... 0 5,606 Light Vehicles Hybrid ...... 597,035 * 506,701

Light Vehicles subtotal ...... 594,061 527,327

Total Sales ...... 602,061 561,327 * Note—This estimate of vehicle sales includes micro-hybrids which the rule does not apply to. This overestimation of hybrid vehicle sales is addressed in the MY2020 column, where propulsion source is provided by AEO.

The Nissan Leaf and other fully hybrids that fall under the rule. Thus, mounting hardware and wiring in order electric vehicles come equipped with an the number of light vehicles that have to power the speaker(s) and receive alert sound system. Based on what to add (or upgrade) an alert sound signal inputs, and a digital signal manufacturers have voluntarily system for costing purposes for MY2020 processor that receives information from provided in their fully electric vehicles, is 561,327 vehicles. the vehicle regarding vehicle operating the agency assumes that fully electric Based on informal discussions with status (to produce sounds dependent vehicles and hydrogen fuel-cell vehicles suppliers and industry experts, in upon vehicle speed, for example.) We will provide an alert sound system addition to confidential documents assume there will be no other structural provided to the agency, we estimate that voluntarily and, therefore, for costing changes or installation costs associated purposes we assumed a small upgrade the total consumer cost for a system that with complying with the rule’s cost in order to bring these existing produces sounds meeting the requirements. We believe the same systems up to compliance. In addition, requirement of this rule is $125.34 per we assume that some hybrid light hybrid light vehicle. In cases where a system can be used for both LSVs and vehicles, particularly those sound system already exists on a light light vehicles. We estimate that the manufactured by Toyota, come vehicle (hybrid vehicles voluntarily added weight of the system would equipped with some form of speaker equipped, electric vehicles, and fuel cell increase fuel costs for light vehicles by system, similar to the ones expected to vehicles), we assume a cost of $50.49. about $4 to $5 over the lifetime of the be found on electric vehicles. This estimate includes the cost of a vehicle. Average vehicle costs reflect the Furthermore, www.energy.gov data dynamic speaker system that is different installation costs determined indicates that these partially-equipped packaged for protection from the by propulsion source and vehicle make light vehicles make up about 67% of the elements and that is attached with as described above.

TABLE 24—COST SUMMARY (IN $M, 2013 ECONOMICS)

3% Discount 7% Discount rate rate ($) ($)

Per vehicle costs: Passenger Cars, Per Vehicle * ...... $79.06 $78.16 Light Trucks, Per Vehicle * ...... 77.27 76.17 Low Speed Vehicles (LSVs), Per Vehicle * ...... 78.91 77.99 Total Cost by Vehicle Type:. Passenger Cars ...... 38.2M 37.8M Light Trucks ...... 3.6M 3.5M Light Vehicles, PCs + LTVs Subtotal ...... 41.8M 41.3M Low Speed Vehicles (LSVs) ...... 0.3M 0.3M Total (PC + LTV + LSV) ...... 42.1M 41.6M

In addition to the quantifiable costs costs. Sound insulation also can over current noise levels. Given the low discussed above, there may be a cost of counteract interior noise, and a increase in overall noise caused by this adding sound to quiet vehicles to sensitivity analysis for sound insulation rule, we expect that any costs that may owners who value quietness of vehicle cost is provided in the accompanying exist due to added sound will be operation and to society at large. FRIA. minimal. NHTSA has not found any NHTSA is not aware of a method to As explained further in the way to value the increase in noise to quantify the value of quietness for a Environmental Assessment (EA), we society at large, and, thus it is a non- driver’s own vehicle. Some sound from expect that the increase in noise from quantified cost. these systems may intrude into the the alert sound will be no louder than C. Comparison of Costs and Benefits passenger compartment. The use of that from an average ICE vehicle and multiple speakers with directional that aggregate sound from these vehicles Comparison of costs and benefits characteristics might mitigate these will not create an appreciable increase expected due to this rule provides a

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00094 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90509

savings of $0.4 million per equivalent MY2020) and therefore this rulemaking impact of the rule on light vehicles and life saved to a cost of $0.04 million per is assumed to be cost beneficial. Since LSVs is a positive one, even with the equivalent life saved across the 3- the lifetime monetized benefits estimated $46 million required to install percent and 7-percent discount levels. (VSL+Economic) of MY2020 light speakers 172 and $3 million in lifetime This falls under NHTSA’s value of a vehicles (and LSVs) is expected to be fuel costs. statistical life of $10.8 million, (for between $197.6M and $244.9M, the net

TABLE 25—DISCOUNTED BENEFITS (PC+LTV) MY2020, 2013$

Total PED + CYC Total Mone- tized Benefits Total ELS

3% discount ...... (PC) ...... $301,146,801 24.25 (LTV) ...... 17,381,812 1.39

Total ...... 318,528,614 25.64

7% discount ...... (PC) ...... 233,031,924 18.74 (LTV) ...... 13,258,335 1.06

Total ...... 246,290,259 19.80

TABLE 26—TOTAL COSTS (PC+LTV) 2013$

Total cost/veh Total costs

3% discount ...... (PC) ...... $79.06 $38,223,782 (LTV) ...... 77.27 3,587,400

Total ...... 78.91 41,811,182

7% discount: (PC) ...... 78.16 37,788,667 (LTV) ...... 76.17 3,536,329

Total ...... 77.99 41,324,996

TABLE 27—NET IMPACTS (PC+LTV) 2013$

Net costs/ELS Net impact/veh Net impact (in $M)

3% Discount ...... (PC) ...... $543.83 $262,923,019 ¥0.1 (LTV) ...... 297.12 13,794,413 0.93

Total ...... 522.22 276,717,432 ¥0.04

7% Discount ...... (PC) ...... 403.84 195,243,258 0.33 (LTV) ...... 209.40 9,722,005 1.67

Total ...... 386.81 204,965,263 0.4

The net impact of this rule on LSVs requirements for these vehicles is discount rate. Thus, the total net impact is also expected to be positive. The net $1,023,934 at the 3-percent discount of the rule considering both the MY2016 benefits of the minimum sound rate and $788,953 at the 7-percent light vehicle and LSV fleet is positive.

172 Based on the assumption in this analysis that manufacturers will install speakers to meet the rule.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00095 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90510 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

TABLE 28—COSTS AND SCALED BENEFITS FOR LSVS, MY2020 173

Scaled Sales ratio Scaled injuries Scaled benefits Discount rate LSV to light Sales Scaled costs (undisc.) Scaled ELS benefits minus vehicle scaled costs

3% ...... 0.47% 2,500 $197,264 11.28 0.1210 $1,189,469 $1,305,543 7% ...... 0.47% 2,500 194,970 11.28 0.0934 848,651 967,019

Alternative 2 (the final rule), the agency 3, but under the Preferred Alternative a would require a sound addition at single listener would perceive a D. Retrospective Review speeds at or below 30 km/h and would noticeable difference in sound level NHTSA has been unable to directly require that covered vehicles produce when standing 7.5 meters from the compare pedestrian and pedalcyclist sound at the stationary but active roadway compared to the no action crash rates for hybrids with and without operating condition. Under Alternative alternative. sound because sufficient data is not yet 3, the agency would require a minimum The results from second analysis available. As a result, we have not been sound pressure level of 48 A-weighted show that changes in overall sound able to directly determine whether lack dB for speeds at or below 20 km/h; there levels near a busy roadway for either of sound is the cause of the difference would be no sound requirement when in pedestrian and pedalcyclist crash action alternative compared to the No the vehicle is stationary. Action Alternative would not exceed 3 rates between hybrids and ICEs. For this In order to determine the potential reason, we intend conduct an expedited dB, the commonly used threshold for environmental impacts of the noticeability by human listeners, even retrospective review of this rule once alternatives, NHTSA estimated the data are available. Although some assuming that up to 20% of vehicles on amount of travel covered by vehicles the road are EVs/HVs, which is nearly hybrid manufacturers began putting and changes in sound level projected to alert sound in their vehicles around three times the deployment level occur under each of the alternatives. currently projected for 2035. When non- 2012, the state data from this period NHTSA separately analyzed the needed for our analysis is just starting urban or urban ambient sound levels are projected environmental impacts of taken into account, the perceived sound to become available. While these each of the three alternatives in both voluntarily equipped vehicles will not level change is further reduced to well urban and non-urban environments under the 3 dB threshold. be fully compliant with this rule, within because differences in population, the next four years we will conduct a vehicle speeds, and deployment of EVs/ In addition to analyzing the projected preliminary study to determine whether HVs in these areas could affect the impact of the action alternatives on an adding sound eliminates some potential environmental impacts. The individual listener, NHTSA computed pedestrian and pedalcyclist crashes EA calculates the potential noise the magnitude of the change in sound should we have sufficient data for such impacts of the alternatives in two levels nationally as a result of the analysis. It will take several more years different ways. alternatives. This analysis takes into until data from fully compliant vehicles In one analysis, NHTSA analyzed the account the National Household Travel are available for analysis. Therefore, we potential for change in sound levels Survey (NHTS) distribution of trip expect to complete our retrospective experienced by an individual listener miles, the Annual Energy Outlook review of this rule within eight years of near a roadway as a result of the final (AEO) forecast of the deployment of when this rule is finalized. For LSVs, alternatives by single vehicle passes by. EVs/HVs, and Environmental Protection sufficient data may not be available and In the second analysis, NHTSA Agency (EPA) drive cycle speed it may be necessary to use a Special compared the sound levels experienced distributions. Because the action Crash Investigation to determine by a single listener among sets of alternatives would only affect specific whether adding sound makes these vehicles with varying percentages of vehicles in certain operating conditions, types of vehicles safer than those EVs/HVs when these vehicles were this analysis calculates the total U.S. without sound should we be able to assumed to have no minimum sound vehicle operations affected by the action identify any such crashes. requirement versus when producing the alternatives as a proportion of total U.S. vehicle operations, and analyzes the E. Environmental Assessment sound level specified under each of the action alternatives. For this analysis, overall change in sound levels projected The agency has prepared an NHTSA calculated the difference in to occur as a result of the action Environmental Assessment (EA) to sound perceived by a person standing alternatives. analyze and disclose the potential either 7.5 or 15 meters (25 or 50 feet, Based on this analysis of national environmental impacts of a reasonable respectively) away from the source to impacts, NHTSA projects that under the range of minimum sound requirements replicate the difference in sound Preferred Alternative, 2.3 percent of all for HVs and EVs, including a preferred between the alternatives experienced by urban U.S. light duty vehicle hours alternative. The alternatives the agency a person standing near a busy roadway. travelled and 0.3 percent of all non- analyzed include a No Action Our first analysis for both action urban U.S. light duty vehicle hours Alterative, under which the agency alternatives suggest that in urban travelled potentially would be impacted would not establish any minimum environments, a single listener would by the minimum sound requirement. sound requirements for EVs/HVs, and not perceive a noticeable difference in Under Alternative 3, NHTSA projects two action alternatives. Under sound when standing 7.5 meters from that 0.9 percent of all urban U.S. light the roadway compared to the no action duty vehicle hours and 0.1 percent of all 173 Scaled benefits and costs for low speed nonurban U.S. light duty vehicle hours vehicles are estimated to be directly proportional to alternative. In a non-urban environment, light vehicles based on sales. Scaled costs include a single listener would not perceive a potentially would be impacted by the both installation costs for the system and fuel costs. noticeable difference under Alternative minimum sound requirement.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00096 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90511

Given the extremely small percentage issues. In some cases, the differences impacts to a variety of resources of vehicle hours travelled impacted by between the regulatory approaches of U.S. including biological resources, waste, this rule and the fact the sounds under agencies and those of their foreign and environmental justice populations. the final rule would only be noticeable counterparts might not be necessary and The findings of the EA are summarized might impair the ability of American to a single listener standing 7.5 meters businesses to export and compete in Section V.D. from the roadway under the single internationally. In meeting shared challenges The Final EA is available in Docket vehicle pass by condition, the involving health, safety, labor, security, No. NHTSA–2011–0100 at http:// environmental impacts of the final rule environmental, and other issues, www.regulations.gov/ as well as on are expected to be negligible. In international regulatory cooperation can NHTSA’s Web site at http:// addition, the EA anticipates no or identify approaches that are at least as www.nhtsa.gov/. Additionally, hard negligible additional impacts on protective as those that are or would be copies may be obtained by contacting wildlife; topography, geology, and soils; adopted in the absence of such cooperation. Mike Pyne, Safety Standards Engineer, International regulatory cooperation can also hazardous materials, hazardous waste, National Highway Traffic Safety reduce, eliminate, or prevent unnecessary Administration, 1200 New Jersey Ave. and solid waste; water resources; differences in regulatory requirements. historical and archeological resources; SE., Washington, DC 20590–0001. farmland resources; air quality and We received several comments I have reviewed the Final EA, which climate; and environmental justice regarding the impact of the rulemaking is hereby incorporated by reference. As populations. schedule on the development of GTR of described in that Final EA and this topic. As discussed in Section IV of summarized above, this rulemaking is VI. Regulatory Notices and Analyses this notice, given the deadlines for anticipated to have no or negligible Executive Order (E.O.) 12866 issuing a final rule provided in the impacts on the human environment. (Regulatory Planning and Review), E.O. PSEA, the agency did not think that it Based on the Final EA, I conclude that 13563, and DOT Regulatory Policies and would be feasible to delay issuing a implementation of any of the action Procedures final rule until after the GTR is alternatives (including the final rule) completed. will not have a significant effect on the The agency has considered the impact NHTSA also received comments human environment and that a ‘‘finding of this rulemaking action under E.O. regarding the approach taken in of no significant impact’’ (see 40 CFR 12866, E.O. 13563, and the Department guidelines developed by the UNECE and 1501.4(e)(1) and 1508.13) is appropriate. of Transportation’s regulatory policies Japan regarding the crossover speed and This statement constitutes the agency’s and procedures. This action was whether HVs and EVs should be ‘‘finding of no significant impact,’’ and reviewed by the Office of Management required to produce sound when they an environmental impact statement will and Budget under E.O. 12866. This are not in motion. For the reasons not be prepared. action is ‘‘significant’’ under the discussed in Section III.D of this notice, Department of Transportation’s we believe that a crossover speed of 30 Regulatory Flexibility Act regulatory policies and procedures (44 km/h is necessary to ensure that blind, Pursuant to the Regulatory Flexibility FR 11034; February 26, 1979). visually-impaired, and sighted Act (5 U.S.C. 601 et seq., as amended by This action is significant because it is pedestrians can safely detect EVs and the Small Business Regulatory the subject of congressional interest and HVs operating at low speeds. For the Enforcement Fairness Act (SBREFA) of because it is a mandate under the PSEA. reasons discussed in Section III.C of this 1996), whenever an agency is required The agency has prepared and placed in notice, we believe that EVs and HVs to publish a notice of rulemaking for the docket a Final Regulatory Impact must produce sound when stationary any proposed or final rule, it must Analysis. with their gear selector is in any prepare and make available for public We estimate the total fuel and position other than park to prevent comment a regulatory flexibility installation costs of this rule to the light collisions and because of the language analysis that describes the effect of the EV, HV and LSV fleet to be $41.8M at of the PSEA. rule on small entities (i.e., small the 3-percent discount rate and $41.3M businesses, small organizations, and at the 7-percent discount rate. We National Environmental Policy Act small governmental jurisdictions). The estimate that the impact of this rule in Concurrently with this final rule, Small Business Administration’s pedestrian and pedalcyclist injury NHTSA is releasing a Final EA, regulations at 13 CFR part 121 define a reduction in light vehicles and LSVs pursuant to the National Environmental small business, in part, as a business will be 30.69 equivalent lives saved at Policy Act, 42 U.S.C. 4321–4347, and entity ‘‘which operates primarily within the 3-percent discount rate and 24.75 implementing regulations issued by the the United States.’’ 174 No regulatory equivalent lives saved at the 7-percent Council on Environmental Quality flexibility analysis is required if the discount rate. The benefits of applying (CEQ), 40 CFR part 1500, and NHTSA, head of an agency certifies the rule will this rule to light EVs and HVs are 49 CFR part 520. NHTSA prepared the not have a significant economic impact estimated to be $260.1 million at the 3- EA to analyze and disclose the potential on a substantial number of small percent discount rate and $209.5 environmental impacts of the entities. SBREFA amended the million at the 7-percent discount rate. requirements of the proposed action and Regulatory Flexibility Act to require Thus, this action is also significant a range of alternatives. The EA analyzes Federal agencies to provide a statement because it has an annual economic direct, indirect, and cumulative impacts of the factual basis for certifying that a impact greater than $100 million. and analyzes impacts in proportion to rule will not have a significant their significance. economic impact on a substantial Executive Order 13609: Promoting Because this rule will increase the International Regulatory Cooperation number of small entities. amount of sound produced by a certain In issuing this rule, I the undersigned The policy statement in Section 1 of segment of the vehicle fleet, the EA hereby certify that this rule will not Executive Order 13609 provides, in part: considers the possible impacts of have a significant economic impact on The regulatory approaches taken by foreign increased ambient noise levels on both a substantial number of small entities. governments may differ from those taken by urban and rural environments. The EA U.S. regulatory agencies to address similar also describes potential environmental 174 13 CFR 121.105(a).

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00097 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90512 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

We believe that the rulemaking will not subdivision of a State may prescribe or standard. As such, NHTSA does not have a significant economic impact on continue in effect a standard applicable intend that this rule preempt state tort the small vehicle manufacturers because to the same aspect of performance of a law that would effectively impose a the systems are not technically difficult motor vehicle or motor vehicle higher standard on motor vehicle to develop or install and the cost of the equipment only if the standard is manufacturers than that established by systems between $50.49 and $125.34 is identical to the standard prescribed today’s final rule. Establishment of a small in proportion to the overall under this chapter. 49 U.S.C. higher standard by means of State tort vehicle cost for most small vehicle 30103(b)(1). It is this statutory command law would not conflict with the manufacturers. by Congress that preempts any non- minimum standard promulgated here. This rule will directly affect motor identical State legislative and Without any conflict, there could not be vehicle manufacturers and final-stage administrative law addressing the same any implied preemption of a State manufacturers that produce EVs and aspect of performance. common law tort cause of action. HVs. The majority of motor vehicle The express preemption provision manufacturers will not qualify as a described above is subject to a savings Executive Order 12988 (Civil Justice small business. There are less than five clause under which ‘‘[c]ompliance with Reform) manufacturers of light hybrid and a motor vehicle safety standard With respect to the review of the electric vehicles that would be subject prescribed under this chapter does not promulgation of a new regulation, to the requirements of this proposal that exempt a person from liability at Section 3(b) of Executive Order 12988, are small businesses. Similarly, there common law.’’ (49 U.S.C. 30103(e)). ‘‘Civil Justice Reform’’ (61 FR 4729; Feb. are several manufacturers of low-speed Pursuant to this provision, State 7, 1996), requires that Executive vehicles that are small businesses. common law tort causes of action agencies make every reasonable effort to Because the PSEA applies to all motor against motor vehicle manufacturers ensure that the regulation: (1) Clearly vehicles (except trailers) in its mandate that might otherwise be preempted by specifies the preemptive effect; (2) to reduce quiet vehicle collisions with the express preemption provision are clearly specifies the effect on existing pedestrians, all of these small generally preserved. However, the Federal law or regulation; (3) provides manufacturers that produce hybrid or Supreme Court has recognized the a clear legal standard for affected electric vehicles are affected by the possibility, in some instances, of conduct, while promoting simplification requirements in today’s final rule. implied preemption of such State and burden reduction; (4) clearly However, the economic impact upon common law tort causes of action by specifies the retroactive effect, if any; (5) these entities will not be significant for virtue of NHTSA’s rules, even if not specifies whether administrative the following reasons. expressly preempted. This second way proceedings are to be required before (1) The cost of the systems is a small that NHTSA rules can preempt is parties file suit in court; (6) adequately proportion of the overall vehicle cost for dependent upon there being an actual defines key terms; and (7) addresses even the least expensive electric conflict between an FMVSS and the other important issues affecting clarity vehicles. higher standard that would effectively and general draftsmanship under any (2) This final rule provides a three be imposed on motor vehicle guidelines issued by the Attorney year lead-time and allows small volume manufacturers if someone obtained a General. This document is consistent manufacturers the option of waiting State common law tort judgment against with that requirement. until the end of the phase-in (September the manufacturer, notwithstanding the Pursuant to this Order, NHTSA notes 1, 2018) to meet the minimum sound manufacturer’s compliance with the as follows. The issue of preemption is requirements. NHTSA standard. Because most NHTSA discussed above. NHTSA notes further that there is no requirement that Executive Order 13132 (Federalism) standards established by an FMVSS are minimum standards, a State common individuals submit a petition for NHTSA has examined today’s rule law tort cause of action that seeks to reconsideration or pursue other pursuant to Executive Order 13132 (64 impose a higher standard on motor administrative proceedings before they FR 43255, August 10, 1999) and vehicle manufacturers will generally not may file suit in court. concluded that no additional be preempted. However, if and when Unfunded Mandates Reform Act consultation with States, local such a conflict does exist—for example, governments or their representatives is when the standard at issue is both a Section 202 of the Unfunded mandated beyond the rulemaking minimum and a maximum standard— Mandates Reform Act of 1995 (UMRA) process. The agency has concluded that the State common law tort cause of requires federal agencies to prepare a the rulemaking would not have action is impliedly preempted. See written assessment of the costs, benefits, sufficient federalism implications to Geier v. American Honda Motor Co., and other effects of proposed or final warrant consultation with State and 529 U.S. 861 (2000). rules that include a Federal mandate local officials or the preparation of a Pursuant to Executive Order 13132 likely to result in the expenditure by federalism summary impact statement. and 12988, NHTSA has considered State, local, or tribal governments, in the The proposed rule would not have whether this rule could or should aggregate, or by the private sector, of ‘‘substantial direct effects on the States, preempt State common law causes of more than $100 million annually on the relationship between the national action. The agency’s ability to announce (adjusted for inflation with base year of government and the States, or on the its conclusion regarding the preemptive 1995). Adjusting this amount by the distribution of power and effect of one of its rules reduces the implicit gross domestic product price responsibilities among the various likelihood that preemption will be an deflator for 2010 results in $136 million levels of government.’’ issue in any subsequent tort litigation. (110.659/81.536 = 1.36). NHTSA rules can preempt in two To this end, the agency has examined As noted previously, the agency has ways. First, the National Traffic and the nature (e.g., the language and prepared a detailed economic Motor Vehicle Safety Act contains an structure of the regulatory text) and assessment in the FRIA. We estimate the express preemption provision: When a objectives of today’s rule and finds that annual total fuel and installation costs motor vehicle safety standard is in effect this rule, like many NHTSA rules, of this final rule to the light EV, HV and under this chapter, a State or a political prescribes only a minimum safety LSV fleet to be $41.8 million at the 3-

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00098 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90513

percent discount rate and $41.3 million are approximately 21 such Act (NTTAA) requires NHTSA to at the 7-percent discount rate. manufacturers. The proposed collection evaluate and use existing voluntary Therefore, this rule is not expected to would occur one per year. consensus standards in its regulatory result in the expenditure by State, local, Estimate of the Total Annual activities unless doing so would be or tribal governments, in the aggregate, Reporting and Recordkeeping Burden inconsistent with applicable law (e.g., or by the private sector, of more than Resulting from the Collection of the statutory provisions regarding $136 million annually. Information: NHTSA estimates that the NHTSA’s vehicle safety authority) or total annual burden is 42 hours (2 hours otherwise impractical. Paperwork Reduction Act per manufacturer per year). Voluntary consensus standards are Under the Paperwork Reduction Act Comments are invited on: technical standards developed or of 1995, a person is not required to • Whether the collection of adopted by voluntary consensus respond to a collection of information information is necessary for the proper standards bodies. Technical standards by a Federal agency unless the performance of the functions of the are defined by the NTTAA as collection displays a valid OMB control Department, including whether the ‘‘performance-based or design-specific number. The final rule contains information will have practical utility. technical specification and related reporting requirements so that the • Whether the Department’s estimate management systems practices.’’ They agency can determine if manufacturers for the burden of the information pertain to ‘‘products and processes, comply with the phase in schedule. collection is accurate. such as size, strength, or technical In compliance with the PRA, this • Ways to minimize the burden of the performance of a product, process or notice announces that the Information collection of information on material.’’ Collection Request (ICR) abstracted respondents, including the use of Examples of organizations generally below has been forwarded to OMB for automated collection techniques or regarded as voluntary consensus review and comment. The ICR describes other forms of information technology. standards bodies include the American the nature of the information collections A comment to OMB is most effective Society for Testing and Materials and their expected burden. This is a if OMB receives it within 30 days of (ASTM), the Society of Automotive request for new collection. publication. Send comments to the Engineers (SAE), and the American Agency: National Highway Traffic Office of Information and Regulatory National Standards Institute (ANSI). If Safety Administration (NHTSA). Affairs, Office of Management and NHTSA does not use available and Title: 49 CFR part 575.141, Minimum Budget, 725 17th Street NW., potentially applicable voluntary Sound Requirements for Hybrid and Washington, DC 20503, Attn: NHTSA consensus standards, we are required by Electric Vehicles. Desk Officer. PRA comments are due the Act to provide Congress, through Type of Request: New collection. within 30 days following publication of OMB, an explanation of the reasons for OMB Clearance Number: Not this document in the Federal Register. not using such standards. assigned. The agency recognizes that the The agency uses certain parts of Form Number: The collection of this collection of information contained in voluntary consensus standard SAE information will not use any standard today’s final rule may be subject to J2889–1, Measurement of Minimum forms. revision in response to public comments Noise Emitted by Road Vehicles, in the Requested Expiration Date of and the OMB review. test procedure contained in this final Approval: Three years from the date of rule. SAE J2889–1 only contains approval. Executive Order 13045 measurement procedures and does not Summary of the Collection of Executive Order 13045 175 applies to contain any minimum performance Information: This collection would any rule that: (1) Is determined to be requirements. The agency did not use require manufacturers of passenger cars, economically significant as defined any voluntary consensus standards for multipurpose passenger vehicles, under E.O. 12866, and (2) concerns an the minimum acoustic requirements trucks, buses, and low speed vehicles environmental, health or safety risk that contained in today’s final rule because subject to the phase-in schedule to NHTSA has reason to believe may have no such voluntary consensus standards provide motor vehicle production data a disproportionate effect on children. If exist. The agency added additional test for one year: September 1, 2018 to the regulatory action meets both criteria, scenarios other than those contained in August 31, 2019. we must evaluate the environmental SAE J2889–1 because those additional Description of the Need for the health or safety effects of the proposed test scenarios address aspects of Information and Use of the Information: rule on children, and explain why the performance not covered in that The purpose of the reporting proposed regulation is preferable to standard. requirements will be to aid NHTSA in other potentially effective and The agency also used voluntary determining whether a manufacturer reasonably feasible alternatives consensus standard ISO 10844 has complied with the requirements of considered by us. ‘‘Acoustics—Test Surface for Road Federal Motor Vehicle Safety Standard This rule will not pose such a risk for Vehicle Noise Measurements,’’ to No. 141, Minimum Sound for Hybrid children. The primary effects of this rule specify the road surface to be used for and Electric Vehicles, during the phase- are to ensure that hybrid and electric compliance testing under this standard. in of those requirements. vehicles produce enough sound so that We also used ANSI S1.11 ‘‘Specification Description of the Likely Respondents pedestrians can detect them. We expect for Octave-Band and Fractional-Octave- (Including Estimated Number, and this rule to reduce the risk of injuries to Band Analog and Digital Filters,’’ to Proposed Frequency of Response to the children and other pedestrians. specify the filter roll-offs to be used Collection of Information): The during the analyses of data collected National Technology Transfer and respondents are manufacturers of hybrid during compliance testing. and electric passenger cars, Advancement Act multipurpose passenger vehicles, Section 12(d) of the National Incorporation by Reference trucks, buses, and low-speed vehicles Technology Transfer and Advancement As discussed earlier in the relevant with a GVWR of 4,536 kg (10,000 lbs.) portions of this document, we are or less. The agency estimates that there 175 62 FR 19885 (Apr. 23, 1997). incorporating by reference various

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00099 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90514 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

materials into the Code of Federal Executive Order 13211 PART 571—FEDERAL MOTOR Regulations in this rulemaking. The VEHICLE SAFETY STANDARDS standards we are incorporating are ANSI Executive Order 13211 177 applies to S1.11–2004, ‘‘Specification for Octave- any rule that: (1) Is determined to be ■ 1. The authority citation for part 571 Band and Fractional-Octave-Band economically significant as defined continues to read as follows: Analog and Digital Filters,’’ the 1994, under E.O. 12866, and is likely to have Authority: 49 U.S.C. 322, 30111, 30115, 2011, and 2014 versions of ISO a significant adverse effect on the 30117, and 30166; delegation of authority at 10844 176 and SAE Standard J2889–1 supply, distribution, or use of energy; or 49 CFR 1.95. Dec. 2014, ‘‘Measurement of Minimum (2) that is designated by the ■ 2. In § 571.5: Noise Emitted by Road Vehicles,’’ Administrator of the Office of ■ a. Redesignate paragraphs (c)(1) Under 5 U.S.C. 552(a)(1)(E), Congress Information and Regulatory Affairs as a through (4) as paragraphs (c)(2) through allows agencies to incorporate by significant energy action. If the (5); reference materials that are reasonably regulatory action meets either criterion, ■ b. Add new paragraph (c)(1); available to the class of persons affected we must evaluate the adverse energy ■ c. Add paragraphs (i)(2) through (4); if the agency has approval from the effects of the proposed rule and explain and ■ Director of the Federal Register. As a why the proposed regulation is d. Redesignate paragraph (l)(49) as part of that approval process, the preferable to other potentially effective paragrapgh (l)(50) and, and add new Director of the Federal Register (in 1 and reasonably feasible alternatives paragraah (l)(49). The additions read as follows: CFR 51.5) directs agencies to discuss (in considered by NHTSA. the preamble) the ways that the § 571.5 Matter incorporated by reference. materials we are incorporating by This rule seeks to ensure that hybrid reference are reasonably available to and electric vehicles are detectable by * * * * * pedestrians. The average weight gain for (c) * * * interested parties. (1) ANSI S1.11–2004, ‘‘Specification NHTSA has worked to ensure that a light vehicle is estimated to be 1.5 pounds (based upon a similar for Octave-Band and Fractional-Octave- standards being considered for Band Analog and Digital Filters,’’ incorporation by reference are waterproof speaker used for marine purposes), resulting in 2.3 more gallons approved February 19, 2004, into reasonably available to the class of § 571.141. persons affected. In this case, those of fuel being used over the lifetime of directly affected by incorporated a passenger car and 2.5 more gallons of * * * * * (i) * * * provisions are NHTSA and parties fuel being used over the lifetime of a (2) ISO 10844:1994(E) ‘‘Acoustics— contracting with NHTSA to conduct light truck. When divided by the life Test Surface for Road Vehicle Noise testing of new vehicles. New vehicle time of the vehicle (26 years for Measurements,’’ First edition, 1994–09– manufacturers may also be affected to passenger cars and 36 years for light 01, into § 571.141. the extent they wish to conduct trucks) the yearly increase in fuel (3) ISO 10844: 2011(E) ‘‘Acoustics— NHTSA’s compliance test procedures on consumption attributed to this proposed Specification of test tracks for their own vehicles. These entities have rule would be negligible. Therefore, this measuring noise emitted by road access to copies of aforementioned proposed rule would not have a vehicles and their tyres,’’ Second standards through ANSI, ISO and SAE significant adverse effect on the use of edition, 2011–02–01 into § 571.141. International for a reasonable fee. These energy. Accordingly, this rulemaking (4) ISO 10844: 2014(E) ‘‘Acoustics— entities have the financial capability to action is not designated as a significant Specification of test tracks for obtain a copy of the material energy action. measuring noise emitted by road incorporated by reference. Other vehicles and their tyres,’’ Third edition, interested parties in the rulemaking Regulation Identifier Number (RIN) 2014–05–15 into § 571.141. process beyond the class affected by the * * * * * regulation include members of the The Department of Transportation assigns a regulation identifier number (l) * * * public, safety advocacy groups, etc. (49) SAE Standard J2889–1, (RIN) to each regulatory action listed in Such interested parties can access the ‘‘Measurement of Minimum Noise the Unified Agenda of Federal standard by obtaining a copy from the Emitted by Road Vehicles,’’ December aforementioned standards development Regulations. The Regulatory Information 2014 into § 571.141. organizations. Service Center publishes the Unified * * * * * Interested parties may also access the Agenda in April and October of each ■ standards through NHTSA. All year. You may use the RIN contained in 3. Section 571.141 is added to read as approved material is available for the heading at the beginning of this follows: inspection at NHTSA, 1200 New Jersey document to find this action in the § 571.141 Standard No. 141; Minimum Avenue SE., Washington, DC 20590, and Unified Agenda. Sound Requirements for Hybrid and at the National Archives and Records Electric Vehicles. List of Subjects in 49 CFR Part 571 Administration (NARA). For S1. Scope. This standard establishes information on the availability of this Imports, Incorporation by reference, performance requirements for material at NHTSA, contact NHTSA’s Motor vehicle safety, Reporting and pedestrian alert sounds for motor Office of Technical Information recordkeeping requirements, Tires. vehicles. Services, phone number (202) 366– S2. Purpose. The purpose of this 2588. Regulatory Text standard is to reduce the number of injuries that result from electric and 176 The 1994 version of ISO 10844 is titled In accordance with the forgoing, hybrid vehicle crashes with pedestrians ‘‘Acoustics—Test Surface for Road Vehicle Noise NHTSA is amending 49 CFR part 571 as by providing a sound level and sound Measurements’’ the 2011 and 2014 versions of ISO follows: 10844 are titled ‘‘Acoustics—Specification of test characteristics necessary for these tracks for measuring noise emitted by road vehicles vehicles to be detected and recognized and their tyres.’’ 177 66 FR 28355 (May 18, 2001). by pedestrians.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00100 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90515

S3. Application. This standard (c) Electric vehicles and hybrid ‘‘Specification for Octave-Band and applies to— vehicles that are low speed vehicles. Fractional-Octave-Band Analog and (a) Electric vehicles with a gross S4. Definitions. Band or one-third Digital Filters’’ (incorporated by vehicle weight rating (GVWR) of 4,536 octave band means one of thirteen one- reference, see § 571.5). Kg or less that are passenger cars, multipurpose passenger vehicles, third octave bands having nominal Band sum means the combination of trucks, or buses; center frequencies ranging from 315 to Sound Pressure Levels (SPLs) from (b) Hybrid vehicles with a gross 5000Hz. These are Bands 25 through 37 selected bands that produce an SPL vehicle weight rating (GVWR) of 4,536 as defined in Table A1, Mid-band representing the sound in all of these Kg or less that are passenger cars, multi- Frequencies for One-Third-Octave-Band bands. Band sum is calculated with the purpose passenger vehicles, trucks, or and Octave-Band Filters in the Audio following equation: buses; and Range, of ANSI S1.11–2004:

where SPLi is the sound pressure level S5.1.1.2 For directivity, the vehicle TABLE 2—ONE-THIRD OCTAVE BAND in each selected band. must emit a sound measured at the MIN. SPL REQUIREMENTS FOR Electric vehicle means a motor vehicle microphone on the line CC’ having at SOUND WHILE IN REVERSE—Contin- with an electric motor as its sole means least the A-weighted sound pressure ued of propulsion. level according to Table 1 in each of Front plane of the vehicle means a four non-adjacent bands spanning no One-third octave band center Min SPL, vertical plane tangent to the leading fewer than 9 of the 13 bands from 315 frequency, Hz A-weighted dB edge of the vehicle during forward to 5000Hz. operation. 500 ...... 43 Hybrid vehicle means a motor vehicle TABLE 1—ONE-THIRD OCTAVE BAND 630 ...... 43 which has more than one means of 800 ...... 44 MIN. SPL REQUIREMENTS FOR 1000 ...... 44 propulsion for which the vehicle’s SOUND WHEN STATIONARY AND propulsion system can propel the 1250 ...... 45 CONSTANT SPEEDS LESS THAN 1600 ...... 41 vehicle in the normal travel mode in at 10KM/H 2000 ...... 42 least one forward drive gear or reverse 2500 ...... 40 without the internal combustion engine 3150 ...... 37 One-third octave band center Min SPL, operating. frequency, Hz A-weighted dB 4000 ...... 35 Rear plane means a vertical plane 5000 ...... 33 tangent to the leading edge of the rear 315 ...... 39 of the vehicle during operation in 400 ...... 39 S5.1.3 Constant pass-by speeds reverse. 500 ...... 40 greater than 0 km/h but less than 20 S5. Requirements. Subject to the 630 ...... 40 km/h. When at a constant speed greater phase-in set forth in S9 of this standard, 800 ...... 41 than 0 km/h but less than 20 km/h the each hybrid and electric vehicle must 1000 ...... 41 vehicle must emit a sound having at meet the requirements specified in 1250 ...... 42 least the A-weighted sound pressure 1600 ...... 39 level according to Table 1 or Table 3 as either S5.1 or S5.2. subject to the 2000 ...... 39 requirements in S5.3. Each vehicle must 2500 ...... 37 applicable based upon vehicle test also meet the requirements in S5.4 and 3150 ...... 34 speed in each of four non-adjacent S5.5. 4000 ...... 32 bands spanning no fewer than 9 of the S5.1 Performance requirements for 5000 ...... 31 13 bands from 315 to 5000 Hz. four-band alert sounds. S5.1.1 Stationary. When stationary S5.1.2 Reverse. For vehicles capable TABLE 3—ONE-THIRD OCTAVE BAND the vehicle must satisfy S5.1.1.1 and of rearward self-propulsion, whenever MIN. SPL REQUIREMENTS FOR CON- S5.1.1.2 whenever the vehicle’s the vehicle’s gear selector is in the STANT PASS-BY SPEEDS GREATER propulsion system is activated and: Reverse position, the vehicle must emit THAN OR EQUAL TO 10 KM/H BUT (i) In the case of a vehicle with an a sound having at least the A-weighted LESS THAN 20 KM/H automatic transmission, the vehicle’s sound pressure level according to Table gear selector is in Neutral or any gear 2 in each of four non-adjacent bands One-third octave band center Min SPL, position other than Park that provides spanning no fewer than 9 of the 13 frequency, Hz A-weighted dB forward vehicle propulsion; bands from 315 to 5000Hz. (iii) in the case of a vehicle with a 315 ...... 45 manual transmission, the vehicle’s 400 ...... 44 TABLE 2—ONE-THIRD OCTAVE BAND 500 ...... 46 parking brake is released and the gear MIN. SPL REQUIREMENTS FOR selector is not in Reverse. 630 ...... 46 S5.1.1.1 For detection, the vehicle SOUND WHILE IN REVERSE 800 ...... 47 1000 ...... 47 must emit a sound having at least the A- One-third octave band center Min SPL, 1250 ...... 48 weighted sound pressure level frequency, Hz A-weighted dB 1600 ...... 44 according to Table 1 in each of four non- 2000 ...... 45 adjacent bands spanning no fewer than 315 ...... 42 2500 ...... 43 9 of the 13 bands from 315 to 5000 Hz. 400 ...... 41 3150 ...... 40

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00101 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 ER14DE16.018 90516 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

TABLE 3—ONE-THIRD OCTAVE BAND TABLE 4—ONE-THIRD OCTAVE BAND TABLE 5—ONE-THIRD OCTAVE BAND MIN. SPL REQUIREMENTS FOR CON- MIN. SPL REQUIREMENTS FOR CON- MIN. SPL REQUIREMENTS FOR 30– STANT PASS-BY SPEEDS GREATER STANT PASS-BY SPEEDS GREATER 32 KM/H PASS-BY—Continued THAN OR EQUAL TO 10 KM/H BUT THAN OR EQUAL TO 20 KM/H BUT LESS THAN 20 KM/H—Continued LESS THAN 30 KM/H—Continued One-third octave band center Min SPL, frequency, Hz A-weighted dB One-third octave band center Min SPL, One-third octave band center Min SPL, frequency, Hz A-weighted dB frequency, Hz A-weighted dB 1250 ...... 59 1600 ...... 55 4000 ...... 38 1250 ...... 54 2000 ...... 55 5000 ...... 36 1600 ...... 51 2500 ...... 54 2000 ...... 51 3150 ...... 51 S5.1.4 Constant pass-by speeds 2500 ...... 50 4000 ...... 49 greater than or equal to 20km/h but less 3150 ...... 47 5000 ...... 47 than 30 km/h. When at a constant speed 4000 ...... 45 5000 ...... 43 equal to or greater than 20 km/h but less S5.2 Performance requirements for than 30 km/h the vehicle must emit a S5.1.5 Constant 30km/h pass-by. two-band alert sounds. When operating sound having at least the A-weighted When at a constant speed of 30–32 km/ under the vehicle speed conditions sound pressure level according to Table h the vehicle must emit a sound having specified in Table 6, the vehicle must 4 in each of four non-adjacent bands at least the A-weighted sound pressure emit sound having two non-adjacent spanning no fewer than 9 of the 13 level according to Table 5 in each of one-third octave bands from 315 to 3150 bands from 315 to 5000 Hz. four non-adjacent bands spanning no Hz each having at least the A-weighted fewer than 9 of the 13 bands from 315 sound pressure level according to the TABLE 4—ONE-THIRD OCTAVE BAND to 5000 Hz. minimum SPL requirements in Table 6 MIN. SPL REQUIREMENTS FOR CON- and spanning no fewer than three one- STANT PASS-BY SPEEDS GREATER TABLE 5—ONE-THIRD OCTAVE BAND third octave bands from 315 to 3150 Hz. THAN OR EQUAL TO 20 KM/H BUT MIN. SPL REQUIREMENTS FOR 30– One of the two bands meeting the LESS THAN 30 KM/H 32 KM/H PASS-BY minimum requirements in Table 6 shall be the band that has the highest SPL of One-third octave band center Min SPL, One-third octave band center Min SPL, the 315 to 800 Hz bands and the second frequency, Hz A-weighted dB frequency, Hz A-weighted dB band shall be the band meeting the minimum requirements in Table 6 that 315 ...... 52 315 ...... 56 400 ...... 51 400 ...... 55 has the highest SPL of the 1000 to 3150 500 ...... 52 500 ...... 57 Hz bands. The two bands used to meet 630 ...... 53 630 ...... 57 the two-band minimum requirements 800 ...... 53 800 ...... 58 must also meet the band sum 1000 ...... 54 1000 ...... 58 requirements as specified in Table 6.

TABLE 6—ONE-THIRD OCTAVE BAND MINIMUM REQUIREMENTS FOR TWO-BAND ALERT

A-weighted SPL, dB(A) Vehicle speed Minimum in each band Band sum

Reverse ...... 40 48 Stationary and up to but not including 10 km/h ...... 40 44 10 km/h up to but not including 20 km/h ...... 42 51 20 km/h up to but not including 30 km/h ...... 47 57 30 km/h ...... 52 62

S5.2.1 When tested according to the through S5.1.5 or S5.2 (Stationary, S5.4 Relative volume change to test procedure in S7.1 the vehicle must Reverse, Pass-by at 10 km/h, 20 km/h, signify acceleration and deceleration. emit a sound measured at the and 30 km/h, respectively), and during The sound produced by the vehicle in microphone on the line CC’ having at testing to any one of those requirements accordance with paragraph S5 shall least two non-adjacent octave bands the vehicle is measured for ten change in volume, as calculated in S7.6, from 315 to 3150 Hz each having at least consecutive times without recording a from one critical operating condition to the A-weighted sound pressure level, valid measurement, or for a total of 20 the next in accordance with the indicated in the ‘‘Minimum in Each times without recording four valid requirements in Table 7. Band’’ column in Table 6 for the measurements because the vehicle’s ICE ‘‘Stationary up to but not including 10 remains active for the entire duration of km/h’’ condition. The two bands used to a measurement or the vehicle’s ICE meet the two-band minimum activates intermittently during every requirements must also meet the Band measurement, the vehicle is exempted Sum as specified in Table 6. from meeting the specific requirement S5.3 If a hybrid vehicle to which that was under evaluation at the time this standard applies is evaluated for the ICE interfered in the prescribed compliance with requirements in S5.1.1 manner.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00102 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90517

TABLE 7—MINIMUM RELATIVE VOLUME S6.1.4 Background noise level. The internal combustion engine does not CHANGE REQUIREMENTS background noise level will be activate, manufacturer instructions will measured and reported as specified in be followed. Minimum S6.7, Ambient correction. (d) Vehicle test weight, including the relative S6.2 Test surface. Test surface will driver and instrumentation, will be Critical operating speed intervals volume meet the requirements of ISO evenly distributed between the left and change, dB 10844:1994, ISO 10844:2011, or ISO right side of the vehicle and will not 10844:2014 (incorporated by reference, exceed the vehicle’s GVWR or GAWR: Between: see § 571.5). (1) For passenger cars, and MPVs, Stationary and 10 km/h ...... 3 S6.3 Instrumentation. trucks, and buses with a GVWR of 4,536 10 km/h and 20 km/h ...... 3 S6.3.1 Acoustical measurement. kg (10,000 pounds) or less, the vehicle 20 km/h and 30 km/h ...... 3 Instruments for acoustical measurement test weight is the unloaded vehicle will meet the requirements of S5.1 of weight plus 180 kg (396 pounds); S5.5 Sameness requirement SAE J2889–1 (incorporated by reference, (2) For LSVs, the test weight is the S5.5.1 Any two vehicles of the same see § 571.5). unloaded vehicle weight plus 78 kg (170 make, model, and model year (as those S6.3.2 Vehicle speed measurement. pounds). terms are defined at 49 CFR 565.12) to Instruments used to measure vehicle (e) Tires will be free of all debris and which this safety standard applies shall speed during the constant speed pass-by each tire’s cold tire inflation pressure use the same pedestrian alert system tests in S7 of this standard will be set to: and shall be designed to have the same capable of either continuous (1) For passenger cars, and MPVs, pedestrian alert sound when operating measurement of speed within ±0.5 km/ trucks, and buses with a GVWR of 4,536 in any given condition for which an h over the entire measurement zone kg (10,000 pounds) or less, the inflation alert sound is required in Section S5 of specified in S6.4 or independent pressure specified on the vehicle this safety standard. measurements of speed within ±0.2 km/ placard in FMVSS No. 110; S5.5.2 For the purposes of this h at the beginning and end of the (2) For LSVs, the inflation pressure requirement, a pedestrian alert system measurement zone specified in S6.4. recommended by the manufacturer for includes all hardware and software S6.3.3 Meteorological GVWR; if none is specified, the components that are utilized to generate instrumentation. Instruments used to maximum inflation pressure listed on an alert sound. Aspects of an alert measure ambient conditions at the test the sidewall of the tires. system which shall be the same include, site will meet the requirements of S5.3 (f) Tires are conditioned by driving if applicable: Alert system hardware of SAE J2889–1 (incorporated by the test vehicle around a circle 30 components including speakers, speaker reference, see § 571.5). meters (100 feet) in diameter at a speed modules, and control modules, as S6.4 Test site. The test site will be that produces a lateral acceleration of evidenced by specific details such as established per the requirements of 6.1 0.5 to 0.6 g for three clockwise laps part numbers and technical of SAE J2889–1 (incorporated by followed by three counterclockwise illustrations; the location, orientation, reference, see § 571.5), including Figure laps; and mountings of the hardware 1, ‘‘Test Site Dimensions’’ with the S6.7 Ambient correction. S6.7.1 Measure the ambient noise components within the vehicle; the definitions of the abbreviations in for at least 30 seconds immediately digital sound file or other digitally Figure 1 as given in Table 1of SAE before and after each series of vehicle encoded source; the software and/or J2889–1 (incorporated by reference, see tests. A series is a test condition, i.e. firmware and algorithms which generate § 571.5). Microphone positions will stationary, reverse, 10 km/h pass-by test, the pedestrian alert sound and/or which meet the requirements of 7.1.1 of SAE 20 km/h pass-by test, or 30 km/h pass- process the digital source to generate a J2889–1 (incorporated by reference, see by test. Ambient noise data files will be pedestrian alert sound; vehicle inputs § 571.5). collected from each microphone including vehicle speed and gear S6.5 Test set up for directivity required by the test procedures in S7. selector position utilized by the alert measurement will be as per S6.4 with the addition of one microphone meeting S6.7.2 For each microphone, system; any other design features determine the minimum A-weighted necessary for vehicles of the same make, the requirements of S6.3.1 placed on the line CC’, 2m forward of the line PP’ at overall ambient SPL during the 60 model, and model year to have the same seconds (or more) of recorded ambient pedestrian alert sound at each given a height of 1.2m above ground level. S6.6 Vehicle condition noise consisting of at least 30 seconds operating condition specified in this (a) The vehicle’s doors are shut and recorded immediately before and at safety standard. locked and windows are shut. least 30 seconds immediately after each S6. Test Conditions. (b) All accessory equipment (air test series. S6.1 Weather conditions. The conditioner, wipers, heat, HVAC fan, S6.7.3 For each of the 13 one-third ambient conditions specified by this audio/video systems, etc.) that can be octave bands, the minimum A-weighted section will be met at all times during shut down, will be off. Propulsion ambient noise level during the 60 the tests described in S7. Conditions battery cooling fans and pumps and seconds (or more) from the two 30 will be measured with the accuracy other components of the vehicle’s second periods of ambient noise required in S6.3.3 at the microphone propulsion battery thermal management recorded immediately before and after ¥ height specified in S6.4 +/ 0.02 m. system are not considered accessory each test series will be determined for S6.1.1 The ambient temperature will equipment. During night time testing each microphone. be between 5 °C (41 °F) and 40 °C (104 test vehicle headlights may be activated. S6.7.4 To correct overall SPL values °F). (c) Vehicle’s electric propulsion for ambient noise, calculate the S6.1.2 The maximum wind speed at batteries, if any, are charged according difference, for each microphone, the microphone height is no greater than to the requirements of S7.1.2.2 of SAE between the measured overall SPL 5 m/s (11 mph), including gusts. J2889–1 (incorporated by reference, see values for a test vehicle obtained in S6.1.3 No precipitation and the test § 571.5). If propulsion batteries must be sections S7.1.4(b) and S7.3.4(b) and the surface is dry. recharged during testing to ensure minimum overall ambient SPL values

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00103 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90518 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

determined in S6.7.2, above. Using S6.7.5 To correct one-third octave required, subtract the appropriate Table 8, determine a correction factor band sound levels for ambient noise, correction factor in Table 9 from the for each microphone. Subtract the calculate the difference, for each uncorrected one-third octave band correction factor from the overall SPL microphone, between the uncorrected sound level to calculate the corrected value measured under sections S7.1.4(b) level for a one-third octave band level for each one-third octave band. If and S7.3.4(b) to calculate the corrected (obtained in sections S7.1.5(b), S7.1.6(b) the level of any ambient one-third overall SPL value. Any test for which and S7.3.5(b)) and the minimum octave band is within 3 dB of the the minimum overall SPL of the ambient level in the same one-third corresponding uncorrected one-third ambient is within 3 dB of the octave band as determined in S6.7.3. octave band level, then that one-third Use Table 9 to determine if a correction uncorrected overall SPL of the vehicle is octave band is invalid and not analyzed is required for each microphone and invalid and not analyzed further. further. one-third octave band. If a correction is

TABLE 8—OVERALL SPL CORRECTIONS FOR AMBIENT NOISE

Difference between vehicle measurement and ambient noise level Correction

Greater than 10 dB ...... 0 dB. Greater than 8 dB but less than or equal to 10 dB ...... 0.5 dB. Greater than 6 dB but less than or equal to 8 dB ...... 1.0 dB. Greater than 4.5 dB but less than or equal to 6 dB ...... 1.5 dB. Greater than 3 dB but less than or equal to 4.5 dB ...... 2.5 dB. Less than or equal to 3 dB ...... Invalid test run.

TABLE 9—1/3 OCTAVE BAND CORRECTIONS FOR AMBIENT NOISE

Difference between vehicle 1/3 octave band sound pressure level and ambient noise level Correction

Greater than 6 dB ...... 0 dB. Greater than 4.5 dB but less than or equal to 6 dB ...... 1.5 dB. Greater than 3 dB but less than or equal to 4.5 dB ...... 2.5 dB. Less than or equal to 3 dB ...... Specific 1/3 octave band is not useable.

S7. Test Procedure. place the vehicle’s gear selector in any or continuously are considered invalid. S7.1 Vehicle stationary forward gear. A valid test requires a valid left side, a S7.1.1 Execute stationary tests and (c) Execute multiple tests to acquire at valid right side, and a valid front-center collect acoustic sound files. least four valid tests within 2 dBA acoustic sound file. overall SPL in accordance with S7.1.2 (b) Sequentially number all tests (a) Position the vehicle with the front and S7.1.3. For each test, measure the which are deemed valid based upon the plane at the line PP’, the vehicle sound emitted by the stationary test chronological order in which they were centerline on the line CC’ and the vehicle for a duration of 10 seconds. conducted. Acoustic files will be starting system deactivated. For vehicle (d) During each test a left (driver’s identified with a test sequence number equipped with a Park position, place the side), a right (passenger side), and a and their association with the left side, vehicle’s gear selector in ‘‘Park’’ and front-center acoustic file will be right side, or front center microphone. engage the parking brake. For vehicles recorded. S7.1.3 Identify first four valid tests not equipped with a Park position, place S7.1.2. Eliminate invalid tests. within 2dBA. the vehicle’s gear selector in ‘‘Neutral’’ (a) Determine validity of sound files (a) For each valid test sound file and engage the parking brake. Activate collected during S7.1.1 tests. identified in S7.1.2, determine a the starting system to energize the Measurements that contain any distinct, maximum overall SPL value, in vehicle’s propulsion system. transient, loud sounds (e.g., chirping decibels. Each SPL value will be (b) For vehicles equipped with a Park birds, overhead planes, trains, car doors reported to the nearest tenth of a position for the gear selector, after being slammed, etc.) are considered decibel. activating the starting system to energize invalid. Measurements that contain (b) Compare the first four left-side the vehicle’s propulsion system, apply sounds emitted by any vehicle system SPL values from S7.1.3(a) of this and maintain a full application of the that is automatically activated and paragraph, and determine the range by service brake, disengage the vehicle constantly engaged during the entire 10 taking the difference between the largest parking brake and then place the second performance test are considered and smallest of the four values. In the vehicle’s gear selector in ‘‘Drive,’’ or any valid. Measurements that contain sound same manner, determine the range of forward gear. For vehicles not equipped emitted by any vehicle system that is SPL values for the first four right-side with a Park position for the gear automatically activated and and the first four front-center sound selector, after activating the starting intermittently engaged at any time files. If the range for the left side, right system to energize the vehicle’s during the stationary performance test, side, and front-center are all less than or propulsion system, apply and maintain are considered invalid. Additionally, equal to 2.0 dB, then the twelve sound a full application of the service brake, when testing a hybrid vehicle with an files associated with the first four valid disengage the vehicle parking brake, internal combustion engine, tests will be used for the one-third disengage the manual clutch (fully measurements that include sound octave band evaluations in S7.1.5. and depress and hold the clutch pedal), and emitted by the ICE either intermittently S7.1.6. If the range of the SPL values for

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00104 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90519

the left side are not within 2 dBA, or for determined that no combination paragraph in each of the four one-third the right side are not within 2 dBA, or meeting this selection criterion can octave bands selected in paragraph for the front-center of the vehicle are not satisfy paragraph S7.1.5(d)(ii). S7.1.6(d)(i) to the required minimum within 2 dBA, an iterative process will (ii) Compare the average corrected level of the corresponding one-third be used to consider sound files from sound pressure level from S7.1.5(c) of octave band specified in paragraph additional sequential tests until the this paragraph in each of the four one- S5.1.1, Table 1, to determine range for all three microphone locations third octave bands selected in paragraph compliance. are within 2 dBA for the same sequence S7.1.5(d)(i) to the required minimum (e) For alerts designed to meet the two number recordings for all three level of the corresponding one-third one-third octave band requirements. locations. octave band specified in paragraph (i) Select the two highest one-third S7.1.4 Compare the average overall S5.1.1, Table 1, to determine octave bands that are non-adjacent to SPL for the left and right side of the test compliance. each other and within the range of 315 vehicle to determine which is lower. (e) For alerts designed to meet the Hz up to and including 3150 Hz to (a) Document the maximum overall two-one-third octave band evaluate according to paragraph (ii), SPL values in each of the eight acoustic requirements: below. One band shall be below 1000 data files (four left side files and four (i) Select the two highest one-third Hz and one band shall be at or greater right side files) identified in S7.1.3. octave bands that are non-adjacent to than 1000 Hz. This step will be repeated (b) Correct each of the eight SPL each other and within the range of 315 until compliance is established or it is values from S7.1.4(a) according to S6.7 Hz up to and including 3150 Hz to determined that no combination using the ambient sound level recorded evaluate according to paragraph (ii), meeting this selection criterion can during the test. The results will be below. This step will be repeated until satisfy paragraph S7.1.6(e)(ii). reported to the nearest tenth of a compliance is established or it is (ii) Compare the average corrected decibel. determined that no combination sound pressure level from S7.1.6(c) of (c) Calculate a left-side average and a meeting this selection criterion can this paragraph in each of the two one- right-side average from the ambient- satisfy paragraph S7.1.5(e)(ii). third octave bands selected in paragraph corrected overall SPL values from (ii) Compare the average corrected S7.1.6(e)(i) to the required minimum S71.4(b), and determine the lower of the sound pressure level from (c) in each of level of the corresponding one-third two sides. The result will be reported to the two one-third octave bands selected octave band specified in paragraph S5.2 the nearest tenth of a decibel. in paragraph S7.1.5(e)(i) to the required Table 6. Also, compare the band sum of (d) If the left-side value from S7.1.4(c) minimum level of the corresponding the two bands to the required minimum is the lower one, then the left side one-third octave band specified in level in Table 6. acoustic data will be further evaluated paragraph S5.2 Table 6. Also, compare S7.2 Reverse. Test the vehicle per for compliance at the one-third octave the band sum of the two bands to the S7.1 (S7.1.1–S7.1.5), except that the rear band levels in accordance with S7.1.5. required minimum level in Table 6. plane of the vehicle is placed on line If the left-side value from S7.1.4(c) is S7.1.6 Procedure for selected one- PP’, no third microphone (front center) not the lower one, the right-side third octave bands to be used for is used, and the vehicle’s gear selector acoustic data will be further evaluated evaluating compliance with directivity is placed in ‘‘Reverse.’’ for compliance at the one-third octave requirements. S7.3 Constant speed pass-by tests at band level in accordance with S7.1.5. (a) Determine the one-third octave speeds greater than 0 km/h but less than S7.1.5 Select one-third octave bands band levels associated with the four 20 km/h. to be used for evaluating compliance front center sound files selected in S7.3.1 Execute pass-by tests at with detection requirements. S7.1.3. 11km/h (+/¥1 km/h) and collect (a) For each of the four left-side or (b) The identified one-third octave acoustic sound files. right-side acoustic files, which ever was band levels in each of the four sound (a) For each test, measure the sound selected in S7.1.4, determine the sound files will be corrected for the measured emitted by the test vehicle while at a pressure level in each one-third octave ambient levels as specified in paragraph constant speed of 11km/h (+/¥ 1km/h) band from 315 Hz up to and including S6.7. throughout the measurement zone 5000 Hz. (c) The four corrected sound pressure specified in S6.4 between lines AA’ and (b) Correct the one-third octave band level values calculated from each of the PP’. Execute multiple test runs at 11km/ levels in all four sound files to adjust for four sound files in each one-third octave h (+/¥1km/h) to acquire at least four the ambient sound level recorded band will be averaged together to get the valid tests within 2dBA in accordance during the test according to paragraph average corrected sound pressure level with S7.3.2 and S7.3.3. S6.7. in each one-third octave band. (b) During each test, record a left (c) For each one-third octave band, (d) For alerts designed to meet the (driver’s side) and a right (passenger average the corrected levels from the four one-third octave band side) acoustic sound file. four sound files. The results will be requirements. S7.3.2 Eliminate invalid tests and reported to the nearest tenth of a (i) Select any four one-third octave acoustic sound files decibel. bands that are non-adjacent to each (a) Determine validity of sound files (d) For alerts designed to meet the other and that span a range of at least collected during S7.3.1 tests. four one-third octave band alert sound nine one-third octave bands in the range Measurements that contain any distinct, requirements: of 315 Hz up to and including 5000 Hz transient, background sounds (e.g., (i) Select any four one-third octave to evaluate according to paragraph chirping birds, overhead planes, car bands that are non-adjacent to each S7.1.6(d)(ii). This step will be repeated doors being slammed, etc.) are other and that span a range of at least until compliance is established or it is considered invalid. Measurements that nine one-third octave bands in the range determined that no combination contain sounds emitted by any vehicle of 315 Hz up to and including 5000 Hz meeting this selection criterion can system that is automatically activated to evaluate according to paragraph satisfy paragraph S7.1.6(d)(ii). and constantly engaged during the S7.1.5(d)(ii). This step will be repeated (ii) Compare the average corrected entire performance test are considered until compliance is established or it is sound pressure level from (c) of this valid. Measurements that contain sound

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00105 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90520 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

emitted by any vehicle system that is ambient conditions as specified in Hz up to and including 3150 Hz to automatically activated, and paragraph S6.7. The test results will be evaluate according to paragraph intermittently engaged at any time reported to the nearest tenth of a S7.3.5(e)(ii). This step will be repeated during the performance test, are decibel. until compliance is established or it is considered invalid. Additionally, when (c) Calculate the average of the four determined that no combination testing a hybrid vehicle with an internal overall ambient-corrected SPL values on meeting this selection criterion can combustion engine that runs each side of the vehicle to derive one satisfy paragraph S7.3.5(e)(ii). intermittently during a specific test, corrected maximum overall SPL value (ii) Compare the average corrected measurements that contain sound for each side of the vehicle. The result sound pressure level from S7.3.5(c) in emitted by the ICE are considered will be reported to the nearest tenth of each of the two one-third octave bands invalid. A valid test requires both a a decibel. selected in paragraph S7.3.5(e)(i) to the valid left side and a valid right side (d) The side of the vehicle with the required minimum level of the acoustic sound file. lowest average corrected maximum corresponding one-third octave band (b) Tests which are deemed valid will overall SPL value will be the side of the specified in paragraph S5.2 and Table 6. be numbered sequentially based upon vehicle that is further evaluated for Also, compare the band sum of the two the chronological order in which they compliance at the one-third octave band bands to the required minimum level in were collected. Sound files will retain levels in accordance with S7.3.5. Table 6. their test sequence number and their S7.3.5 Complete one-third octave S7.3.6 Repeat S7.3.1–S7.3.5 using association with the left side or right band evaluation for compliance any other constant vehicle speed equal side microphone. verification. S7.3.3 Identify ‘‘first four valid tests (a) The side of the vehicle selected in to or greater than 10 km/h but less than within 2 dBA’’. S7.3.4 will have four associated 20 km/h. (a) For each valid test sound file individual acoustic sound data files. S7.4 Constant speed pass-by tests at identified in S7.3.2, determine a Each sound file shall be broken down speeds greater than or equal to 20 km/ maximum overall SPL value, in into its one-third octave band levels. h but less than 30 km/h. Repeat the test decibels. The SPL value will be reported (b) The identified octave band levels of S7.3 at 21 km/h (+/¥1km/h). In to the nearest tenth of a decibel. in each of the four sound files will be S7.3.6, the 21km/h (+/¥1km/h) test (b) Compare the first four left side corrected for the measured ambient speed can be replaced using any maximum overall SPL values. Of the levels as specified in paragraph S6.7. constant speed greater than or equal to four SPL values calculate the difference (c) The four corrected sound pressure 20 km/h but less than 30 km/h. between the largest and smallest level values calculated from each of the S7.5 Constant speed pass-by tests at maximum SPL values. The same process four sound files in each one-third octave 30 km/h. Repeat the test of S7.3 at 31 will be used to determine the difference band will be averaged together to get the km/h (+/¥1km/h) between the largest and smallest average corrected sound pressure level S7.6 Relative volume change. The maximum SPL values for the first four in each one-third octave band. valid test run data selected for each right side maximum SPL values. If the (d) For alerts designed to meet the critical operating scenario in S7.1 difference values on the left and right four one-third octave band (S7.1.5(c)), S7.3 (S7.3.5(c)), S7.4 and sides of the test vehicle are both less requirements. S7.5 will be used to derive relative than or equal to 2.0 dB, then the eight (i) Select any four one-third octave volume change as required in S5.4 as sound files associated with the first four bands that are non-adjacent to each follows: other and that span a range of at least valid tests will be used for the final one- S7.6.1 Calculate the average sound nine one-third octave bands in the range third octave band evaluation in pressure level for each of the 13 one- of 315 Hz up to and including 5000 Hz accordance with S7.3.4. and S7.3.5. If third octave bands (315 Hz to 5000 Hz) to evaluate according to paragraph the first four test sound files on each using the four valid test runs identified S7.3.5(d)(ii). This step will be repeated side of the vehicle are not within 2 dBA, for each critical operating scenario from until compliance is established or it is an iterative process will be used to S7.1.3 and S7.3.3 (stationary, 10 km/h determined that no combination consider sound files from additional (11+/¥1km/h), 20 km/h (21+/¥1km/h), meeting this selection criterion can sequential tests until the range for both and 30 km/h (31+/¥1km/h)). microphone locations are within 2 dBA satisfy paragraph S7.3.5(d)(ii). for the same sequence number (ii) Compare the average corrected S7.6.2 For each critical operating recordings for both locations. sound pressure level from S7.3.5(c) in scenario, normalize the levels of the 13 S7.3.4 Determine average overall each of the four one-third octave bands one-third octave bands by subtracting SPL value on each side (left and right) selected in paragraph S7.3.5(d)(i) to the the corresponding minimum SPL values of test vehicle. required minimum level of the specified in Table 1 for the stationary (a) Document the maximum overall corresponding one-third octave band operating condition from each of the SPL value in decibels for each of the specified in paragraph S5.1.3, Table 3, one-third octave band averages eight acoustic sound data files (four left- to determine compliance. calculated in S7.6.1. side files and four right-side files) (e) For alerts designed to meet the two S7.6.3 Calculate the NORMALIZED identified in S7.3.3. one-third octave band requirements. BAND SUM for each critical operating (b) Each of the eight acoustic sound (i) Select the two highest one-third scenario (stationary, 10 km/h (11+/ data file maximum overall SPL values octave bands that are non-adjacent to ¥1km/h), 20 km/h (21+/¥1km/h), and will be corrected for the recorded each other and within the range of 315 30 km/h (31+/¥1km/h)) as follows:

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00106 Fmt 4701 Sfmt 4725 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 ER14DE16.019 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations 90521

Where: 2019, shall comply with this safety Vehicles manufactured on or after i represents the 13 one-third octave bands standard. September 1, 2018, and before and Normalized Band Leveli is the ■ 4. Section 571.500 is amended by September 1, 2019. normalized one-third octave band value adding paragraph S5.(b)(12) to read as derived in S7.6.2. § 585.131 Definitions. follows: S7.6.4 Calculate the relative volume (a) All terms defined in 49 U.S.C. change between critical operating § 571.500 Standard No. 500; Low-speed 30102 are used in their statutory scenarios (stationary to 10km/h; 10km/ vehicles. meaning. h to 20 km/h; 20km/h to 30 km/h) by * * * * * (b) Bus, gross vehicle weight rating or subtracting the NORMALIZED BAND S5.(b) * * * GVWR, low-speed vehicle, SUM of the lower speed operating (12) An alert sound as required by multipurpose passenger vehicle, scenario from the NORMALIZED BAND § 571.141. passenger car, truck, and motorcycle are SUM of the next higher speed operating * * * * * used as defined in § 571.3 of this scenario. For example, the relative chapter. volume change between 10 km/h (11+/ PART 585—PHASE-IN REPORTING (c) Production year means the 12- ¥1km/h) and 20 km/h (21+/¥1km/h) REQUIREMENTS month period between September 1 of would be the NORMALIZED BAND one year and August 31 of the following ■ 4. The authority citation for part 585 SUM level at 21+/¥1km/h minus the year, inclusive. is revised to read as follows: (d) Electric Vehicle, and hybrid NORMALIZED BAND SUM level at vehicle are used as defined in § 571.141 11+/¥1km/h. Authority: 49 U.S.C. 322, 30111, 30115, S8 Prohibition on altering the sound 30117, and 30166; delegation of authority at of this chapter. 49 CFR 1.95 of a vehicle subject to this standard. No § 585.132 Response to inquiries. ■ entity subject to the authority of the 5. Add Subpart N to read as follows: At any time during the production National Highway Traffic Safety Subpart N—Minimum Sound Requirements year ending August 31, 2018, each Administration may: for Hybrid and Electric Vehicles Reporting manufacturer shall, upon request from (a) Disable, alter, replace or modify Requirements the Office of Vehicle Safety Compliance, any element of a vehicle installed as Sec. provide information identifying the original equipment for purposes of 585.128 Scope. vehicles (by make, model and vehicle complying with this Standard, except in 585.129 Purpose. identification number) that have been connection with a repair of a vehicle 585.130 Applicability. certified as complying with the malfunction related to its sound 585.131 Definitions. requirements of Standard No. 141, emission or to remedy a defect or non- 585.132 Response to inquiries. Minimum Sound Requirements for compliance with this standard; or 585.133 Reporting requirements. 585.134 Records. Hybrid and Electric Vehicles (49 CFR (b) Provide any person with any 571.141). The manufacturer’s mechanism, equipment, process or Subpart N—Minimum Sound designation of a vehicle as a certified device intended to disable, alter, replace Requirements for Hybrid and Electric vehicle is irrevocable. or modify the sound emitting capability Vehicles Reporting Requirements of a vehicle subject to this standard, § 585.133 Reporting requirements. except in connection with a repair of § 585.128 Scope. (a) Phase-in reporting requirements. vehicle malfunction related to its sound This subpart establishes requirements Within 60 days after the end of the emission or to remedy a defect or non- for manufacturers of hybrid and electric production year ending August 31, compliance with this standard. passenger cars, trucks, buses, 2018, each manufacturer shall submit a S9 Phase-in schedule. multipurpose passenger vehicles, and report to the National Highway Traffic S9.1 Hybrid and Electric Vehicles low-speed vehicles to submit a report, Safety Administration concerning its manufactured on or after September 1, and maintain records related to the compliance with the requirements of 2018, and before September 1, 2019. For report, concerning the number of such Standard No. 141 Minimum Sound hybrid and electric vehicles to which vehicles that meet minimum sound Requirements for Hybrid and Electric this standard applies manufactured on requirements of Standard No. 141, Vehicles (49 CFR 571.141) for its or after September 1, 2018, and before Minimum Sound Requirements for vehicles produced in that year. Each September 1, 2019, except vehicles Hybrid and Electric Vehicles (49 CFR report shall provide the information produced by small volume 571.141). specified in paragraph (b) of this section manufacturers, the quantity of hybrid and in § 585.2 of this part. and electric vehicles complying with § 585.129 Purpose. (b) Phase-in report content— this safety standard shall be not less The purpose of these reporting (1) Basis for phase-in production than 50 percent of one or both of the requirements is to assist the National goals. Each manufacturer shall provide following: Highway Traffic Safety Administration the number of hybrid vehicles and (a) A manufacturer’s average annual in determining whether a manufacturer electric vehicles manufactured in the production of hybrid and electric has complied with the minimum sound current production year or, at the vehicles on and after September 1, 2015, requirements of Standard No. 141, manufacturer’s option, in each of the and before September 1, 2018; Minimum Sound for Hybrid and three previous production years. A (b) A manufacturer’s total production Electric Vehicles (49 CFR 571.141). manufacturer that is, for the first time, of hybrid and electric vehicles on and manufacturing vehicles for sale in the after September 1, 2018, and before § 585.130 Applicability. United States must report the number of September 1, 2019. This subpart applies to manufacturers vehicles manufactured during the S9.2 Hybrid and Electric Vehicles of hybrid and electric passenger cars, current production year. manufactured on or after September 1, trucks, buses, multipurpose passenger (2) Production of complying 2019. All hybrid and electric vehicles to vehicles, and low-speed vehicles subject vehicles— which this standard applies to the phase-in requirements of Each manufacturer shall report for the manufactured on or after September 1, § 571.141, S9.1 Hybrid and Electric production year being reported on, and

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00107 Fmt 4701 Sfmt 4700 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2 90522 Federal Register / Vol. 81, No. 240 / Wednesday, December 14, 2016 / Rules and Regulations

each preceding production year, to the Hybrid and Electric Vehicles (49 CFR Issued on November 10, 2016 in extent that vehicles produced during the 571.141). Washington, DC, under authority delegated preceding years are treated under in 49 CFR 1.95 and 501.5. Standard No. 141 as having been § 585.134 Records. Mark R. Rosekind, produced during the production year Each manufacturer shall maintain Administrator. being reported on, information on the records of the Vehicle Identification [FR Doc. 2016–28804 Filed 12–13–16; 8:45 am] number of vehicles that meet the Number for each vehicle for which BILLING CODE 4910–59–P requirements of Standard No. 141, information is reported under § 585.133 Minimum Sound Requirements for until December 31, 2023.

VerDate Sep<11>2014 18:14 Dec 13, 2016 Jkt 241001 PO 00000 Frm 00108 Fmt 4701 Sfmt 9990 E:\FR\FM\14DER2.SGM 14DER2 sradovich on DSK3GMQ082PROD with RULES2