The Long Duration New Horizons (NH) Mission to Pluto, Its

IAC-04-A.8.01

NEW HORIZONS MISSION TO PLUTO/CHARON: REDUCING COSTS OF A LONG-DURATION MISSION

Alice F. Bowman Johns Hopkins University Applied Physics Laboratory, Laurel, MD (USA) [email protected]

Albert A. Chacos, Christopher C. DeBoy, R. Michael Furrow, Karl E. Whittenburg Johns Hopkins University Applied Physics Laboratory, Laurel, MD (USA) [email protected], [email protected], [email protected], [email protected]

ABSTRACT

The long-duration and long light-time delay of NASA’s planned New Horizons mission to Pluto, its moon Charon, and the extended mission to one or more Kuiper Belt Objects poses unique challenges to mission operations, especially in this time of limited space exploration budgets. A number of courses of action can be followed to reduce wear on Observatory hardware, reduce operations staffing costs, and reduce Deep Space Network usage and costs without sacrificing the health and safety of the Observatory or risking the successful completion of the primary mission. Major components in this system are an autonomy subsystem that can react quickly enough to safe the Observatory when it is out of contact with the ground station; the use of a beacon to indicate the health of the Observatory during the dormant phases of the mission; command loading and verification strategies to accommodate the long light-time delays; and the combining of operations personnel to take advantage of similarities of Observatories, supporting ground station setup, and procedures. When planned early in the mission development phase, these components are easily integrated into the operations concept and Observatory hardware to form a co- hesive plan to mitigate cost and risk. Cost reduction measures for long-duration missions, such as those planned for New Horizons, enable funding for a greater number of equally important space exploration missions from a limited space exploration budget.

1. BACKGROUND design, development, and mission operations are delegated to the Johns Hopkins University/Applied Physics Laboratory The New Horizons (NH) mission is part of (JHU/APL) in Laurel, MD, USA. The Tom- NASA’s New Frontiers Program. It was baugh Science Operations Center (TSOC), awarded in the fall of 2001, with a start date named for the discoverer of Pluto, would of January 2002 at a cost of less than serve as the center of scientific research $600M. Dr. S. Alan Stern of Southwest Re- and data repository for this mission. search Institute (SwRI) in Boulder, CO, Throughout this paper, the term “Observa- USA, is the mission’s Principal Investigator tory” is defined as the integrated spacecraft (PI) and is responsible for the overall NH and science instrument payloads. mission. Responsibilities of Observatory

1 TIMEFRAME The primary mission would be to conduct a MISSION (nominal- OBJECTIVE detailed first reconnaissance flyby of the PHASE planned) Pluto/Charon system, with observations KBO 1 80 days pre and Science Observa- planned to begin 150 days prior to and 60 (extended 90 days post KBO tion Final Rehears- days after closest approach (C/A). The ex- mission) 1 C/A als, Navigation & tended mission would be to conduct a flyby Targeting, Science Collection, Data of one or more Kuiper Belt Objects (KBOs). Return The NASA decision to fund the extended Cruise 4 91 days post KBO Beacon-Hibernation, mission would be made sometime after (extended 1 C/A to 81 days Annual Checkouts, launch of the NH Observatory. Table 1 gives mission) pre KBO 2 Precession Maneu- vers, Navigation & an overview of the currently planned NH Targeting mission phases. KBO 2 80 days pre and Science Observa- (extended 90 days post KBO tion Final Rehears- The NH Observatory primary launch window mission) 2 C/A als, Navigation & Targeting, Science would span 35 days from January 11 to Feb- Collection, Data ruary 14, 2006, giving C/A arrivals from July Return 2015 to July 2020. If launched in the first 23 days of the launch window, a Jupiter Gravity Assist (JGA) would be used to reduce flight time to Pluto/Charon by as much as 5 years; if launched in the last 12 days, the trajectory would be a direct flight to Pluto/Charon. Fwd LGA (+Y) Launch would be from Cape Canaveral, FL, MGA (+Y) 2.1m HGA USA, using an Atlas 551 with a Star 48B PEPSSI upper stage. Figure 1 depicts the current F-8 RTG configuration of the NH Observatory. SWAP

Table 1 NH Mission Phase Overview Thrusters TIMEFRAME MISSION Sun Sensors (nominal- OBJECTIVE PHASE planned) Heat Shield Launch & Launch + 60 days Observatory (Obs.) Thrusters Early Opera- (Jan–Mar 2006) Checkout, Naviga- LORRI tions (LEOps) tion & Targeting SDC Cruise 1 Cruise following Instrument Commis- PERSI/Alice Star Trackers LEOps to 61 days sioning, Flight Tests, PERSI/Ralph before Jupiter C/A Navigation & Target- (Mar–Dec 2006) ing, etc. Figure 1 NH Observatory Current Con- Jupiter 60 days pre to 40 JGA, Navigation & figuration days post JGA Targeting, Jupiter (Jan–Apr 2007) Science Cruise 2 41 days post JGA Beacon-Hibernation, to 201 days pre Annual Checkouts, Pluto/Charon C/A Precession Maneu- (Apr 2007–Jan vers, Navigation & 2015) Targeting, Re- hearsal 2. INTRODUCTION Pluto/Charon 200 days pre to Science Observa- 14 days post tion Final Rehears- Pluto/Charon C/A als, Navigation & The NH mission operations team (MOps) is (Jan–Jul 2015) Targeting, Science expected to face a number of unique opera- Collection, First- tional challenges as a result of the long- Look Data Return duration and light-time delay of this mission. Pluto/Charon 15 days to 270 Return of Recognition and mitigation of these chal- Data Retrieval days post Pluto/ Pluto/Charon Sci- Charon C/A (Aug ence Data, Naviga- lenges began from the onset. To help ad- 2015–Apr 2016) tion & Targeting dress the long duration and the subsequent Cruise 3 271 days post Beacon-Hibernation, Deep Space Network (DSN) usage and (extended Pluto/Charon C/A Annual Checkouts, costs, a beacon-hibernation phase would be mission) to 81 days pre Precession maneu- KBO 1 C/A vers, Navigation & implemented for the cruise between JGA Targeting and Pluto/Charon C/A denoted as Cruise 2.

2 To address the long light-time delay that will gency telemetry. Dormant periods would be reach approximately 4.5 hours one way at when the Observatory is in a passive spin- Pluto/Charon, an autonomy safing strategy hibernation (PS-H) state with no active G&C and a modification to nominal command control. The majority of time during Cruise 2 load and verification would be implemented. would be spent in PS-H with the operations Both of these mitigations would result in team relying upon the beacon tone to indi- changes to the standard staffing concept of cate the health of the Observatory. In the the mission operations team and would be current concept, the Observatory could be done with minimal risk to the primary mis- placed in PS-H for up to 11 months at a sion objectives occurring, at the earliest, 9.5 time. Cruise 2 would last approximately 7.5 years after launch. years, assuming a 2015 Pluto/Charon arri- val. 3. MISSION DURATION 3.1.1 Active Periods Because of the long duration of this mission and the resulting relative cost of DSN sup- Active periods would comprise about 2 port as compared with the total mission cost, months of each year. Planned periods of as well as the number of spacecraft compet- Observatory activity would include preces- ing for the DSN stations, it was decided dur- sion maneuvers, TCMs, and annual check- ing the concept study phase that DSN costs outs. Precession maneuvers would be con- and usage would be minimized in order for ducted to maintain MGA pointing to Earth for the proposal to be considered by NASA. beacon tone transmission/reception and One way of reducing costs is to reduce the emergency commanding at 7.8 bps. The number of required DSN passes by placing high gain antenna (HGA), MGA, and forward the Observatory into a beacon-hibernation low gain antenna (LGA) would be co-aligned mode. Deep Space 1 first demonstrated the (Figure 1). Another precess to the edge of feasibility of using a set of beacon tones to the MGA would be conducted before the indicate a spacecraft’s health.1 During the pointing drifts outside of the HGA deadband, Deep Space 1 technical demonstration, one (Figure 2). These maneuvers would require of four beacon tones was transmitted for the DSN 70-m antenna support of one 8-hour short periods between telemetry contacts. pass per day for a week, for a total of ap- Based on these results, a beacon- proximately 504 hours of 70-m antenna hibernation concept was developed for the time. Preliminary analysis indicates that ap- NH mission, utilizing eight beacon tones proximately nine precession maneuvers (four each on two carriers), with one “green” would be required outside of the annual and seven “red” tones, each indicating a checkout periods. While TCMs require DSN specific Observatory state of health. support of one 8-hour pass per day for a week, all needed TCMs would be planned to 3.1 Cruise 2 occur during the annual checkout periods.

The mission operational concept for Cruise 2 (the phase between Jupiter and Pluto/Charon) is to have “active periods” and n co “dormant periods”. Active periods would be ea Precess when the Observatory is in either an active B GA Angle spin or a three-axis stabilized state, meaning M that Guidance and Control (G&C) is controlling the Observatory attitude. (See the Appendix for a discussion of NH Observa- H tory modes and states.) The active periods GA Cmd would include planned precession maneuvers to orient the medium gain antenna (MGA) to Earth, trajectory correction maneuvers (TCMs), annual checkouts, and Figure 2 NH Precession Angles response to “red” beacon tones or emer-

3 The annual checkout period would be de- requiring attention, the autonomy rule facility voted to accessing the NH Observatory sub- (ARF) would initiate one of seven “red” bea- system and instrument states of health, ob- con tones, disable the preloaded weekly taining navigation data to support any “green” tone, and broadcast the “red” tone needed TCMs and the next period of PS-H, continuously. In certain cases, the ARF and performing routine maintenance on des- would command the Observatory to go the ignated subsystems and instruments. The Active Spin-Earth Acquisition (AS-EA) safe first annual checkout period would occur state, in which case telemetry would be within 6 months after the end of the Jupiter broadcast at the emergency rate (10 bps) phase. Each of the eight planned checkouts instead of a “red” beacon tone, and a 14-day spans 50 days. DSN contacts would vary “demote to Active Spin–Sun Acquisition (AS- from two to three 8-hour passes per week SA) state” timer would be initiated. Dormant except for the one annual checkout that in- periods will comprise on average approxi- cludes the Pluto/Charon rehearsal. At the mately 10 months each year and require end of the checkout period, the MGA-to- 281 hours of 34-m and 182 hours of 70-m Earth pointing would be precessed to the DSN beacon support over the 7.5-year pe- edge of the pointing deadband. The mission riod of Cruise 2. operations team would load the time tag commands to broadcast the weekly “green” 4. LONG LIGHT-TIME DELAY beacon tone corresponding to the already scheduled weekly DSN beacon contacts and The NH mission end of life would be realized the time tag commands to “wake up” the at a solar distance of 50 AU. Analysis early Observatory for precession maneuvers, in the concept study phase determined that TCMs, and annual checkout periods. Each probability was high that two KBOs with di- nominal annual checkout period would util- ameters greater than 50 km could be ize approximately sixteen 8-hour 70-m DSN 2 encountered within 50 AU. As a result, the passes (or 128 hours), while the one annual MGA is being sized to provide emergency checkout period that contains the first commanding out to 50 AU. Primary mission Pluto/Charon rehearsal is currently esti- objectives would occur at approximately 32 mated to require 392 hours of 70-m DSN AU, with the extended mission objectives support time. This gives a total of 1288 (first KBO encounter) predicted at 40 AU. hours of 70-m support over Cruise 2 to sup- One Way Light Time (OWLT) delays would port all annual checkouts. be 4.3 hours and 5.3 hours respectively. Due to these long OWLT, Observatory saf- 3.1.2 Dormant Periods ing and command loading and verification concepts would be modified from missions During the beacon-hibernation phase of with shorter OWLTs. Cruise 2, the Observatory would be placed in PS-H with the “green” beacon tone broad- 4.1 Observatory Safing cast weekly for a 24-hour period centered around the planned DSN contact. Each During all cruise phases of this mission ex- weekly DSN beacon contact would be cept for the cruise from Earth to Jupiter scheduled for 1.5 hours, although it is ex- (Cruise 1), the Observatory would be placed pected to take much less time for the DSN in a PS-H state for long periods, relying to receive and analyze the tone (less than upon the ARF to determine the state of 45 minutes). These measures would be health and report this state to the NH mis- taken to cover any DSN or operations con- sion operations center (MOC) via a beacon tingency that might arise. For distances of or telemetry. The ARF would report in three less than 25 AU, the 34-m stations would be different ways, “green” beacon tone, “red” used to support beacon contacts. At dis- beacon tone, or emergency telemetry by tances of 25 AU and beyond, use of the 70- commanding the Observatory to AS-EA. m stations would be necessary to detect the Many components would be turned off dur- beacon tone. The “green” tone will not be ing this phase, partly to reduce wear and broadcast continuously to maximize the life- partly to reduce risk. time of the traveling wave tube’s cathode filament. In the case of an onboard failure

4 Within the ARF, conditions would be defined Consultative Committee for Space Data that merit a “red” beacon tone. Definition of Systems (CCSDS) Command Operations the conditions triggering a “red” beacon tone Procedure 1 (COP 1) protocol for TCTF up- will be chosen carefully to maximize safety link verification. The Observatory Command and minimize false or non-mission threaten- and Data Handling subsystem (C&DH) will ing conditions that do not require operations accept the TCTF and its command contents intervention. Because there are seven “red” after performing validity checks that include beacon tones, if a “red” tone were received, a check of the TCTF sequence number. The the mission operations team would have a entire TCTF would be rejected if any good indication of the type of onboard frame/command check fails or the sequence anomaly just from the particular tone re- number is not the one expected. The failure ceived. of one telemetry frame would cause a re- transmit request to be sent to the ground Because DSN telemetry contacts would not and all subsequent telemetry frames to fail be scheduled during the dormant periods of until the expected sequence number is re- the cruise phases, it will be imperative that ceived. To mitigate this potential failure and the ARF request for a telemetry contact be to increase the chances that the first attempt done only when needed to avoid impacting to transmit the TCTFs to the NH Observa- the DSN schedule. The seven “red” tones tory is successful, the ground station soft- would be divided into priority categories, ware would be modified to include the ability with not all “red” tones requiring immediate to send each TCTF up to “n” times. The NH ground intervention. When ARF places the Observatory C&DH would accept only the Observatory in AS-EA, ground intervention first valid expected TCTF and would ignore must occur within 14 days or the ARF would other TCTFs with the same sequence num- command the Observatory to AS-SA and ber. When OWLTs are short, TCTFs will broadcast the highest priority “red” tones. routinely be sent one time. As the OWLTs increase, the mission operations team would 4.2 Command Loading and have the ability to increase the number of Verification times the TCTFs are sent. It is expected that no more than n = 3 will be used operation- ally. Early in the mission, tests would be per- The nominal uplink rate for the NH Observa- formed to verify the sending and receipt of tory will be 500 bps, with the capability of a TCTFs at n > 1. 2000-bps rate when needed. It is expected that at 2000 bps, the longest command load Command load verification will be accom- would take 15 minutes to radiate. 70-m plished by loading the command and time downlink rates for the Observatory would be tag macros in a disabled state, commanding variable, ranging from about 37 kbps at Jupi- a dump of the onboard memory containing ter to about 1000 bps at Pluto/Charon. While the load and verifying the dump against the these downlink rates limit telemetry data expected image on the ground. After ground return, especially at Pluto/Charon and be- verification, commands would be sent to yond, the largest telemetry delay would be enable the macros and to enable the load due to the OWLT. For this reason, ways to transition autonomy rule, which would allow decrease the number of OWLTs required for activation of the new load when the current loading commands on the Observatory were load completes. This process requires a de- explored. lay of three OWLTs plus an OWLT to verify the “command load enable” status. When Nominally, the operations center ground OWLTs reach a pre-determined duration, station would send each telecommand the command loading and verification strat- transfer frame (TCTF) once to the DSN for egy would be modified so that the command upload to the Observatory. Each TCTF load, verification, and enabling time is de- would contain command or time tag macros. creased to one OWLT plus an OWLT to ver- Command macros contain individual com- ify the “command load enable” status, and mands; time tag macros contain the execu- optimally would not span more than one 8- tion time for each command macro. NH hour DSN pass. would implement a modified version of the

5 With the modified strategy, the command pending on the current staffing need and and time tag macros would be loaded with a Observatory events. In some cases this is ground-calculated checksum. As part of the problematic and results in the requirement C&DH acceptance of the command macro, that staff be experts in more than a few dis- an onboard checksum would be performed ciplines in order to minimize staffing cost. and compared to the ground-calculated checksum uploaded with command macro. 6. DSN USAGE AND COST Only if the checksums match would the command macro be enabled. Furthermore, SAVINGS the load transition autonomy rule would allow activation of the new load only if all Both the 34-m and the 70-m DSN antenna macros in the loaded range were reporting resources are heavily subscribed, and any “enabled” (i.e., they have all passed the reduction in use would help ease this strain. checksum compare.) At the end of the DSN This is especially true of the 70-m antenna pass or the beginning of the next pass, the resources. Unfortunately, the distances of command load “enabled” status is verified this mission require heavy reliance on the via telemetry. The mission operations team 70-m resources to get the downlink rates to also would have the ability to delay the support telemetry and data retrieval in a dump of the on-board memory containing timely manner. After Jupiter, all DSN teleme- the command load so that it corresponds to try passes would be run with the 70-m an- the next scheduled DSN pass. When this tennas. By incorporating a beacon, the mis- strategy is followed, the dump would be sion operations team would not have to rely compared with the expected ground image, upon a DSN telemetry pass to ascertain the and the command to enable the transitional health of the Observatory; the receipt of a autonomy rule would be sent during this beacon tone via a DSN beacon pass would second pass. This would result in the mis- give this information. Also, these DSN bea- sion operations team having three strategies con passes would make use of the 34-m to perform a command load and verification antennas out to 25 AU, thereby reducing this choosing the strategy that best fits the mission’s 70-m antenna requirements. Cou- OWLT delay and mission phase. ple these 70-m resource savings with the fact that a DSN beacon pass would take much less time (conservatively estimated at 5. MISSION OPERATIONS 1.5 hours) than scheduling one or more CENTER STAFFING weekly 8-hour DSN telemetry pass, and the DSN 70-m antenna usage would be de- The lengthy periods spent with the Observa- creased by approximately 2274 hours. tory in PS-H during Cruise 2 would allow for a reduction in mission operations center 7. SUMMARY staffing. An average of 5 operations center staff per month would be needed to support As currently planned, the NH mission would operations for the first 2 years after Jupiter be a long-duration mission with long and then would decrease to 3.5 operations OWLTs, both of which pose challenges to center staff per month to support the remain- the project budget and mission operations ing years of Cruise 2 before the team. Developing a concept of operations Pluto/Charon phase begins. This translates that meets these challenges early in the life into significant savings compared with an of the project would allow hardware and average of about 9.5 operations center staff software to be developed to support these per month needed to support a more tradi- concepts from the very beginning, saving tional mission concept. cost. Use of a beacon-hibernation mode supported by an autonomy system with em- Additionally, a staffing concept would be phasis on simplicity and safing the Observa- developed to allow sharing of staff among tory for long periods of time would save in Observatories with common onboard hard- DSN usage, cost, and staffing. Co-locating ware and ground systems. This allows a mission operations centers for Observato- pool of staff equally well versed in opera- ries with common onboard hardware and tions to move between Observatories de-

6 common ground systems would allow staff 3A-N mode/state would be the typical con- sharing between missions, decreasing costs figuration of the Observatory when in 3A for both projects. Developing a strategy for mode. It would be used for instrument com- command load verification that minimizes missioning and calibration, selected encoun- OWLTs would decrease DSN usage and ter science observations, and optical naviga- time devoted to command loading. In addi- tion. tion, the ability of the ground station to send telemetry frames more than once and the 3A-TCM mode/state would be entered via Observatory to accept the first valid frame MOps ground command for G&C controlled and wait for the next expected frame would TCMs and would be maintained only long add additional robustness to Observatory enough to complete the TCM. command loads. AS-N mode/state would be the typical state Many of the cost reductions would also add of the Observatory when in AS mode. Ex- robustness and lessen risk to the mission. amples of its planned use are science ob- Incorporating the beacon to support the servations not requiring 3A mode, checkout beacon-hibernation concept and the ARF to activities not requiring 3A mode, and pre- trigger the appropriate beacon tone when in cessions to maintain the HGA to Earth. PS-H would reduce DSN usage and operations center staff. It also would reduce risk AS-TCM mode/state would be entered via to the mission, as it would not be possible to MOps ground command for G&C controlled maintain real-time contact with the Observa- TCMs and would be maintained only long tory given the long light-time delays. Using enough to complete the TCM. shared staff between common Observato- ries and ground stations also would lessen AS-EA mode/state would be entered after risk, as it would allow a larger pool of staff the occurrence of Observatory faults such during inactive periods on both Observato- as low voltage sensing (LVS) or a loss of ries, increasing team depth more than would communications. After 14 days, if no contact normally be possible. has been made with the Observatory or if internal reference is lost for 24 hours, the 8. APPENDIX autonomy rule facility (ARF) would time out and would command the Observatory to AS- A brief description of the NH Observatory SA mode/state. modes and states is provided for clarifica- tion. AS-SA mode/state would be entered only when the Observatory cannot point to Earth The NH Observatory would have three (loss of inertial reference capability) or after modes of operation (three axis [3A], active an extended loss of communications (14 day spin [AS], and passive spin [PS]) and six in AS-EA mode). states (trajectory correction maneuver [TCM], normal [N], hibernation [H], encoun- PS-N mode/state would be the typical state ter [E], Earth acquisition [EA], and Sun of the Observatory while in PS mode. This acquisition [SA]). There would be a total of mode/state would be used to transition the ten mode/state combinations, since not all Observatory out of PS-H and verify that the states are allowed in every mode, as shown Observatory subsystems are nominal before in Table 2. transitioning to PS-TCM or an AS mode. In PS-N, G&C would be on and operating in a 3A-E mode/state would be used for instru- passive state, i.e., not actively controlling the ment pointing and scanning for science ob- spin axis. It also would be used during non- servations. Autonomous exiting of this state active periods in Cruise 1 to save propellant. is allowed only with the expiration of a long- term backup timer (currently envisioned to PS-TCM mode/state would be entered via be 14 days). Due to the propellant required MOps ground command and would be used to maintain this state, it is used only when for MOps initiated and controlled TCMs, required. spin-ups, and spin-downs. This mode/state

7 would be maintained only long enough for 9. ACKNOWLEDGMENTS the completion of these activities. The authors gratefully acknowledge the PS-H mode/state would be the mode/state support and dedication of all the scientists used for journeys from Jupiter to Pluto (if on and engineers on the New Horizons team a JGA trajectory) or from Launch + 1 year to whose efforts result in the realization of this Pluto (if on the Pluto-direct trajectory), from concept of operation. Special mention is Pluto to KBO 1, and from KBO 1 to KBO 2. made of the Deep Space Mission System Most subsystems would be powered off (in- (DSMS) personnel supporting the New Hori- cluding the G&C subsystem), and the pro- zons program at NASA’s Jet Propulsion pulsion system latch valves would be Laboratory. Furthermore, we acknowledged closed. The Observatory would remain in that it is due to the unyielding dedication and PS-H mode/state except during the annual support of the Mission Principal Investigator, checkout periods and for those precession Dr. S. Alan Stern, that this mission is well on maneuvers occurring outside of the annual its way to becoming the first mission to checkout period. Pluto/Charon despite various impediments along the way. Table 2 NH Observatory Modes and States State Mode 10. REFERENCES Active Passive 3-Axis Spin Spin TCM 3A- AS-TCM PS-TCM 1. DeCoste et al., “Deep Space 1 Technol- TCM ogy Validation Report – Beacon Monitor Normal 3A-N AS-N PS-N Operations Experiment,” Jet Propulsion Hibernation Not Not PS-H allowed allowed Laboratory, California Institute of Technol- Encounter 3A-E Not Not ogy. February 2000. allowed allowed Earth Not AS-EA Not 2. J. Spencer, M. Buie, L. Young, Y. Guo, Acquisition allowed allowed Sun Not AS-SA Not and A. Stern 2003. “Finding KBO Flyby Tar- Acquisition allowed allowed gets for New Horizons,” Earth, Moon and Planets 92, 483-491.