Bridge Linking Engineering and Society

Winter 2004

The BRIDGE LINKING ENGINEERING AND SOCIETY

Cool Robots: Scalable Mobile Robots for Deployment in Polar Climates Laura R. Ray, Alexander D. Price, Alexander Streeter, Daniel Denton, and James H. Lever The Challenges of Landing on Mars Tommaso Rivellini The Future of Engineering Materials: Multifunction for Performance-Tailored Structures Leslie A. Momoda Modeling the Stuff of the Material World: Do We Need All of the Atoms? Rob Phillips Capturing and Simulating Physically Accurate Illumination in Computer Graphics Paul Debevec Spatial Audio Reproduction: Toward Individualized Binaural Sound William G. Gardner

Promoting the technological welfare of the nation by marshalling the knowledge and insights of eminent members of the engineering profession. The BRIDGE

NATIONAL ACADEMY OF ENGINEERING

Craig R. Barrett, Chair Wm. A. Wulf, President Sheila E. Widnall, Vice President W. Dale Compton, Home Secretary George Bugliarello, Foreign Secretary William L. Friend, Treasurer

Editor in Chief (interim): George Bugliarello Managing Editor: Carol R. Arenberg Production Assistant: Penelope Gibbs The Bridge (USPS 551-240) is published quarterly by the National Academy of Engineering, 2101 Constitution Avenue, NW, Washington, DC 20418. Periodicals postage paid at Washington, DC. Vol. 34, No. 4, Winter 2004 Postmaster: Send address changes to The Bridge, 2101 Constitution Avenue, N.W., Washington, DC 20418. Papers are presented in The Bridge on the basis of general interest and time- liness. They reflect the views of the authors and not necessarily the position of the National Academy of Engineering. The Bridge is printed on recycled paper. © 2004 by the National Academy of Sciences. All rights reserved.

A complete copy of each issue of The Bridge is available in PDF format at http://www.nae.edu/TheBridge. Some of the articles in this issue are also available as HTML documents and may contain links to related sources of information, multimedia files, or other content. The Volume 34, Number 4 • Winter 2004 BRIDGE LINKING ENGINEERING AND SOCIETY

Editor’s Note 3 Cutting-Edge Research for a Changing World Pablo G. Debenedetti

Features 5 Cool Robots: Scalable Mobile Robots for Deployment in Polar Climates Laura R. Ray, Alexander D. Price, Alexander Streeter, Daniel Denton, and James H. Lever Low-cost mobile robots can advance scientific research on the Arctic plateau. 13 The Challenges of Landing on Mars Tommaso Rivellini Each generation of landing technology addresses the challenges posed by the previous generation. 18 The Future of Engineering Materials: Multifunction for Performance-Tailored Structures Leslie A. Momoda Multifunctional materials are emerging as a new interdisciplinary field. 22 Modeling the Stuff of the Material World: Do We Need All of the Atoms? Rob Phillips From a model-building perspective, the goal is to “make things as simple as possible, but no simpler.” 28 Capturing and Simulating Physically Accurate Illumination in Computer Graphics Paul Debevec Someday we may be able to make a photoreal computer model of anything—no matter what it is made of or how it reflects light. 37 Spatial Audio Reproduction: Toward Individualized Binaural Sound William G. Gardner Sound is inherently a spatial perception.

NAE News and Notes 43 NAE Newsmakers 44 NAE President Honored by ASME

(continued on next page) The BRIDGE

44 2004 Annual Meeting 45 Chairman’s Remarks 47 President’s Address 51 Remarks by Eli Ruckenstein, 2004 Founders Award Recipient 53 Remarks by John Brooks Slaughter, 2004 Bueche Award Recipient 55 Opportunities and Challenges for Engineering Education as Seen from an Interdisciplinary Telecommunications Master’s Degree Program, by Frank S. Barnes 61 Call for Nominations for 2005–2006 Awards 62 Report of the Home Secretary 63 New Program Officer for the Committee on Engineering Education 63 Education Research Scholar Joins CASEE 64 CASEE Fall Student Intern 64 Janice Tsai, NAE Intern 65 Tenth U.S. Frontiers of Engineering Symposium 66 Calendar of Meetings and Events 66 In Memoriam

67 Publications of Interest

The National Academy of Sciences is a private, nonprofit, self- The Institute of Medicine was established in 1970 by the National perpetuating society of distinguished scholars engaged in scientific Academy of Sciences to secure the services of eminent members of and engineering research, dedicated to the furtherance of science and appropriate professions in the examination of policy matters pertaining technology and to their use for the general welfare. Upon the author- to the health of the public. The Institute acts under the responsibility ity of the charter granted to it by the Congress in 1863, the Academy given to the National Academy of Sciences by its congressional char- has a mandate that requires it to advise the federal government on ter to be an adviser to the federal government and, upon its own scientific and technical matters. Dr. Bruce M. Alberts is president of the initiative, to identify issues of medical care, research, and education. National Academy of Sciences. Dr. Harvey V. Fineberg is president of the Institute of Medicine.

The National Academy of Engineering was established in 1964, The National Research Council was organized by the National under the charter of the National Academy of Sciences, as a parallel Academy of Sciences in 1916 to associate the broad community of organization of outstanding engineers. It is autonomous in its adminis- science and technology with the Academy’s purposes of furthering tration and in the selection of its members, sharing with the National knowledge and advising the federal government. Functioning in Academy of Sciences the responsibility for advising the federal gov- accordance with general policies determined by the Academy, the ernment. The National Academy of Engineering also sponsors engi- Council has become the principal operating agency of both the neering programs aimed at meeting national needs, encourages edu- National Academy of Sciences and the National Academy of Engi- cation and research, and recognizes the superior achievements of neering in providing services to the government, the public, and the engineers. Dr. Wm. A. Wulf is president of the National Academy scientific and engineering communities. The Council is administered of Engineering. jointly by both Academies and the Institute of Medicine. Dr. Bruce M. Alberts and Dr. Wm. A. Wulf are chair and vice chair, respectively, of the National Research Council. www.national-academies.org 3 Editor’s Note

Cutting-Edge Research in the sky-crane system that will be used in the 2009 for a Changing World Mars Science Laboratory mission. The session closed with Kent Joosten’s talk on accessing the lunar poles in Every year, NAE sponsors future human exploration. Recent robotic and earth- a Frontiers of Engineering based studies suggest that the lunar poles may offer (FOE) Symposium, which advantages in terms of thermal conditions, availability brings together some of solar energy, and access to resources. 100 outstanding, competi- Kristi Anseth and Diann Brei organized the session tively selected, young (ages on designer materials. Greg Carman explained the 30–45) engineering leaders development and potential uses of thin films of materi- from academia, industry, and als that exhibit energy coupling. Thin-film active mate- Pablo G. Debenedetti is the Class government laboratories for rials are still in their scientific infancy but may someday of 1950 Professor in Engineering three days of sharing ideas be used as sensors and actuators in applications ranging and Applied Science at Princeton and learning about cutting- from drug delivery to microfluidics. Leslie Momoda, University and an NAE member. edge research on a broad whose paper is published here, described ongoing range of engineering topics. research on performance-tailored structures that can Modern engineering is characterized by globalization, adapt their performance or morphology on demand and rapid technological change, and the crossing of interdis- the multifunctional materials that underlie such sys- ciplinary boundaries, and the emerging engineering lead- tems. Jennifer West discussed recent advances in vas- ers who attend FOE symposia represent a wide spectrum cular tissue engineering. Biomimetic strategies, of backgrounds, interests, and talents. The event offers a unique opportunity for them to learn about the fron- including genetic modification of vascular cells and pul- tiers in engineering areas other than their own. Six of satile stressing of smooth muscle cells, offer hope that the papers from this year’s symposium are included in this novel substitutes for blood vessels may be fabricated for issue of The Bridge. For the past two years, it has been my use in coronary artery bypass graft surgery, where treat- privilege to chair the FOE organizing committee, which ment options are severely limited. selects the speakers and topics for the symposium. Multiscale modeling, that is to say, the computational The tenth FOE Symposium was held September 9–11, analysis of systems with structures and dynamics that 2004, at the Beckman Center in Irvine, California. The span many length and time scales, was the topic of the program encompassed four themes: engineering for third session, organized by Dimitrios Maroudas and extreme environments, designer materials, multiscale Grant Heffelfinger. Yannis Kevrekidis discussed the modeling, and engineering and entertainment. “equation-free” modeling of complex systems, whereby The opening session, organized by Mary Kae Lock- information from microscopic (e.g., atomic-level) wood and John Weatherly, focused on engineering for dynamics is used, not to derive macroscopic equations, extreme environments. Laura Ray, whose paper appears but to perform coarse-grained computational experi- in this issue, described the engineering challenges in ments that probe the system’s behavior over macro- designing scalable mobile robots for deploying instru- scopic lengths and times. Rob Phillips’ presentation ment networks on the Antarctic plateau, an ideal loca- focused on the computational modeling of complex tion for studying the upper atmosphere at high magnetic problems, such as protein/DNA interactions. In his latitudes. John Berkoe illustrated the role of simulation paper (published here), he gives examples of cases where and modeling in extreme engineering projects through an all-atom approach is inadequate because it generates three examples, including the Tacoma Narrows Bridge an enormous amount of data without providing a con- mooring system and the conceptual design of the Cher- comitant increase in knowledge. His elegant solution nobyl New Safe Confinement. Tommaso Rivellini involves maintaining the microscopic physics only to (paper published in this issue) traced the evolution of the extent it is needed. Adam Arkin’s talk addressed landing technology over the last 40 years, culminating biological models of events that occur on different time The 4 BRIDGE scales, from single-enzyme kinetics to the evolutionary from year to year. The 2004 program featured presenta- time scale. He described how improvements in mea- tions by representatives of three government agencies surement technology and the abundance of genetic- (Carey Schwartz of DARPA, Kenneth Harwell of the sequence information are beginning to link disparate U.S. Department of Defense, and Doug Stetson of time scales. Bjorn Stevens discussed the computational NASA’s Advanced Planning and Integration Office) challenges to simulating the climate system arising from that fund cutting-edge research. Their descriptions of interactions between small-scale processes, such as current projects elicited an excellent discussion and aerosol formation, that occur over the micrometer scale, audience participation. and the planetary scale. FOE symposia traditionally have an evening speaker, The relationship between engineering and entertain- and this year was no exception. Alex Singer, a film ment was the subject of the fourth session, chaired by director who has directed more than 280 television Chris Kyriakakis. Paul Debevec (paper published here) shows and five theatrical features during his 40-year discussed the state of the art in the simulation of illumi- career, spoke about the integration of technology and nation in computer graphics. The accuracy of the sim- entertainment. In his lively talk, entitled “Unlikely ulation depends on re-creating the way light permeates Partners: DARPA and Me,” he described an ongoing a given scene, as it is reflected by some objects and trav- DARPA-funded short-film project envisioning a future els through others. William Gardner addressed the of augmented cognition in the year 2030. challenges of accurately reproducing the spatial proper- I know of no meeting as interdisciplinary, diverse, and ties of sound, as well as progress toward the development stimulating as FOE, and I hope the six papers included of individualized binaural technology. His paper is also in this issue convey some of the excitement we experi- published in this issue. enced in September at the 2004 symposium. The technical talks are always followed by extended, lively Q&A sessions with broad audience participation. The program surrounding this basic framework changes Low-cost mobile robots can advance scientific research on the Arctic plateau.

Cool Robots: Scalable Mobile Robots for Deployment in Polar Climates

Laura R. Ray, Alexander D. Price, Alexander Streeter, Daniel Denton, and James H. Lever

Laura R. Ray Alexander D. Price Alexander Streeter Daniel Denton James H. Lever

The Antarctic plateau is a unique location to study the upper atmosphere at high magnetic latitudes because it provides a stable environment for sensi- tive instruments that measure interactions between the solar wind and the earth’s magnetosphere, ionosphere, and thermosphere. Existing stations on the edge of the continent and at the South Pole, and six low-power (50 W) automatic geophysical observatories, have demonstrated the value of distributed, ground-based observations of solar-terrestrial physics. Increasing the spatial density of these observations offers great scientific opportunities. A National Research Council report, The Sun to the Earth and Beyond, empha- sizes the need for mobile instrument networks. The report recommends

Laura R. Ray is associate professor of engineering, Alexander D. Price is a master of engineering management candidate, Alexander Streeter is an M.S. candidate, and Daniel Denton is an A.B. candidate at the Thayer School of Engi- neering, Dartmouth College, Hanover, New Hampshire. James H. Lever is a mechanical engineer with the U.S. Army Cold Regions Research and Engineering Laboratory, Hanover, New Hampshire. The 6 BRIDGE

“comprehensive new approaches to the design and main- long periods of time in drifting snow and will require tenance of ground-based, distributed instrument net- a stable environment with low vibration and electro- works, with proper regard for the severe environments in magnetic noise. Robots and deployed sensors should be which they must operate” (NRC, 2002). retrievable with high reliability to minimize environ- This paper describes scalable mobile robots that can mental impact and cost. In this paper, related robotics enable the deployment of instrument networks in research is summarized, including mobile robot design Antarctica. The drivetrain, power system, chassis, and concepts for polar environments, technical challenges navigation algorithms can be scaled for payloads of associated with their development, and enabling tech- roughly 5 to 25 kilograms. One can envision deploying nologies for cost-effective mobile robots. robots from the South Pole to locations on the plateau for long-term or short-term observation and retrieving State-of-the-Art Robots for Extreme or repositioning the network through iridium-based Environments communication. Potential missions include: deploy- NOMAD, a gasoline-powered robot for polar and ing arrays of magnetometers, seismometers, radio desert environments, was developed at Carnegie Mel- receivers, and meteorological instruments; measuring lon University (Apostolopoulos et al., 2000; Carnegie disturbances in the ionosphere through synchro- Mellon University, 2004b). NOMAD, which is 2.4 × nization of Global Positioning System (GPS) signals; 2.4 × 2.4 m in size and weights 725 kg (Figure 1), can using ground-penetrating radar (GPR) to survey travel up to 50 centimeters per second (cm/s) and crevasse-free routes for field parties; and conducting deploy instruments, such as a magnetometer. In 1997, glaciological surveys with GPR. Robot arrays could NOMAD executed its first mission in the Atacama also provide high-bandwidth communications links Desert of southern Chile, traversing 223 km through and power systems for field scientists. teleoperation. Subsequently, NOMAD successfully Based on this concept, a single robot is under con- found and classified five indigenous meteorites on an struction that would carry a triaxial fluxgate magne- Antarctic mission. For our purposes, however, tometer, an iridium modem, and a modest set of weather NOMAD’s large size, high cost, and nonrenewable fuel instruments. The magnetometer will be a payload test are disadvantages. In addition, its deployment experi- case. Magnetometer arrays are already used in low ence suggests that navigation cameras may work poorly latitudes and midlatitudes, but polar regions provide in polar climates due to reflection of sunlight off of unique windows for observing the effects of solar wind the snowfield. on the Earth’s magnetosphere. With mobile networks, Spirit and Opportunity are Mars exploration rovers sensor locations could potentially be tuned to events in developed by the National Aeronautics and Space the magnetosphere. Also, synchronized data from polar Administration Jet Propulsion Laboratory. Each 2.3 × networks could potentially discover spatial characteris- 1.6 × 1.5-m rover weighs 174 kg and has a top speed of tics of narrow-band spectral features in geomagnetic field data, identify magnetospheric boundaries, and refine models accordingly (Lanzerotti et al., 1999). The deployment of a remote observatory on the Antarctic plateau via transport and small aircraft is expensive and entails dangerous takeoffs and landings at remote sites. For large-scale, widely distributed (>500 km radius) networks, relatively low-cost mobile robots could substantially reduce per-instrument deployment costs. Semiautonomous network deployment would also free limited aircraft and human resources for other missions. The harsh weather of polar environments, long-range requirements, navigation issues, and variable terrain pose significant design challenges for inexpensive FIGURE 1 NOMAD gasoline-powered robot. Source: Carnegie Mellon University, unmanned vehicles. Instruments would be deployed for 2004b. WINTER 2004 7

5 cm/s (NASA/JPL, 2004). The power source for Mars rovers, a multipanel solar array and two rechargeable lithium-ion batteries, enables the rover to generate 140 W of power for four hours per sol, when the panels are fully illuminated. The warm electronics box, which contains the batteries, electronics, and computer, can only operate in the range of –40°C to +40°C. Gold- painted, insulated walls, solid silica aerogel, thermostat and heaters, and a heat rejection system protect the body from the 113°C temperature swing during the Martian day. The payload includes a panoramic camera, a miniature thermal emission spectrometer, a Mössbauer spectrometer, an alpha-particle X-ray spectrometer, and a rock abrasion tool. Each of the six wheels is driven by its own in-wheel motor, and the two front and two rear wheels have steering motors for point turns. The rovers are well suited for their mission on Mars but are too expensive for the deployment of instrument networks. Hyperion (Figure 2), designed for sun-synchronous FIGURE 2 Hyperion rover. Source: Carnegie Mellon University, 2004a. exploration, is a 157-kg, 2 × 2.4 × 3-m vehicle that includes a 3.45-m2 nearly vertical solar panel; its maxi- during the austral summer. Key design issues are out- mum speed is 30 cm/s (Carnegie Mellon University, lined below. 2004a; Wettergreen et al., 2001). The chassis is inten- Figure 3 shows a satellite photo of Antarctica. The tionally simple—a 1.5 N-m, 150 W brushless DC motor vast central plateau covers more than five million combined with a harmonic drive for an 80:1 reduction square kilometers of relatively flat, crevasse-free terrain. ratio drives a wheel through a bicycle chain. A passively A second large area of operation is the Ross Ice Shelf. articulated steering joint provides two free rotations, Generally, Antarctic snowfields consist of dense, wind- enabling moderate maneuverability and mechanical blown snow. Aside from wind-sculpted sastrugi, dune- simplicity. This design has many appealing features for like features that are identifiable on satellite imagery, instrument-network deployment, including renewable there are few obstacles. The central plateau receives fuel, simplicity, and potentially low cost. less than 50 mm precipitation (<500 mm snowfall) Commercial all-terrain robots made by iRobot and in an average year. During summer months at the ActivMedia weigh 39 to 100 kg and carry payloads South Pole, wind speed averages 2 m/s (Valenziano and of 7 to 100 kg. Powered by two DC servomotors and a 4-wheel differential-drive system, these battery- operated robots run for 2 to 6 hours at speeds between 1 and 2 m/s. With no navigation instruments or soft- ware, these robots cost from $7,000 to $22,000. They could potentially be retrofitted for solar-power operation, but they are not designed for low-temperature operation, and the solar panels alone could comprise the entire payload budget.

Cool Robot Concept Our robot is designed to operate in interior Antarc- tica, which is characterized by low snowfall, moderate winds, and extreme cold. We envision networks of robots, guided by GPS and onboard sensors, that are launched and retrieved from the South Pole Station FIGURE 3 Satellite photo of Antarctica. Source: USGS, 2004. The 8 BRIDGE

Dall’Oglio, 1999), and the five-year maximum speed is 20.5 m/s (CMDL, 2004). The average daily temperature is –20°C to –40°C. An Antarctic robot must traverse firm snow and occa- sional softer drifts, sustain mobility in windy conditions, have a minimal environmental impact, and operate in temperatures down to –40°C. We envision a lightweight, solar- powered, wheeled robot that can be transported in a Twin Otter aircraft and is capable of traversing 500 km within two weeks. After reaching a target location, the robot could collect data over a period of two to FIGURE 4 Daily average horizontal insolation at the South Pole, 2002. Source: CMDL, 2004. three months before return- ing to the South Pole for the winter. The design includes a low center-of-gravity vehi- in the dense snow of the Antarctic plateau, sinkage cle with four direct-drive brushless electric motors, an should be small. The estimated total resistance of 0.25 for enclosed thermally controlled volume for instrumenta- a 90-kg vehicle will require a net traction force of 221 N. tion and batteries, and a solar panel “box” for renewable Travel of 500 km in two weeks will require an average energy. Table 1 provides design specifications for a speed of 0.41 m/s, with an average power requirement of wheeled robot and a price point for economic viability 90 W and maximum power of 180 W for the top speed of for the deployment of networks of such robots. 0.8 m/s. Allowing up to 40 W for housekeeping power Motion resistance in snow is attributable to sinkage and power-system efficiencies, the target power budget is and varies with the firmness of the snow pack immedi- approximately 220 W. ately in front of the wheel, the length of the tire in con- Despite a harsh climate and low sun angles, Antarc- tact with the snow, and the width of the tire (Richmond tica is an ideal place for a solar-powered robot. The et al., 1995). Given the target ground pressure (< 20 kPa) summer sun provides a 24-hr energy source, and the central plateau receives scant precipitation and infre- TABLE 1 Robot Specifications quent fog. The Antarctic plateau is nearly completely covered in snow, with albedo averaging 95 percent ≥ Maximum speed 0.80 m/s across visible and ultraviolet wavelengths (Grenfell et Mass (excluding payload) ≤ 75 kg al., 1994) and fairly uniform scattering in all directions (Warren et al., 1998). The high altitude and dry air ≥ Payload mass 15 kg block less incoming radiation, and there is a small ben- Ground pressure ≤ 20 kPa efit due to the proximity of the Earth to the sun during the summer. The sun remains at approximately the Operating temperature range 0°C to –40°C same elevation throughout the day (especially near the Dimensions ≤ 1.4 × 1.15 × 1-m South Pole), resulting in relatively constant energy input. With low elevation angles and significant Cost ≤ $20,000 reflected solar energy, nearly vertical solar panels will WINTER 2004 9

be optimal. Also, the efficiency of solar cells increases chassis enclosed by a five-panel box—along with pre- as temperature decreases. dicted panel capacities extrapolated from the model for Average horizontal insolation data for 2002 (Figure 4) nominal Antarctic solar radiation, 20 degree sun eleva- show a range of horizontal irradiance of 300 to 500 W/m2 tion, and 90 percent albedo. The panel outputs are at the South Pole (CMDL, 2004). Adjusting for sun ele- reported as a percentage of their standard capacities vation angle gives a net insolation between 800 W/m2 (rated at 1,000 W/m2 insolation). The front panel and 1,200 W/m2, which is consistent with earlier studies (directly facing the sun) has a capacity of 128 percent (Hanson, 1960). At other locations in Antarctica where (more than 100 percent due to reflected energy). Signi- cloud cover and fog are more frequent, the average inso- ficantly, the top and side panels contribute nearly as lation is about half this, but the sunny days are almost as much power as the panel facing the sun. Even the back bright. For comparison, consider a clear winter day in panel receives substantial radiation because the robot’s New England. At a sun elevation of 20 degrees, the total shadow is not as large as the area of snow that reflects insolation is between 600 and 800 W/m2. Table 2 sum- light to the panel. marizes insolation data at various locations on the Enabling technology for the robot is the affordable, 20 percent efficient, A-300 solar cell by Sunpower, Inc., TABLE 2 Insolation for Various Sun Conditions which became available in 2003. Figures 6 and 7 show predicted power available to the motors for a robot using Condition/Location Nominal Insolation 54 of these cells per panel (each cell is 12.5 × 12.5 cm) 2 Maximum for Antarctica, continent 1,200 W/m2 at 1,000 W/m insolation and 90 percent albedo. The resulting robot will fit in the Twin Otter cargo bay. 2 Average for South Pole, Nov–Feb 1,000 W/m These results include efficient maximum-power-point- Average for South Pole, at solstice 1,100 W/m2 tracking (MPPT) circuits for each panel and subtract housekeeping power. The robot can drive at full speed > 2 Average for south polar plateau 800 W/m even in below-average insolation. Under minimal inso- Average for Ross Ice Shelf 400 W/m2 lation, the robot still has enough power to drive slowly or charge the batteries and drive in short bursts on bat- January 2004 measurements, Hanover, New Hampshire 660 W/m2 tery power. Diffuse incoming radiation—light scattered by the atmosphere—is an unmodeled benefit because diffuse light is expected to be received perpendicular to Antarctic continent and elsewhere. The average solar the panels from all directions. The total diffuse radia- energy input during the November to February operating tion is expected to be 50 to 100 W/m2 providing 40 to window is approximately 1,000 W/m2, with an average 80 W, enough to drive the robot in bursts and maintain sun elevation of about 20 degrees. instrument operation (Hanson, 1960). To determine the optimum size of the solar panel, we A positive feature of the deployment of instruments developed a model to predict power as a function of solar insolation, sun eleva- Top: 34% tion, and azimuth for solar (direct sun only) panels in a snowfield. The model assumes diffuse reflection from the snow at Back: 11% Front: 128% (in shadow) a specified albedo. We vali- (direct + reflected) dated the model using data collected with a commercial Sides: 34% 20-W panel during January- (reflected light only) February 2004 in Hanover, New Hampshire. Figure 5 shows the resulting robot design concept—a wheeled FIGURE 5 Panel power capacities in nominal Antarctic sun. The 10 BRIDGE

(e.g., downed trees), is in force for short periods, after Power vs. Robot Angle to Sun which the hiker returns to 400 the path. In the initial stages of research, global 380 navigation is being used, 360 Sun Elevation: primarily through GPS and 23 degrees speed control, with sensors 340 to detect unbalanced wheel 20 degrees speeds and hence poten- 320 17 degrees tial problems with traction, and low-bandwidth path- 300 correction algorithms to Power Available to Motors 280 reduce “dither” around the 0102030405060708090 path. In the traversal of long distances, GPS-induced Angle to Sun, degrees (0 = facing sun) path deviations are toler- able. Traction control can FIGURE 6 Power availability as a function of sun elevation angle and the robot azimuth angle. be layered onto the basic global path-following algorithm, along with sensors for tilt and wheel slip. Local- mode navigation would be Power vs. Sun Elevation Angle invoked if sensors detect 500 extreme tilting or slippage. Navigation and motion 400 control also depend on the Robot at 0°* power system. The robot 300 Robot at 45° will move along the path 660 W/m2, 0 only when adequate solar 200 Full Speed power is available. In Half Speed cloudy conditions, the 100 robot will move under bat- * nominal sun of Power Available to Motors 1,000 W/m tery power if necessary to 0 prevent drifting in, and in 091827 windy conditions, the robot Sun Elevation (degrees above horizon) may face a direction that minimizes drag and the potential for tipping. We FIGURE 7 Power available and required at various sun angles. do not anticipate the robot having a vision system on the Antarctic plateau is relatively flat, straight paths because sastrugi are visible on satellite imagery and can on dense snow, which minimizes planning the com- be avoided through route selection. The lack of con- plexity of the path. One of the most promising navigation trast in snow-covered terrain would make onboard nav- architectures for such paths is “mixed-mode” operation, igation challenging and potentially expensive. which mimics human behavior (e.g., in hiking a known path over a long distance) (Simmons et al., 1995). The Design Embodiment and Enabling Technologies global objective is to stay on the path, but a local mode, We have attempted to minimize vehicle mass by in which the hiker goes around unanticipated objects using stiff honeycomb composites for the solar panels WINTER 2004 11

and chassis and custom-designed wheel rims and hubs. that can be configured for speed or torque control. We The enabling technologies and cost-design trade-offs tested a single motor-gearbox combination in a cold and an estimate of the costs of parts and the mass for the room in a box insulated as configured on the robot. The prototype robot are highlighted below. results showed that both long-term operation and start- Because of their newness, complete panels for A-300 stop operation would be efficient, and controllability solar cells are not yet available. Moreover, traditional would be maintained at cold temperatures. panel construction with its steel backing is not viable The power-system architecture includes three lithium- for the robot. Thus, we will construct the solar panels ion batteries in series and five solar panels, each of which in house, using quarter-inch honeycomb sandwich pan- can operate under varying insolation and temperature els (Nomex core with fiberglass facing). The cell will be conditions. To deliver power to a common power bus, encapsulated in silicone. Similar honeycomb compos- each panel requires an MPPT, a device that allows each ite will be used to make the chassis box (Figure 8). The panel to operate at the bus voltage established by the bat- construction and joinery of honeycomb panels are teries while meeting power demands for the motors. mature technologies in the aerospace field, and these Custom-designed MPPTs that weigh less than 250 grams materials supply area densities of 1.4 kg/m2 for solar pan- and have better than 97 percent efficiency have been els and 2.5 kg/m2 for chassis walls. designed and constructed. We estimate that the five-panel robot, without payload, will weigh ~73 kg, with a total material cost of less Motor controller Panel support than $15,000. The design is relatively insensitive to tube payload up to about 20 kg. The robot will be equipped Motor/gearhead with instruments to assess power-input and mobility models during field trials in Antarctica, anticipated for the 2005–2006 austral summer.

Conclusion Solar-powered mobile robots for operation on the Antarctic plateau are feasible in terms of power availability, mechanical design, and power-system design. Waypoint navigation on the relatively obstacle-free Batteries plateau through GPS can provide long-distance travel Payload space appropriate to the scientific missions envisioned. Solar panel Mobile robots capable of reliable, long-term operation support rib on the Antarctic plateau can potentially advance scien- FIGURE 8 Internal chassis and components. tific research through instrument deployment, mapping, and the provision of portable mobile power. The wheels are sized to provide adequate ground clearance and low rolling resistance. We considered many References different tires and trade-offs between traction, weight, Apostolopoulos, D., M.D. Wagner, B. Shamah, L. Pedersen, pressure, and rolling resistance. Taking into account K. Shillcutt, and W.L. Whittaker. 2000. Technology and availability and cost, a 16 × 6–8 ATV tire, which has low field demonstration of robotic search for antarctic mete- mass and excellent traction, was selected. Low-mass rims orites. International Journal of Robotics Research 19(11): and hubs are not available for ATV tires, however, so 1015–1032. these were custom designed and machined in house Carnegie Mellon University. 2004a. Hyperion: Sun Synchro- to meet the strength and deflection requirements for a nous Navigation. Available online at: http://www.ri.cmu.edu/ 90-kg robot. For a mass-produced robot, rims and hubs projects/project_383.html (June, 2004). could be cast or stamped at low cost. Carnegie Mellon University. 2004b. Robotic Antarctic High-efficiency, brushless motors with 90 percent Meteorite Search: The NOMAD Robot. Available online efficient gear trains and lubricant for operation at –50°C at: http://www.frc.ri.cmu.edu/projects/meteorobot/Nomad/ drive the wheels directly. Each motor has a controller Nomad.html#Mechanical (June, 2004). The 12 BRIDGE

CMDL (Climate Monitoring and Diagnostic Laboratory). N.H.: Cold Regions Research and Engineering Laboratory. 2004. Daily Average Horizontal Insolation at the South Available online at: http://www.stormingmedia.us/82/8273/ Pole. Available online at: http://www.cmdl.noaa.gov/ A827392.html. info/ftpdata.html (March, 2004). Simmons, R., E. Krotkov, L. Chrisman, F. Cozman, R. Good- Grenfell, T.C., S.G. Warren, and P.C. Mullen. 1994. Reflec- win, M. Hebert, L. Katraqadda, S. Koenig, G. Krish- tion of solar radiation by the Antarctic snow surface at naswamy, Y. Shinoda, W. Whittaker, and P. Klarer. 1995. ultraviolet, visible, and near-infrared wavelengths. Journal Experience with Rover navigation for lunar-like terrains. of Geophysical Research 99(D9): 18-669–18-684. Journal of Engineering and Applied Science 1: 441–446. Hanson, K.J. 1960. Radiation studies on the south polar USGS (U.S. Geological Survey). 2004. Satellite Image Map of snowfield. IGY Bulletin 31: 1–7. Washington, D.C.: Antarctica. Available online at: http://terraweb.wr.usgs.gov/ National Academy of Sciences. TRS/projects/Antarctica/AVHRR.html (June, 2004). Lanzerotti, L., A. Shona, H. Fukunishi, and C.G. Maclennan. Valenziano, L. and G. Dall’Oglio. 1999. Millimetre astro- 1999. Long-period hydromagnetic waves at very high geo- nomy from the high Antarctic plateau: site testing at Dome magnetic latitudes. Journal of Geophysical Research C. Publications of the Astronomical Society of Australia 104(A12): 8, 423. 16: 167–174. NASA/JPL (National Aeronautics and Space Adminis- Warren, S.G., R.E. Brandt, and P. O’Rawe Hinton. 1998. tration Jet Propulsion Laboratory). 2004. Spacecraft: Effect of surface roughness on bidirectional reflectance of Surface Operations: Rover. Available online at: Antarctic snow. Journal of Geophysical Research http://www.marsrovers.jpl.nasa.gov/mission/spacecraft_rover_ 103(E11): 25-789–25-807. energy.html (June, 2004). Wettergreen, D., B. Shamah, P. Tompkins, and W.L. Whit- NRC (National Research Council). 2002. The Sun to the taker. 2001. Robotic Planetary Exploration by Sun- Earth and Beyond: A Decadal Research Strategy in Solar Synchronous Navigation. In Proceedings of the 6th Inter- and Space Physics. Washington, D.C.: National Academy national Symposium on Artificial Intelligence, Robotics Press. and Automation in Space (i-SAIRAS 2001), Montreal, Richmond, P.W., S.A. Shoop, and G.L. Blaisdell. 1995. Cold Canada, June 2001. Regions Mobility Models. CRREL Report 95-1. Hanover, Each generation of landing technology addresses the challenges posed by the previous generation.

The Challenges of Landing on Mars

Tommaso Rivellini

People have been fascinated with the idea of exploring Mars since the very beginning of the space age. Largely because of the belief that some form of life may have existed there at one time, surface exploration has been the ultimate ambition of this exploration. Unfortunately, engineers and scientists discovered early on that landing a spacecraft on the surface of Mars would be one of the most difficult and treacherous challenges of robotic Tommaso Rivellini is the lead space exploration. mechanical engineer for Mars Upon arrival at Mars, a spacecraft is traveling at velocities of 4 to 7 kilo- Science Laboratory Entry, Descent, meters per second (km/s). For a lander to deliver its payload to the surface, 100 percent of this kinetic energy must be safely removed. Fortunately, and Landing, at the Jet Propulsion Mars has an atmosphere substantial enough for the combination of a high- Laboratory, California Institute of drag heat shield and a parachute to remove 99 percent and 0.98 percent Technology. respectively of the kinetic energy. Unfortunately, the Martian atmosphere is not substantial enough to bring a lander to a safe touchdown. This means that an additional landing system is necessary to remove the remaining kinetic energy. On previous successful missions, the landing system consisted of two major elements, a propulsion subsystem to remove an additional 0.002 percent (~50 to 100 meters per second [m/s]) of the original kinetic energy and a dedicated touchdown system. The first-generation Mars landers used legs to accomplish touchdown. The second generation of touchdown systems used The 14 BRIDGE air bags to mitigate the last few meters per second of The decreased stability because of the high center of residual velocity. The National Aeronautics and Space mass is exacerbated if a mission carries a large rover to Administration (NASA) is currently developing a the surface. Because of the rover’s configurational third-generation landing system in an effort to reduce requirements, it is typically placed on top of the lander. cost, mass, and risk while simultaneously improving per- The Soviet Lunokhod lunar landers (Figure 2) are an formance as measured by payload fraction to the surface excellent example of this type of configuration. and the roughness of accessible terrain. A second major challenge of the legged-landing architecture is ensuring safe engine cutoff. To prevent the Legged Landing Systems guidance and control system from inadvertently destabi- The legs of the 1976 Viking mission lander represent lizing the lander during touchdown, contact sensors have the first-generation landing system technology (Pohlen been used to shut down the propulsion system at the et al., 1977). Basic landing-leg technology was devel- moment of first contact. On sloped terrain, this causes oped for the lunar Surveyor and Apollo programs in the the lander to free fall the remaining distance, which can early 1960s. In conjunction with a variable-thrust liq- significantly increase the total kinetic energy present at uid propulsion system and a closed-loop guidance and touchdown and, in turn, decrease landing stability and control system, legs represented an elegant solution increase mission risk. Implementation and testing of to the touchdown problem. They are simple, reliable fault protection for engine cutoff logic has been, and mechanisms that can be added to an integrated struc- continues to be, a difficult problem. ture that houses the scientific and engineering sub- The first in-flight problem associated with engine systems for a typical surface mission (Figure 1). shut off occurred on the lunar Surveyor lander mission when the propulsion system failed to shut off at touchdown, resulting in a significant amount of postimpact hopping. Fortunately, the terrain was benign, and the problem was not catastrophic. The second in-flight problem occurred on the Mars 98 lander mission when the engines were inadvertently shut off prematurely because of a spurious contact signal generated by the landing gear during its initial deployment. This problem resulted in a catastrophic loss of the vehicle. As a result, the Apollo missions all reverted to a man in the loop to perform engine shut off. A third major challenge with a legged landing system for missions with rovers is rover egress. Once the lander has come to rest on the surface, the rover must be brought to the surface. For legged landers, a ramped egress system is the most logical configuration. Because FIGURE 1 First-generation landing system used on the Viking lander, which landed rovers are bidirectional, the most viable arrangement on Mars in 1976. has been considered two ramps, one at the front and one at the rear of the lander. The Soviet Lunokhod The first challenge for a legged system is to enable the missions landed in relatively benign terrain, and in all lander to touch down safely in regions with rocks. For cases, both ramps were able to provide safe paths for the this the legs must either be long enough to raise the rover. In the Mars Pathfinder mission, one of the two belly of the lander above the rocks, or the belly of ramps was not able to provide a safe egress path for the the lander must be made strong enough to withstand Sojourner rover, but the second ramp did provide safe contact with the rocks. Neither solution is attractive. egress. For vehicles designed to explore a larger frac- Either the lander becomes top heavy and incapable of tion of the Martian surface and, therefore, land in more landing on sloped terrain or a significant amount of diverse terrain, combinations of slopes and rocks could structural reinforcement must be carried along for the conceivably obstruct or render useless the two primary remote chance that the lander will directly strike a rock. egress paths. WINTER 2004 15

air-bag testing were significant challenges for both the Mars Pathfinder and MER missions.

The Sky-Crane Landing System As Mars surface explo- rations mature, roving is becoming more important in the proposed mission architectures. The MER missions demonstrated the FIGURE 2 Lunokhod Soviet lunar rover leaving the legged lander that delivered it to the surface. value of a fully functional rover not reliant on the lan- Air-Bag Landing Systems der to complete its surface mission. In the 2009 Mars The second-generation landing system was developed Science Laboratory (MSL) and other future missions, for the Mars Pathfinder mission and subsequently the rover’s capabilities and longevity will be extended. improved upon for the Mars Exploration Rover (MER) Future missions are also being designed to access larger missions (Figure 3). These second-generation systems areas of the planet and, therefore, will require more have a combination of fixed-thrust solid rocket motors robust landing systems that are tolerant to slope and and air bags to perform the touchdown task. The solid rock combinations that were previously considered too rocket motors, which are ignited two to three seconds hazardous to land or drive on. The third-generation prior to impacting the surface, slow the lander down to landing system, the sky-crane landing system (SLS), a stop 10 meters above the surface, from an initial veloc- currently being developed for the MSL mission, will ity of approximately 120 meters/second. The lander is directly address all of the major challenges presented by then cut away from the over-slung rockets and free falls for the remaining distance. The air-bag system, which was developed to reduce cost and increase landing robustness, is designed to provide omnidirectional protection of the payload by bouncing over rocks and other surface hazards. Because the system can also right itself from any orientation, the challenge of stability during landing has been completely eliminated. Because the lander comes to rest prior to righting itself, the challenge of rock strikes has been reduced to strikes associated with the righting maneuver, which are significantly more benign. The challenge of thrust termination, in this case cutting the lander away from the rockets, remains but has been decoupled from the problem of landing stability. The problems of rover egress were addressed systemati- cally on the MER missions; a triple ramp-like system provided egress paths in any direction, 360 degrees around the lander. Although the air-bag landing system has addressed some of the challenges and limitations of legged landers, it has also introduced some challenges of its own. FIGURE 3 Second-generation landing system used on the Mars Pathfinder and Horizontal velocity control using solid rockets and MER landers. The 16 BRIDGE the first- and second-generation landing systems. It will coupled with an active guidance and control system also eliminate the problem of rover egress. controls the velocity and position of the vehicle. At SLS eliminates the dedicated touchdown system and 35 meters, the sky-crane landing maneuver is initiated, lands the fully deployed rover directly on the surface of and the rover is separated from the propulsion module. Mars, wheels first. This is possible because the rover is The rover is lowered several meters as the entire system no longer placed on top of the lander. In the SLS, the continues to descend. The two-body system then propulsion module is above the rover, so the rover can descends the final few meters to set the rover onto the be lowered on a bridle, similar to the way a cargo heli- surface and cut it away from the propulsion module. The copter delivers underslung payloads (Figure 4). propulsion module then performs an autonomous The landing sequence for future missions will be sim- fly-away maneuver and lands 500 to 1,000 meters away. ilar to the Viking mission, except for the last several The central feature of the SLS architecture is that the seconds when the sky-crane maneuver is performed propulsion hardware and terrain sensors are placed high (Figure 5). After separating from the parachute, the SLS above the rover during touchdown. As a result, their follows a Viking-lander-like propulsive descent profile in operation is uninterrupted during the entire landing a one-body mode from 1,000 meters above the surface sequence. One important result of this feature is that down to approximately 35 meters above the surface. the velocity control of the whole system is improved, During this time, a throttleable liquid-propulsion system and, therefore, the rover touches down at lower velocity. Thus, there is no last-meter free fall associated with engine cutoff, and, because dust kick-up is minimal, the radar antennas can continue to operate even while the rover is being set down on the surface. The lower impact velocity has two effects. First, the touchdown velocities can now be reliably brought down to the levels the rover has already been designed for so it can traverse the Martian surface. Second, the low velocity, coupled with the presence of bridles until the rover’s full weight has been transferred to the surface, results in much more stability during landing. Because the rover does not have to be protected from the impact energy at landing and because there is no need to augment stability at landing, there is no longer a need for a dedicated touchdown system. This, in turn, eliminates the need for a dedicated egress system. The SLS takes advantage of the fact that the rover’s mobility system is inherently designed to interact with rough, sloping natural terrain. Rovers are designed to have high ground clearance, high static stability, reinforced belly pans, and passive terrain adaptability/ conformability. These are all features of an ideal touchdown system.

Touchdown Sensing Touchdown sensing can be done in several ways. The simplest and most robust way is to use a logic routine that monitors the commanded up-force generated by the guidance and control computer. The landing sequence is specifically designed to provide a constant descent velocity of approximately 0.75 m/s until touchdown has FIGURE 4 Sky-crane landing system shown with the rover already deployed. been declared. Prior to surface contact, the commanded WINTER 2004 17

FIGURE 5 Sky-crane landing sequence showing the three main phases and events. Nominal touchdown velocities are 0.75 m/s vertically and 0.0 m/s horizontally. up-force is equal to the mass of the rover plus the mass of Conclusion the descent stage (which are roughly equal) times the As Mars explorers have learned the hard way, it’s not gravity of Mars. During the touchdown event, the com- typically the fall that kills you, it’s the landing. Landing manded up-force fluctuates depending on the specific technology has matured significantly in the 40 years geometry of the terrain. since NASA began exploring extraterrestrial surfaces. Once the weight of the rover has been fully trans- Each generation of landing technology has attempted to ferred to the surface of Mars, the commanded up-force resolve the challenges posed by the previous generation. takes on a new steady-state value equal to the mass of The SLS represents the latest stage in that evolution. the descent stage times the gravity of Mars, approximately one-half of its pretouchdown magnitude. The Acknowledgment system declares touchdown after the new lower com- The research described in this paper was carried out manded up-force has lasted for at least 1.5 seconds. This at the Jet Propulsion Laboratory of the California Insti- approach provides an unambiguous touchdown signa- tute of Technology under a contract with the National ture without the use of dedicated sensors. Aeronautics and Space Administration. The fly-away phase of the landing sequence is initiated when touchdown has been declared. During the fly- Reference away phase, separation of the rover is accomplished by Pohlen, J., B. Maytum, I. Ramsey, and U.J. Blanchard. 1977. the pyrotechnic cutting of the bridle and umbilical lines The Evolution of the Viking Landing Gear. JPL Technical connecting the rover and descent stage. The descent Memorandum 33-777. Pasadena, Calif.: Jet Propulsion stage then uses its onboard computer to guide the Laboratory. propulsion module up and away from the rover and land it several hundred meters away. Multifunctional materials are emerging as a new interdisciplinary field.

The Future of Engineering Materials: Multifunction for Performance- Tailored Structures

Leslie A. Momoda

In the future, new functional and reduced-scale materials that are currently in the forefront of technology will be hybridized into designer materials that can perform dramatic “tailorable” functions in large engineered systems. These performance-tailored structures will have the ability to change or adapt the performance or style of a structure on demand. Today, engineers can imagine designing adaptive flight profiles from morphing aircraft-wing Leslie A. Momoda is director of the structures; comfort-tailored performance, such as active structural vibration Sensors and Materials Laboratory and noise suppression or temperature compensation, from louvered or pore- at HRL Laboratories, LLC, in Mal- based “smart skins”; energy-efficient structures, such as tropical-plant- inspired solar structures; adaptive structures that can compensate for ibu, California. distortion or heal themselves; and structures reconfigured to satisfy style pref- erences. Imagine, for example, being able to commute to work in a stately professional car that can be reconfigured into a sportier car for the weekend. As system-operating scenerios become more constrained by space and logistics limitations, the ability to adapt a structure’s performance at will is becoming increasingly attractive. Currently multi-mission objectives are met with multiple structures (i.e., one car to drive to work and another one for the weekend). These solutions work if there is excess capacity in the system (e.g., a two-car garage), but as the number of mission objectives increases, the procurement, storage, and maintenance of a large number of structures become prohibitive. As a consequence, engineered subsystems WINTER 2004 19

that provide structural adaptability are under development in several programs, such as the Defense Advanced Research Pro- Type I Type II Type III jects Agency (DARPA) Phases are mounted, Phases are distinct, Phases are highly Morphing Aircraft Struc- coated, or laminated with one phase integrated with indistin- tures Program (Wax et al., onto one another. embedded in another. guishable materials. 2003), the General Motors FIGURE 1 Schematic representations of the three types of multifunctional materials. Source: Matic, 2003. Autonomy Concept (Burns et al., 2002), and many structural health-monitoring programs. All of these scales as different “types” (Figure 1). Type I material is programs are designed to provide tailored performance comprised of phases in which one function is simply in large multicomponent system structures. mounted, coated, or laminated to another, usually a Researchers are now thinking about how these same structural component. Type II materials are comprised functionalities can be achieved in materials used to con- of distinct phases in which one function is embedded in struct the system themselves, for example, a thin, smart- another, usually a structural component. Type III mate- material skin that undergoes a radical but controlled rials are truly integrated; the phases are intermeshed, change under mechanical strain; a coating that changes blurring the physical distinctions between them. The color on demand; a shell that reconfigures its shape to true promise of multifunctional materials for perfor- meet styling or performance criteria. These materials mance-tailored structures is most likely to be realized by would enable the same system-level goals that are cur- Type III materials. rently designed as subsystems, but could be more easily Multifunctional materials are designed for improved integrated into larger engineered structures because they overall system performance. Thus, their performance will be lighter, smaller, less difficult to interface, and eas- metrics are inherently different from their single com- ier to maintain than current subsystems. Driven by ponent phases, in which improvement of a single func- recent advances in biomaterials and nanotechnology, tion, such as electrical conductivity, mechanical strain multifunctional materials are emerging as a new inter- or force, or energy density, is maximized or minimized. disciplinary field that promises to provide a new level of A multifunctional material requires a new design functionality, adaptability, and tailorability for future methodology in which system-level performance is engineered systems. emphasized over the optimization of individual functions. Optimizing system-level performance involves Multifunctional Material Systems optimization methodologies that are not commonly A multifunctional material is typically a composite or used in materials science. hybrid of several distinct material phases, in which each Frontiers in Multifunctional Materials Technology phase performs a different but necessary function, such as structure, transport, logic, or energy storage. Because Materials Technology each phase of the material performs an essential func- The goal of multidisciplinary research on multifunction, and because there is little or no parasitic weight or tional materials, much of it under the auspices of the volume, multifunctional materials promise more DARPA Synthetic Multifunctional Materials Program weight-efficient, volume-efficient performance flexibil- (Christodoulou and Venables, 2003), is to demonstrate ity and potentially less maintenance than traditional weight and volumetric efficiencies and performance multicomponent brass-board systems. The finer and enhancements. Most research is focused on integrating more distributed the integration scale in the material, two functions, usually a transport and structural func- the faster and more autonomous the reaction times. tion, and typically with low interconnectivity (e.g., Multifunctionality in a material can be integrated on Type I or Type II). Many of these studies rely heavily several dimensional scales with increasing interconnec- on inherently two-phase structural materials, such as tivity between phases and engineering difficulty as the fiber composites, laminates, foams, and other porous scale decreases. Matic (2003) has categorized these structures, as the matrix for multifunction. Even at The 20 BRIDGE this early stage, however, system-level benefits have hybrid laminate (Figure 3), which can produce a two- been noted. orders-of-magnitude change in stiffness, essentially from Structural batteries that reduce weight and complex- the stiffness of rubber to the hardness of steel, to accom- ity by integrating energy storage directly into the load- pany a shaping actuation force (G.P. McKnight, per- bearing structure have been developed by several teams sonal communication). using fibers, laminate, and nanotube construction. The Self-healing composite structures are also under energy density of the storage medium, such as a battery development (Chen et al., 2002). In these studies, a or supercapacitor, is reduced as the result of the incor- second phase, such as an adhesive or toughening agent, poration of less conductive structural materials. How- is added into the structure to ensure resealing on impact. ever, the decrease in parasitic structure results in an Materials that can integrate other functions into struc- overall weight savings and, therefore, improved energy tures, such as electromagnetism, thermal management, density for the system as a whole. and optics capability, are also being investigated. The integration of actuation or sensing mechanisms into tailorable structural materials, which is essential for Optimization and Computational Design mechanical reconfigurability and structural morphing, Advances have been made in the development of is another area under investigation. Research using optimization tools for designing integrated multifunc- metallic foams or highly engineered mesostructured tional materials. Sigmund and Torquato (1999) have materials (dos Santos e Lucato et al., 2004) and elas- done extensive work on topological optimization meth- tomeric polymers (Pei et al., 2002) has shown tremen- ods to determine the best morphological materials dous promise for producing large structural motions architectures to optimize performance from a highly along with the capacity to integrate actuation for feed- integrated Type III material embedding of very dissimi- back and control. Research conducted at HRL Labo- lar physical mechanisms. Many functional combina- ratories, LLC, using hybrid, shape-memory materials tions, with as many as three phases, have been (Figure 2), has demonstrated the feasibility of using simulated. Although their work is purely theoretical, integrated actuating tendons in a variably stiff matrix to the results of their simulations have been validated by produce low-energy flexure during reconfiguration but the similarity of the optimized topological solutions to shape fixity for structural hold without energy when the microstructures and mesostructures found in biological motion is complete (W. Barvosa-Carter, personal com- systems. Macroscopic optimization tools for the design munication). This concept has been extended to a of less integrated Type I and II multifunctional materials have also been developed (Qidwai et al., 2002).

Challenges and Prospects for the Future Two-phase multifunctional systems show the promise of true materials integration. However, truly smart materials systems, analogous to biological systems, will require a combination of three or more functions, including logic, sensing, energy storage, a b structure, and actuation. Biological systems have per- FIGURE 2 Hybrid shape-memory materials approach to reconfigurable structures. (a) Shape-memory tendons actuate a variably stiff fected multifunction on a shape-memory polymer beam that can be flexed with low-input energy. (b) Upon cooling, the beam becomes stiff, locking shape small scale. With the deformations in place to handle exterior loads (here 500 g). design of a priori multiple WINTER 2004 21

troubleshoot a problem after design. In the new paradigm, however, the materials scientist must be actively involved from the inception of the system design, t providing a finely engineered material on the meso-, micro-, or nanoscale to meet the overall system goals. • n t This will require that the materials scientist be completely familiar with system design tools and computational tools, from the system scale to the micro- or nanoscale. In the future, the design of a new car, air- Constant Stiffness (Material 1) plane, or satellite will truly begin on the atomic scale.

Variable Stiffness (Material 2) References Burns, L.D., J.B. McCormick, and C.E. Barroni-Bird. 2002. Vehicles of change. Scientific American 287(4): 64–73. FIGURE 3 Variable stiffness laminate materials concept for producing large structural reconfigurations. Chen, X., M.A. Dam, K. Ono, A. Mal, H. Shen, S.R. Nutt, K. Sheran, and F. Wudl. 2002. A thermally re-mendable cross-linked polymeric material. Science 295(5560): functions into a materials system, these concepts will 1698–1702. be extended into large-scale structures. The complex- Christodoulou, L., and J.D. Venables. 2003. Multifunctional ities of these higher order systems will require a sophis- material systems: the first generation. JOM: The Member ticated understanding of how basic physical mechanisms Journal of The Minerals, Metals & Materials Society can be manipulated to create new, potentially less singly 55(11): 39–45. optimal means of achieving function and multivariable dos Santos e Lucato, S.L., J. Wang, P. Maxwell, R.M. McMeek- optimization tools. For example, to increase the ing, and A.G. Evans. 2004. Design and demonstration of mechanical strength of energy storage systems, either a high authority shape morphing structure. International the ionic conduction mechanism could be investigated Journal of Solids and Structures 41(13): 3521–3543. for new logic capabilities or electrical conductivity Matic, P. 2003. Overview of Multifunctional Materials. Pp. mechanisms could be investigated to determine how 61–69 in Smart Structures and Materials 2003, edited by they might influence mechanical strength. D.C. Lagoudas. Active Materials: Behaviors and Mechan- As our understanding of materials on the nanoscale ics. Proceedings of SPIE vol. 5053. Bellingham, Wash.: improves, we will be able to improve our control and SPIE. increase the range of physical properties of materials Qidwai, M.A., J.P. Thomas, and P. Matic. 2002. Design and even as we decrease the integration scale. Right now, Performance of Composite Multifunctional Structure- while we are on the verge of understanding and har- Battery Materials. Pp. 1–8 in Proceedings of the American nessing physics on the nanoscale, we have a tremendous Society for Composites Seventeenth Technical Confer- amount of work to do to learn how to fabricate large- ence, edited by C.T. Sun and H. Kim. Boca Raton, Fla: scale materials from nanoscale elements. Although self- CRC Press. assembly and biological processing techniques look Pei, Q., R. Pelerine, S. Stanford, R. Kornbluh, M. Rosenthal, promising, they are not yet mature enough for the fabri- K. Meijer, and R. Full. 2002. Multifunctional Elastomer cation of multicomponent systems. Rolls. Materials Research Society Symposium Proceedings New system-level design methodology for materials 698: 165–170. will change not only the tools a materials scientist must Sigmund, O., and S. Torquato. 1999. Design of smart com- know and understand, but also the role of the materials posite materials using topology optimization. Smart Mate- scientist in the system-design process. Typically, system rials and Structures 8(9): 365–269. designers choose from a toolbox of materials that have Wax, S.G., G.M. Fisher, and R.R. Sands. 2003. The past, pre- already been developed, and the materials scientist present and future of DARPA’s investment strategy in smart designs these materials to improve a single function. materials. JOM: The Member Journal of The Minerals, Often the materials scientist, who acts independently of Metals & Materials Society 55(11): 17–23. the design team, is present only to characterize data or From a model-building perspective, the goal is to “make things as simple as possible, but no simpler.”

Modeling the Stuff of the Material World: Do We Need All of the Atoms?

Rob Phillips

The advent of computers ushered in a new way of doing science and engineering in which a host of complex problems ranging from weather prediction to the microstructural evolution of multiphase alloys to the DNA/protein interactions that mediate gene regulation could be explored explicitly using computer simulation. Indeed, some say that the physical sciences are now based on a triumvirate of experiment, theory, and simulation, Rob Phillips is a professor of with simulation complementing more traditional techniques for under- mechanical engineering and applied standing problems involving many interacting degrees of freedom. One class physics at the California Institute of of problems for which simulation is increasingly important is associated with the understanding and control of materials. When we speak of materials, we Technology. mean “stuff” as diverse as the materials of which man is made (soft, squishy stuff) and technologies (stuff with desirable properties, such as strength or conductivity) (Amato, 1997). Clearly, the use of computation to understand and even design complex materials is one of the major challenges that will make it possible to replace the enlightened empiricism that gave rise to the great material ages (e.g., the Iron Age, Bronze Age, and silicon-based Information Age) with rational design. Similar roles are anticipated for simulation in many other fields as well. One of the flagship techniques for examining problems involving complex materials is molecular dynamics in which the microscopic trajectories of each and every atom are followed explicitly. Despite their promise, WINTER 2004 23

however, these simulations sometimes generate enor- problems currently being tackled in arenas ranging from mous quantities of information (i.e., terabyte data sets) molecular biology to atmospheric science are those in without necessarily delivering the promised con- which structures or processes at one scale influence the comitant increase in understanding. physics at another scale. As a response to these chal- Terabyte data sets engendered by simulations repre- lenges, modelers have begun to figure out how to con- sent a staggering quantity of information. A simple esti- struct models in which the microscopic physics is mate reveals that the entire 10 floors worth of books in maintained only where needed. One benefit of these the Caltech Millikan Library corresponds roughly to a approaches is that they not only reduce the computa- terabyte of information. More impressively, the tional burden associated with simulations of complex genomes of many viruses have an information content systems, but they also provide a framework for figuring that can be stored comfortably on a 256-megabyte mem- out which features of a given problem dictate the way ory stick alongside the genomes of even more complex the “stuff” of interest behaves. Several examples of this organisms from bacteria to yeast. Indeed, even organ- type of thinking are described below. isms as complex as humans have genomes that are much smaller than a terabyte. And yet our computers are overflowing with terabyte data sets, and worse yet, discussions of petabyte data sets are becoming routine. For “Multiscale modeling” is not example, a molecular dynamics calculation on a 100,000-atom system run for only 10 nanoseconds, woe- a fundamentally new idea. fully inadequate for accessing most materials processes, already generates a terabyte worth of data. Clearly, there is a mismatch between the quality of information Before embarking on a discussion of case studies, it is generated in our simulations and the information pre- worth discussing the metrics that might be used in sent in genomes and libraries. deciding whether or not a particular coarse-grained The question of how to build quantitative models of model should be viewed as a success. From the most complex systems with many interacting degrees of free- fundamental point of view, the job of theoretical mod- dom is not new. Indeed, one of the threads through the els is to provide a predictive framework for tying history of physics, the development of continuum theo- together a range of different phenomena. For example, ries, resulted in two compelling examples of this kind of in the case of elasticity described above, there are vast theory—elasticity and hydrodynamics. These theories numbers of seemingly unrelated problems (from flying share the idea of smearing out the underlying discrete- buttresses to the mechanical response of ion channels) ness of matter with continuum field variables. In addi- that may be brought under the same intellectual roof tion, with both theories, material properties are through reference to Hooke’s law. With elasticity the- captured in simple parameters, such as eleastic moduli or ory, we can predict how the cantilever of an atomic- viscosity, which reflect the underlying atomic-level force microscope will deflect when tugging on a protein interactions without specifically mentioning atoms. tethered to a surface. In this sense, elasticity theory has One lesson of these examples is that “multiscale mod- to be viewed as an unqualified success in the coarse- eling” is neither the exclusive domain of computational grained modeling of materials and shows just how high model building nor a fundamentally new idea. Indeed, the bar has been raised for multiscale models worthy in the deepest sense, the sentiment that animates all of the name. efforts at model building, whether analytical or computational, is of finding a minimal but predictive descrip- A Case Study in Multiscale Modeling: tion of the problem of interest. The Quasicontinuum Method One feature that makes problems like those described One of the computational responses to problems here especially prickly is that they often involve multi- involving multiple scales is multiresolution models that ple scales in space or time or both. An intriguing attempt to capture several scales at the same time. There response to the unbridled proliferation of simulated data has been great progress along these lines in recent years has been a search for streamlined models in which there from a number of different quarters, and presently we will is variable resolution. Many of the most interesting consider one example, namely the quasicontinuum The 24 BRIDGE method that permits the treatment of defects in crystalline solids.1 The main idea of the quasicontinuum method is to allow for atomic-level detail in regions where interesting physical processes, such as dislocation nucleation, dislocation intersections, and crack propagation are occurring, while exploiting a more coarse- grained description away from the key action. The motivation for the method is based on a recognition that when treating defects in solids there are both long-range elastic interactions between these defects and atomic- scale processes involving the arrangements of individual atoms. What makes these problems so difficult is that both the short-range and long-range effects can serve as Figure 1 Schematic illustration of a multiresolution mesh used to describe nano- equal partners in dictating material response. indentation of a crystalline solid. In the region just beneath the indenter, the mesh The numerical engine that permits a response to has full atomic resolution. In the “far fields,” the mesh is much coarser. problems of this type is finite elements that allow for nonuniform meshes and introduce geometric con- For a concrete example, consider a crystalline solid straints on atomic positions through the presence of subjected to external loading in the form of an inden- interpolation functions (so-called finite-element shape ter like the one shown in Figure 1. The quasicon- functions). Just as those of us who learned how to inter- tinuum philosophy is to discretize the system in such a polate on logarithm or trigonometric tables remember, way that there is full atomic resolution where the the key idea of the finite-element procedure is to char- action is (such as beneath the indenter) and a select, acterize the geometric state of the system by keeping representative subset of atoms that serve as nodes of the track of the positions of a few key atoms that serve as finite-element regions where all-atom resolution is sur- nodes on the finite-element mesh. The positions of all rendered. For the particular case of two-dimensional other atoms in the system can be found, if needed, by dynamical nanoindentation considered here, the cal- appealing to simple interpolation. culation involves a total of 5,000 nodes as opposed To simulate material response, geometry is not to the 107 atoms that would be needed in a full atom- enough. We not only have to know where the nodes istic calculation. This point is driven home in an even are, but also what forces act on them. To that end, the more compelling fashion in the case of a fully three- quasicontinuum method posits that the forces on the dimensional calculation for which the full atom- nodes can be obtained by appealing to interatomic istic calculation would have implicated in excess of potentials that describe interactions between individual 1011 atoms (Knap and Ortiz, 2003). atoms. Using the interpolated atomic positions, a The main point of this example is to reveal the kind neighborhood of atoms around each node is con- of thinking now being used to address complex prob- structed, and the energies and forces are then computed lems, such as material deformation. As exemplified by using standard atomistic techniques. This is an elegant the quasicontinuum method, the underlying micro- prescription because it ensures that the material scopic physics of bond stretching and bond breaking is response is strictly determined by the underlying micro- treated explicitly where needed, and only approxi- scopic physics without any ad hoc material assumptions. mately elsewhere. Once the geometric mesh has been constructed and the forces on the nodes computed, the simulation itself can The Problem of Living Materials take place by either minimizing the energy with respect Understanding the workings of living materials pre- to nodal coordinates or by using F = ma physics to com- sents even more compelling multiscale challenges than pute the trajectories of the system over time. those encountered in the traditional materials setting. Indeed, we are now realizing that almost no individual macromolecule in the living world acts alone. Rather, 1 There are many articles on the quasicontinuum method, but the inter- the cell can be viewed as “a collection of protein ested reader is invited to consult www.qcmethod.com, which has an extensive list of papers dealing with this method. machines,” assemblies of individual macromolecules WINTER 2004 25

(Alberts, 1998). One of the pressing challenges to have mediated by molecules, such as lac repressor, a protein arisen from the stunning successes of structural biology that sits on the DNA and prevents the genes respon- is the study of assemblies, such as viruses and the many sible for lactose digestion from being expressed. Lac “SOMES” (i.e., macromolecular assemblies that work in repressor binds to several sites in the vicinity of the concert to maintain the mass and energy budget of the promoter for the genes responsible for lactose digestion cell), such as nucleosomes, ribosomes, proteosomes, and and prohibits expression of those genes while simulta- assemblies that mediate gene expression, such as repli- neously creating a loop of DNA between the two somes, spliceosomes, and so on. repressor binding sites. Models of the function of assemblies such as From a modeling perspective, a minimal description “SOMES,” will remain out of reach of traditional of this system involves the DNA molecule itself, RNA atomic-level techniques for the foreseeable future. Con- polymerase, lac repressor, and an activator molecule sider the process of translation mediated by the ribo- called CAP. The kinds of questions that are of interest some. Even if we very generously assume that a new from a quantitative modeling perspective include the amino acid is added only once every millisecond, a mol- extent of gene expression as a function of the number of ecular dynamics simulation of translation would have copies of each molecular actor in this drama, as well as to be run for 1012 time steps for the addition of even a the distance between the DNA binding sites for lac single amino acid to the nascent polypeptide. The num- repressor and other features. ber of atoms (including the surrounding water) engaged in this process is well in excess of 100,000, implying a whopping 1017 positions corresponding to all of the atoms during the entire molecular dynamics trajectory. Multiscale methods are taking Similar estimates can be made for the workings of many other macromolecular assemblies that mediate the root in the biological setting. processes of a cell. All of these estimates lead to the same general conclusion—that even as we continue to pursue atomic-level calculations, we must redouble our One recent multiscale attempt to simulate the inter- efforts to understand the workings of “SOMES” from a action between DNA and lac repressor uses a mixed coarse-grained perspective. atomistic/continuum scheme, in which the lac repressor So the hunt is on to find methods of modeling and the surrounding complement of water molecules are processes of biological relevance involving assemblies of treated in full atomistic detail while the looped DNA diverse molecular actors, such as proteins, lipids, and region is treated using elasticity theory (Villa et al., DNA, without having to pay the price in excessive data 2004). The advantage of this approach is that it per- of all-atom simulation. One way to guarantee a rich mits the DNA to present an appropriate boundary con- interplay between experiment and models is through dition to the lac repressor simulation without having to the choice of case studies that are well developed from do a full atomistic simulation of both DNA and the pro- the standpoint of molecular biology and for which we tein. Figure 2 shows an example of the simulation box have compelling quantitative data. and the elastic rod treatment of DNA. The key point of One example of great importance is the lac operon, this example is not to illustrate what can be learned which has served as the “hydrogen atom” of gene regu- about the lac operon using mixed atomistic-continuum lation. This gene regulatory network, which controls methods, but to illustrate how multiscale methods have the digestion of the sugar lactose in bacteria, has been begun to take root in the biological setting, just as they the cornerstone of the development of our modern have in the conventional materials setting. picture of gene regulation. An intriguing history of this A second scheme, even more coarse-grained than the episode in the history of molecular biology can be found multiscale simulations of the lac repressor, is a statistical in the books of Judson (1996) and Echols (2001). mechanics treatment of molecular decision makers, such The basic idea is that only when a bacterium is as the repressor and its activator counterpart CAP. The deprived of glucose and has a supply of lactose does relevant point is that all of the atomic-level specificity the bacterium synthesize the enzymes needed to digest is captured by simple binding energies that reflect the lactose. The “decisions” made by the bacterium are affinity of these molecules for DNA and for each other. The 26 BRIDGE

This statistical mechanics perspective is a natural model must be by its ability to make predictions quantitative counterpart to the cartoons describing gene about as-yet undone experiments. The outcome of the regulation used in classic texts of molecular biology. As so-called “thermodynamic models” (Ackers et al., 1982) shown in Figure 3, these cartoons depict various states described above is a predictive framework that charac- of occupancy of the DNA in the neighborhood of the terizes the extent to which genes are expressed as a func- site where RNA polymerase binds. The statistical tion of concentrations of the relevant decision-maker mechanics perspective adds the ability to reckon explic- molecules (i.e., the transcription factors), the distance itly the statistical weights of each distinct state of occu- between the looping sites on the DNA, etc. Paradoxi- pancy of the DNA. From these statistical weights, a cally, as a result of the great successes of structural biol- quantitative prediction can be made of the probability ogists in determining the atomic-level structures of that a given gene will be expressed as a function of the important complexes, such as DNA and its binding number of molecules of each species. partners, we are now faced with the challenge of elimi- Ultimately, one of the primary ways of judging a nating molecular details in models of their function.

Summary The critical question for building models of the material world is the extent to which we can suppress an atom-by-atom description of the function of materials. As emergence of “multiscale modeling” reveals, even with increasing computational power, a host of important problems remain out of reach of strictly brute-force approaches. In the analysis of the material world, whether of the complex, rigid metallic structures used to construct cities or the soft, squishy materials that make up the organisms that populate them, the key action often takes place at the level of individual atoms—whether of a bond breaking at a crack tip or the active site of an enzyme. Many of the atoms that are part of these processes are interlopers, however, that seem to be little more than passive Figure 2 Illustration of a mixed atomistic/continuum description of the interaction of lac repressor protein with DNA. The lac repres- observers that provide sor molecule is shown in dark gray, the part of the DNA treated explicitly is shown in medium gray, and the part of the DNA treated boundary conditions for the via continuum mechanics is shown as a ribbon. Water molecules surrounding the protein are shown in light gray. Source: Courtesy atoms actively involved in of Klaus Schulten and Elizabeth Villa. the process of interest. WINTER 2004 27

ε Activation as a function of A for different ap (KBT)

−β∆ε e PD ε RNAP Activation

−β∆ε e AD ε AD ε AP

−β(∆ε +∆ε +ε ) e PD AD AP (b)

ε ε AD PD (a)

Figure 3 a. Schematic illustration showing the relation between cartoon models of various states of the genetic network and their corresponding weights in the statistical mechanics framework. b. Graph showing activation as a function of the number of activator molecules. The three curves correspond to different strengths for the interaction between the RNA polymerase and the activator.

From a model-building perspective, the goal is to References “make things as simple as possible, but no simpler” (i.e., Ackers, G.K., A.D. Johnson, and M.A. Shea. 1982. Quanti- to eliminate as many molecular details as possible). tative model for gene regulation by l phage repressor. Pro- In my opinion, this is one of the key design criteria for ceedings of the National Academy of Sciences 79(4): multiscale models. Although multiscale computational 1129–1133. models are receiving most of the effort and attention Alberts, B. 1998. The cell as a collection of protein right now, I believe the hunt should continue for analytic machines: preparing the next generation of molecular models that can capture the key features of complex biologists. Cell 92(3): 291–294. materials and lead to the kind of insight that can be dis- Amato, I. 1997. Stuff: The Materials the World Is Made Of. cussed at a blackboard. New York: Basic Books. Echols, H. 2001. Operators and Promoters. Berkeley: Uni- Acknowledgments versity of California Press. I am grateful to Jané Kondev, Michael Ortiz, Ellad Judson, H.F. 1996. The Eighth Day of Creation. Plainview, Tadmor, Ron Miller, Vijay Shenoy, Vivek Shenoy, N.Y.: Cold Spring Harbor Laboratory Press. David Rodney, Klaus Schulten, Darren Segall, and Lau- Knap, J., and M. Ortiz. 2003. Effect of indenter-radius size on rent Dupuy for collaboration and conversation. All of Au(001) nanoindentation. Physical Review Letter 90(22). them played a key role in the development of my under- Article No. 226102. standing of multiscale modeling. Farid Abraham first Villa, E., A. Balaeff, L. Mahadevan, and K. Schulten. 2004. illustrated the meaning of a terabyte to me by compar- Multi-scale method for simulating protein-DNA coming it to the informational content of a library. plexes. Multiscale Modeling and Simulation 2(4): 527–553. Someday we may be able to make a photoreal computer model of anything—no matter what it is made of or how it reflects light.

Capturing and Simulating Physically Accurate Illumination in Computer Graphics

Paul Debevec

Anyone who has seen a recent summer blockbuster has seen the results of dramatic improvements in the realism of computer-generated graphics. Visual-effects supervisors now report that bringing even the most challenging visions of film directors to the screen is no longer in question; with today’s techniques, the only questions are time and cost. Based both on recently developed techniques and techniques that originated in the 1980s, Paul Debevec is executive producer computer-graphics artists can now simulate how light travels within a scene of graphics research at the Insti- and how light reflects off of and through surfaces. tute for Creative Technologies, Uni- Radiosity and Global Illumination versity of Southern California, One of the most important aspects of computer graphics is simulating the Marina del Rey, California. illumination in a scene. Computer-generated images are two-dimensional arrays of computed pixel values, with each pixel coordinate having three numbers indicating the amount of red, green, and blue light coming from the corresponding direction in the scene. Figuring out what these numbers should be for a scene is not trivial, because each pixel’s color is based both on how the object at that pixel reflects light and the light that illuminates it. Furthermore, the illumination comes not only directly from the light sources in the scene, but also indirectly from all of the surrounding surfaces in the form of “bounce” light. The complexities of the behavior of light—one reason the world around us appears rich and interesting—make WINTER 2004 29

generating “photoreal” images both conceptually and The result that Goral et al. obtained (Figure 1) cor- computationally complex. rectly modeled that the white wall would subtly pick up As a simple example, suppose we stand at the back of the red and blue of the neighboring surfaces. Soon after a square white room in which the left wall is painted red, this experiment, when Cornell researchers constructed the right wall is painted blue, and the light comes from such a box with real wood and paint, they found that the ceiling (Figure 1). If we take a digital picture of this photographs of the box matched their simulations so room and examine its pixel values, we will indeed find closely that people could not tell the difference under that the red wall is red and the blue wall is blue. But controlled conditions. The first “photoreal” image had when we look closely at the white wall in front of us, we been rendered! will notice that it isn’t perfectly white. Toward the right it becomes bluish, and toward the left it becomes pink. The reason for this is indirect illumination: toward the right, blue light from the blue wall adds to the illumination on the back wall, and toward the left, red light does the same. Indirect illumination is responsible for more than the somewhat subtle effect of white surfaces picking up the colors of nearby objects—it is often responsible for most, sometimes all, of the illumination on an object or in a scene. If I sit in a white room illuminated by a small sky- light in the morning, the indirect light from the patch of sunlight on the wall lights the rest of the room, not the direct light from the sun itself. If light did not bounce between surfaces, the room would be nearly dark! In early computer graphics, interreflections of light between surfaces in a scene were poorly modeled. Light FIGURE 1 A simulation of indirect illumination in a scene, known as the “Cornell falling on each surface was computed solely as a function box.” The white wall at the back picks up color from the red and blue walls at of the light coming directly from light sources, with per- the sides. Source: Goral et al., 1984. (The Cornell box may be seen in color at haps a roughly determined amount of “ambient” light http://www.graphics.cornell.edu/online/box/history.html.) added irrespective of the actual colors of light in the scene. The groundbreaking publication showing that One limitation of this work was that the time indirect illumination could be modeled and computed required to solve the linear system increased with the accurately was presented at SIGGRAPH 84, when cube of the number of patches in the scene, making the Goral et al. of Cornell University described how they technique difficult to use for complex models (especially had simulated the appearance of the red, white, and blue in 1984). Another limitation was that all of the surfaces room example using a technique known as radiosity. in the radiosity model were assumed to be painted in Inspired by physics techniques for simulating heat matte colors, with no shine or gloss. transfer, the Cornell researchers first divided each wall of A subsequent watershed work in the field was pre- the box into a 7 × 7 grid of patches; for each patch, they sented at SIGGRAPH 86, by Jim Kajiya from Caltech, determined the degree of its visibility to every other who published the “The Rendering Equation,” which patch, noting that patches reflect less light if they are generalized the ideas of light transport to any kind of farther apart or facing away from each other. The final geometry and any sort of surface reflectance. The titu- light color of each patch could then be written as its lar equation of Kajiya’s paper stated in general terms inherent surface color times the sum of the light coming that the light leaving a surface in each direction is a from every other patch in the scene. Despite the fact function of the light arriving from all directions upon that the illumination arriving at each patch depends on the surface, convolved by a function that describes how the illumination arriving (and thus leaving) every other the surface reflects light. The latter function, called the patch, the radiosity equation could be solved in a bidirectional reflectance distribution function (BRDF), straightforward way as a linear system of equations. is constant for diffuse surfaces but varies according to The 30 BRIDGE the incoming and outgoing directions of light for sur- light coming from every direction, which dramatically faces with shine and gloss. and subtly shape the appearance of the forms in a scene. Kajiya described a process for rendering images Computer-generated scenes, when constructed from according to this equation using a randomized numeri- simple shapes, textured with ideal plastic and metallic cal technique known as path tracing. Like the earlier reflectance properties, and illuminated by simple point fundamental technique of ray tracing (Whitted, 1980), and area light sources, lack “realism” no matter how path tracing generates images by tracing rays from a accurate or computationally intensive the lighting sim- camera to surfaces in the scene, then tracing rays out ulation. As a result, creating photoreal images was still from these surfaces to determine the incidental illumi- a matter of skilled artistry rather than advanced tech- nation on the surfaces. In path tracing, rays are traced nology. Digital artists had to adjust the appearance of not only in the direction of light sources, but also ran- scenes manually. domly in all directions to account for indirect light from Realistic geometry in computer-generated scenes was the rest of the scene. The demonstration Kajiya pro- considerably advanced in the mid-1980s when 3-D dig- duced for his paper is shown in Figure 2. This simple itizing techniques became available for scanning the scene shows realistic light interactions among both dif- shapes of real-world objects into computer-graphics pro- fuse and glossy surfaces, as well as other complex effects, grams. The Cyberware 3-D scanner, an important part such as light refracting through translucent objects. of this evolution, transforms objects and human faces Although still computationally intensive, Kajiya’s ran- into 3-D polygon models in a matter of seconds by mov- domized process for estimating solutions to the rendering a stripe of laser light across them. An early use of ing equation made the problem tractable both this scanner in a motion picture was in Star Trek IV for conceptually and computationally. an abstract time-travel sequence showing a collage of 3-D models of the main characters’ heads. 3-D digitiza- tion techniques were also used to capture artists’ models of extinct creatures to build the impressive digital dinosaurs for Jurassic Park.

Digitizing and Rendering with Real-World Illumination Realism in computer graphics advanced again with techniques that can capture illumination from the real world and use it to create lighting in computer-generated scenes. If we consider a particular place in a scene, the light at that place can be described as the set of all colors and intensities of light coming toward it from every direction. As it turns out, there is a relatively straightforward FIGURE 2 A synthetic scene rendered using path tracing. Source: Kajiya, 1986. way to capture this function for a real-world location by taking an image of a mirrored sphere, which reflects light coming from the whole environment toward the camera. Bringing Reality into the Computer Other techniques for capturing omnidirectional images Using the breakthroughs in rendering techniques include fisheye lenses, tiled panoramas, and scanning developed in the mid-1980s, it was no simple endeavor panoramic cameras. to produce synthetic images with the full richness and The first and simplest form of lighting from images realism of images in the real world. Photographs appear taken from a mirrored sphere is known as environment “real” because shapes in the real world are typically dis- mapping. In this technique, the image is directly warped tinctive and detailed, and surfaces in the real world and applied to the surface of the synthetic object. The reflect light in interesting ways, with different charac- technique using images of a real scene was used inde- teristics that vary across surfaces. And also, very impor- pendently by Gene Miller and Mike Chou (Miller and tantly, light in the real world is interesting because Hoffman, 1984) and Williams (1983). Soon after, the typically there are different colors and intensities of technique was used to simulate reflections on the silvery, WINTER 2004 31

a b

FIGURE 3 Environment-mapped renderings from the early 1980s. (a) An environment-mapped shiny dog. Source: Miller and Hoffman, 1984. (b) An environment-mapped shiny robot. Source: Williams, 1993. computer-generated spaceship in the 1986 film Flight of of light visible in a scene. In a typical scene, directly vis- the Navigator and, most famously, on the metallic T1000 ible light sources are usually hundreds or thousands of “terminator” character in the 1991 film Terminator 2. In times brighter than indirect illumination from the rest all of these examples, the technique not only produced of the scene, and both types of illumination must be cap- realistic reflections on the computer-graphics object, but tured to represent the lighting accurately. This wide also made the object appear to have truly been in the dynamic range typically exceeds the dynamic range of background environment. This was an important both digital and film cameras, which are designed to advance for realism in visual effects. Computer-graphics capture a range of brightness values of just a few hun- objects now appeared to be illuminated by the light of dred to one. As a result, light sources typically become the environment they were in (Figure 3). “clipped” at the saturation point of the image sensor, Environment mapping produced convincing results leaving no record of their true color or intensity. This is for shiny objects, but innovations were necessary to not a major problem for shiny metal surfaces, because extend the technique to more common computer- shiny reflections would become clipped anyway in the graphics models, such as creatures, digital humans, and final rendered images. However, when lighting more cityscapes. One limitation of environment mapping is typical surfaces—surfaces that blur the incidental light that it cannot reproduce the effects of object surfaces before reflecting it back toward the camera—the effect shadowing themselves or of light reflecting between sur- of incorrectly capturing the intensity of direct light faces. The reason for this limitation is that the lighting sources in a scene can be significant. environment is applied directly to the object surface We developed a technique to capture the full dynamic according to its surface orientation, regardless of the range of light in a scene, up to and including direct light degree of visibility of each surface in the environment. sources (Debevec and Malik, 1997). Photographs are For surface points on the convex hull of an object, cor- taken using a series of varying exposure settings on the rect answers can be obtained. However, for more typi- camera; brightly exposed images record indirect light cal points on an object, appearance depends both on from the surfaces in the scene, and dimly exposed images which directions of the environment they are visible to record the direct illumination from the light sources with- and light received from other points on the object. out clipping. Using techniques to derive the response A second limitation of the traditional environment curve of the imaging system (i.e., how recorded pixel val- mapping process is that a single digital or digitized pho- ues correspond to levels of scene brightness), we assemble tograph of an environment rarely captures the full range this series of limited-dynamic-range images into a single The 32 BRIDGE high-dynamic image representing the full range of each surface point and uses this sum as the point’s illu- illumination for every point in the scene. Using IEEE mination. The elegance of this approach is that it pro- floating-point numbers for the pixel values of these high- duces all of the effects of the real object’s appearance dynamic-range images (called HDR images or HDRIs), illuminated by the light of the environment, including ranges exceeding even one to a million can be captured self-shadowing, and it can be applied to any material, and stored. from metal to plastic to glass. The following year we presented an approach to illu- We first demonstrated HDRI and IBL in a short com- minating synthetic objects with measurements of real- puter animation called Rendering with Natural Light, world illumination known as image-based lighting shown at the SIGGRAPH 98 computer-animation fes- (IBL), which addresses the remaining limitations of tival (Figure 4, left). The film featured a still life of dif- environment mapping (Debevec, 1998). The first step fuse, shiny, and translucent spheres on a pedestal in IBL is to map the image onto the inside of a surface, illuminated by an omnidirectional HDRI of the light in such as an infinite sphere, surrounding the object, rather the eucalyptus grove at UC Berkeley. We later used our than mapping the image directly onto the surface of the lighting capture techniques to record the light in St. object. We then use a global illumination system (such Peter’s Basilica, which allowed us to add virtual tum- as the path-tracing approach described by Kajiya [1986]) bling monoliths and gleaming spheres to the basilica’s to simulate this image of incidental illumination actu- interior for our SIGGRAPH 99 film Fiat Lux (Figure 4, ally lighting the surface of the object. In this way, the right). In Fiat Lux, we used the same lighting tech- global illumination algorithm traces rays from each niques to compute how much light the new objects object point out into the scene to determine what is would obstruct from hitting the ground; thus, the syn- lighting it. Some of the rays have a free path away from thetic objects cast shadows in the same way they would the object and thus strike the environmental lighting have if they had actually been there. surface. In this way, the illumination from each visible The techniques of HDRI and IBL, and the techniques part of the environment can be accounted for. Other and systems derived from them, are now widely used in rays strike other parts of the object, blocking the light it the visual-effects industry and have provided visual- would have received from the environment in that effects artists with new lighting and compositing tools direction. If the system computes additional ray that give digital actors, airplanes, cars, and creatures the bounces, the color of the object at the occluding surface appearance of actually being present during filming, point is computed in a similar way; otherwise, the algo- rather than added later via computer graphics. Exam- rithm approximates the light arriving from this direc- ples of elements illuminated in this way include the tion as zero. The algorithm sums up all of the light transforming mutants in X-Men and X-Men 2, virtual arriving directly and indirectly from the environment at cars and stunt actors in The Matrix Reloaded, and whole

FIGURE 4 Still frames from the animations Rendering with Natural Light (left) and Fiat Lux (right) showing computer-generated objects illuminated by and integrated into real-world lighting environments. WINTER 2004 33

cityscapes in The Time Machine. In our latest computer have well defined surfaces or easy-to-model reflectance animation, we extended the techniques to capture the properties. The object can have arbitrary self-shadowing, full range of light of an outdoor illumination environ- interreflection, translucency, and fine geometric detail. ment—from the pre-dawn sky to a direct view of the This is helpful for modeling and rendering human faces, sun—to illuminate a virtual 3-D model of the which exhibit all of these properties, as do many objects Parthenon on the Athenian Acropolis (Figure 5). we encounter in our everyday lives. Recently, our group constructed two additional light stages. Light Stage 2 (Figure 6) uses a rotating semi- circular arm of strobe lights to illuminate the face from a large number of directions in about eight seconds, much more quickly than Light Stage 1. For this short a period of time, an actor can hold a steady facial expression for the entire capture session. By blending the geometry and reflectance of faces with different facial expressions, we have been able to create novel animated performances that can be realistically rendered from new points of view and under arbitrary illumination (Hawkins et al., in press). Mark Sagar and his colleagues at Sony Pictures Imageworks used related FIGURE 5 A virtual image of the Parthenon synthetically illuminated with a lighting techniques to create the digital stunt actors of Tobey environment captured at the USC Institute for Creative Technologies. Maguire and Alfred Molina from light stage data sets for the film Spider-Man 2. For Light Stage 3, we built a complete sphere of Applying Image-Based Lighting to Actors 156 light sources that can illuminate an actor from all In my laboratory’s most recent work, we have exam- directions simultaneously (Debevec et al., 2002). Each ined the problem of illuminating real objects and peo- light consists of a collection of red, green, and blue LEDs ple, rather than computer-graphics models, with light interfaced to a computer so that any light can be set to captured from real-world environments. To accomplish any color and intensity. The light stage can be used to this we use a series of light stages to measure directly reproduce the illumination from a captured lighting how an object transforms incidental environmental illu- environment by using the light stage as a 156-pixel dis- mination into reflected radiance, a data set we call the play device for the spherical image of incidental illumi- reflectance field of an object. nation. A person standing inside the sphere then The first version of the light stage consisted of a spot- becomes illuminated by a close approximation of the light attached to a two-bar rotation mechanism that light that was originally captured. When composited rotated the light in a spherical spiral about a person’s face over a background image of the environment, it appears in approximately one minute (Debevec et al., 2000). nearly as if the person were there. This technique may At the same time, one or more digital video cameras improve on how green screens and virtual sets are used recorded the object’s appearance under every form of today. Actors in a studio can be filmed lit as if they were directional illumination. From this set of data, we could somewhere else, giving visual-effects artists much more render the object under any form of complex illumina- control over the realism of the lighting process. tion by computing linear combinations of the color In our latest tests, we use a high-frame camera to channels of the acquired images. The illumination could capture how an actor appears under several rapidly be chosen to be measurements of illumination in the real cycling basis lighting conditions throughout the course world (Debevec, 1998) or the illumination present in a of a performance (Debevec et al., 2004). In this way, we virtual environment, allowing the image of a real person can simulate the actor’s appearance under a wide variety to be photorealistically composited into a scene with the of different illumination conditions after filming, pro- correct illumination. viding directors and cinematographers with never- An advantage of this photometric approach for cap- before-available control of the actor’s lighting during turing and rendering objects is that the object need not postproduction. The 34 BRIDGE

varying color and shininess components (Gardner et al., 2003). We found that by moving a neon tube light source across a relatively flat object and recording the light’s reflections using a video camera we could independently estimate the diffuse color and the specular properties of every point on the object. For example, we digitized a 15th-century illuminated manuscript with colored inks and embossed gold lettering (Figure 7a). Using the derived maps for diffuse and specular FIGURE 6 Light Stage 2 is designed to illuminate an object or a person from all possible directions in a short period of time, allow- reflection, we were able to ing a digital video camera to capture directly the subject’s reflectance field. Source: Hawkins et al., in press. render computer-graphics versions of the manuscript under any sort of lighting A Remaining Frontier: Digitizing Reflectance environment with realistic glints and reflections from Properties different object surfaces. Significant challenges remain in the capture and sim- A central complexity in digitizing reflectance prop- ulation of physically accurate illumination in computer erties for more general objects is that the way each graphics. Although techniques for capturing object point on an object’s surface responds to light is a com- geometry and lighting are maturing, techniques for cap- plex function of the direction of incidental light and turing object reflectance properties—the way the sur- the viewing direction—the surface’s four-dimensional faces of a real-world object respond to light—are still BRDF. In fact, the behavior of many materials and weak. In a recent project, our laboratory presented a objects is even more complicated than this, in that relatively simple technique for digitizing surfaces with incidental light on other parts of the object may scatter

a bc

Figure 7 (a) A digital model of a digitized illuminated manuscript lit by a captured lighting environment. Source: Gardner et al., 2003. (b) A synthetic model of a face using a simulation of subsurface scattering. Source: Jensen et al., 2001. (c) A digital model of a translucent alabaster sculpture. Source: Goesele et al., in press. WINTER 2004 35

within the object material, an effect known as sub- on Computer Graphics and Interactive Techniques 97 surface scattering (Hanrahan and Krueger, 1993). International Conference. New York: ACM Press/Addi- Because this effect is a significant component of the son-Wesley. appearance of human skin, it has been the subject Debevec, P.E., and J. Malik. 1997. Recovering high dynamic of interest in the visual-effects industry. New tech- range radiance maps from photographs. Pp. 369–378 of the niques for simulating subsurface scattering effects on Proceedings of the Association for Computing Machinery computer-generated models (Jensen et al., 2001) have Special Interest Group on Computer Graphics and Inter- led to more realistic renderings of computer-generated active Techniques 97 International Conference. New actors (Figure 7b) and creatures, such as the Gollum York: ACM Press/Addison-Wesley. character in Lord of the Rings. Debevec, P., A. Wenger, C. Tchou, A. Gardner, J. Waese, and Obtaining models of how real people and objects T. Hawkins. 2002. A lighting reproduction approach to scatter light in their full generality is a subject of on- live-action compositing. Association for Computing going research. In a recent study, Goesele et al. (in Machinery Transactions on Graphics 21(3): 547–556. press) used a computer-controlled laser to shine a nar- Gardner, A., C. Tchou, T. Hawkins, and P. Debevec. Linear row beam onto every point of a translucent alabaster light source reflectometry. 2003. Association for Comput- sculpture (Figure 7c) and recorded images of the resulting Machinery Transactions on Graphics 22(3): 749–758. ing light scattering using a specially chosen high- Goesele, M., H.P.A. Lensch, J. Lang, C. Fuchs, and H-P. dynamic-range camera. By making the simplifying Seidel. In press. DISCO—Acquisition of Translucent assumption that any point on the object would respond Objects. Paper from the Association for Computing equally to any incidental and radiant light direction, the Machinery Special Interest Group on Computer Graphics dimensionality of the problem was reduced from eight to and Interactive Techniques 31st International Conference, four dimensions yielding a full characterization of the August 8–14, 2004. object’s interaction with light under these assumptions. Goral, C.M., K.E. Torrance, D.P. Greenberg, and B. Battaile. As research in this area continues, we hope to develop 1984. Modelling the interaction of light between diffuse the capability of digitizing anything—no matter what surfaces. Association for Computer Machinery Computer it is made of or how it reflects light—so it can become Graphics (3): 213–222. an easily manipulated, photoreal computer model. For Hanrahan, P., and W. Krueger. 1993. Reflection from layered this, we will need new acquisition and analysis tech- surfaces due to subsurface scattering. Pp. 165–174 of the niques and continued increases in computing power and Proceedings of the Association for Computing Machinery memory capacity. Special Interest Group on Computer Graphics and Inter- active Techniques 1993 International Conference. New References York: ACM Press/Addison-Wesley. Debevec, P. 1998. Rendering synthetic objects into real Hawkins, T., A. Wenger, C. Tchou, A. Gardner, F. Goransson, scenes: bridging traditional and image-based graphics and P. Debevec. In press. Animatable facial reflectance with global illumination and high dynamic range photog- fields. Paper from the 15th Eurographics Symposium on raphy. Pp. 189–198 of the Proceedings of the Association Rendering in Norrkoping, Sweden, June 21–23, 2004. for Computing Machinery Special Interest Group on Jensen, H.W., S.R. Marschner, M. Levoy, and P. Hanrahan. Computer Graphics and Interactive Techniques 98 Inter- 2001. A practical model for subsurface light transport. Pp. national Conference. New York: ACM Press/Addison- 511–518 of the Proceedings of the Association for Com- Wesley. puting Machinery Special Interest Group on Computer Debevec, P., A. Gardner, C. Tchou, and T. Hawkins. 2004. Graphics and Interactive Techniques 2001 International Postproduction re-illumination of live action using time- Conference. New York: ACM Press/Addison-Wesley. multiplexed lighting. Technical Report No. ICT TR Kajiya, J.T. 1986. The rendering equation. Association for 05.2004. Los Angeles: Institute for Creative Technologies, Computing Machinery Special Interest Group on Com- University of Southern California. puter Graphics and Interactive Techniques Computer Debevec, P., T. Hawkins, C. Tchou, H.P. Duiker, W. Sarokin, Graphics 20(4): 143–150. and M. Sagar. 2000. Acquiring the reflectance field of a Miller, G.S., and C.R. Hoffman. 1984. Illumination and human face. Pp. 145–156 of the Proceedings of the Asso- Reflection Maps: Simulated Objects in Simulated and Real ciation for Computing Machinery Special Interest Group Environments. Course Notes for Advanced Computer The 36 BRIDGE

Graphics Animation. New York: Association for Com- Williams, L. 1983. Pyramidal parametrics. 1983. Associa- puting Machinery Special Interest Group on Computer tion for Computing Machinery Special Interest Group on Graphics and Interactive Techniques Computer Graphics. Computer Graphics and Interactive Techniques Computer Whitted, T. 1980. An improved illumination model for Graphics 17(3): 1–11. shaded display. Communications of the Association for Computing Machinery 23(6): 343–349. Sound is inherently a spatial perception.

Spatial Audio Reproduction: Toward Individualized Binaural Sound

William G. Gardner

The compact disc format, which records audio with 16-bit resolution at a sampling rate of 44.1 kHz, was engineered to reproduce audio with fidelity exceeding the limits of human perception. And it works. However, sound is inherently a spatial perception. We perceive the direction, distance, and size of sound sources, and reproducing the spatial properties of sound accurately remains a challenge. In this paper, I review the technologies for spa- William G. Gardner is founder and tial sound reproduction and discuss future directions, focusing on the promise president of Wave Arts Inc. located of individualized, binaural technology. in Arlington, Massachusetts. Hearing People hear with two ears, and the two audio signals received at the eardrums completely define the auditory experience. An amazing feature of the auditory system is that with only two ears sounds can be perceived from all directions, and the listener can even sense the distance and size of sound sources. The perceptual cues for sound localization include the amplitude of the sound at each ear, the arrival time at each ear, and the spectrum of the sound, that is, the relative amplitude of the sound at different frequencies. The spectrum of a sound is modified by interactions between sound waves and the torso, head, and external ear (pinna). Furthermore, spectral modification depends on the location of the source in a complex way. The auditory The 38 BRIDGE system uses spectral modifications as cues to the location A complete description of a subject’s head response of sound, but because the complex shape of the pinna requires hundreds of HRTF measurements from all varies significantly among individuals, the cues for sound directions surrounding the subject. Any sound source localization are idiosyncratic. Each individual’s auditory can be virtually located by filtering the sound with the system is adapted to the idiosyncratic spectral cues pro- HRTFs corresponding to the desired location and pre- duced by his or her head features. senting the resulting binaural signal to the subject using properly equalized headphones. When this procedure is Binaural Audio individualized by using the subject’s own HRTFs, the Binaural audio refers specifically to the recording and localization performance is equivalent to free-field lis- reproduction of sound at the ears. Binaural recordings tening (Wightman and Kistler, 1989a,b). can be made by placing miniature microphones in the Figure 1 shows the magnitude spectra for right ear ear canals of a human subject. Exact reproduction of HRTFs measured for three different human subjects the recording is possible through properly equalized with a sound source located on the horizontal plane at headphones. If the recording and playback are for the 60 degrees right azimuth. Note that the spectra are sim- same subject and there are no head movements, the ilar up to 6 kHz; the significant differences in HRTFs at results are stunningly realistic. higher frequencies are attributable to variations in pinna Many virtual-reality audio applications attempt to shape. Figure 2 shows the magnitude spectra of HRTFs position a sound arbitrarily around a listener wearing measured from a dummy head microphone for all loca- headphones. They rely on a stored database of head- tions on the horizontal plane. Note how the spectral fea- related transfer functions (HRTFs), that is, mathemati- tures change as a function of source direction. cal descriptions of the transformation of sound by the Most research has been focused on localization; sub- torso, head, and external ear. HRTFs for the left and jects presented with an acoustic stimulus are asked to right ears of a subject specify how sound from a particu- report the apparent direction. Their localization is lar direction is transformed en route to the ear drums. then compared to free-field listening to assess the quality of reproduction. But this method does not 10 account for many attri- butes of sound perception, including distance, timbre, 5 and size. In an experimental paradigm developed by Hartmann and Wittenburg 0 (1996), the virtual stimulus is reproduced using open- air headphones that allow -5 dB free-field listening. Thus, real and virtual stimuli can be compared directly. In -10 these experiments, subjects are presented with a stimu- -15 lus and asked to decide if it is real or virtual. If a virtual stimulus cannot -20 be distinguished from a 2 3 4 10 10 10 real stimulus, then the freq,Hz reproduction error is within the limits of perception. FIGURE 1 Spectrum magnitude for right-ear HRTFs for three different human subjects with a sound source at 60 degrees right azimuth When this experimental on the horizontal plane. Due to variations in ear shape, the HRTFs differ significantly above 6 kHz. paradigm was used to study WINTER 2004 39

“non-individualized” system (Wenzel et al., 1993). The use of non-individualized HRTFs is limited by a lack of externalization (the sounds are localized in the head or very close to the head), incorrect perception of elevation angle, and front/back reversals. Externaliza- tion can be improved somewhat by adding dynamic head tracking and reverberation. Nevertheless, the lack of realistic externalization is often cited as a problem with these systems. The great challenge in binaural technology is to devise a practical method by which binaural signals can be individualized to a specific listener. There are several possible approaches to meeting this challenge: acoustic measurement, statistical models, calibration procedure, simplified geometrical models, and accurate head models solved using computational acoustics. With the proper equipment, measuring the HRTFs of a listener is a straightforward procedure, although not practical for commercial applications. Micro- phones are placed in the ears of the listener, either probe microphones placed somewhere in the ear canal or microphones that block the entrances to the ear canals. Measurement signals are produced from speakers surrounding the listener to measure the impulse response of each source direction to each ear. Because tens or hundreds of directions may be measured, the listener is positioned either in a rotating chair or in a fixed position surrounded by hundreds of speakers. The measurements are often made in an anechoic (echo- free) chamber. Various statistical methods have been used to ana- FIGURE 2 Magnitude spectra of KEMAR dummy-head HRTFs as a function of lyze databases of HRTF measurements in an effort to azimuth for a horizontal source (Gardner, 1998). 2a. Ipsilateral (same side) ear. tease out some underlying structure in the data. One 2b. Contralateral (opposite side) ear. White indicates +10 dB; black indicates important study applied principal component analysis –30 dB. Notch features are labeled in (a) according to Lopez-Poveda and Meddis (PCA) to a database of HRTFs from 10 listeners at (1996). The figure shows the complex, yet systematic, variation in spectrum as a 256 directions (Kistler and Wightman, 1992). Using function of source direction. the log magnitude spectra of the HRTFs as input, the analysis indicated that 90 percent of the variance in the the externalization of virtual sound, the results demon- data could be accounted for using only five principal strated that individualized spectral cues are necessary components. The study tested the localization perfor- for proper externalization. mance using individualized HRTFs approximated by The major limitation of binaural techniques is that weighted sums of the five principal components. When all listeners are different. Binaural signals recorded for the listener’s own HRTFs were used, the results were subject A may not sound correct to subject B. Never- nearly identical. The study gathered only directional theless, by necessity, binaural systems are seldom indi- judgments, and externalization was not considered. vidualized. Instead, a reference head, often a model that The study showed that a five-parameter model is suffi- represents a typical listener or HRTFs known to perform cient for synthesizing individualized HRTF spectra, at adequately for a range of different listeners, is used to least in terms of directional localization and for a single encode binaural signals for all listeners. This is called a direction. Unfortunately, the five parameters must be The 40 BRIDGE calculated for each source direction, which means individualized measurements are still necessary. One can imagine a simple calibration procedure that would involve the listener adjusting knobs to match a parameterized HRTF model with the listener’s characteristics. The listener could be given a test stimulus and asked to adjust a knob until some attribute of his perception was maximized. After adjusting several knobs in this manner, the parameter values of the internal model would be optimized for the listener, and the model would be able to generate individualized HRTFs for that individual. Some progress has been made in this area. For example, it has been demonstrated that calibrating HRTFs according to overall head size improves localization performance (Middlebrooks et al., 2000). However, to date, detailed methods of modeling and calibrating the data have not been found. Many researchers have developed geometrical models for the torso, head, and ears. The head and torso can be modeled using ellipsoids (Algazi and Duda, 2002), and the pinna can be modeled as a set of simple geometrical objects (Lopez-Poveda and Meddis, 1996). For simple geometries, the acoustic-wave equation can be solved to FIGURE 3 Mesh model of one-half of a KEMAR dummy head using 15,000 elements. determine head response. For more complicated geome- The model is sufficiently detailed to produce HRTFs that match acoustical measure- tries, head response can be approximated using a multi- ments, even at high frequencies. Source: Kahana, 1999. path model, wherein each reflecting or diffracting object contributes an echo to the response (Brown and Duda, There are, however, a number of practical difficulties 1997). In theory, head models should be easy to fit to with this method. First, scanning the head is compli- any particular listener by making anthropometric mea- cated by the presence of hair, obscured areas behind the surements of the listener and plugging these into the ears, and the obscured internal features of the ear. Sec- model. However, studies have shown that although ond, replicating the interior features of the ear requires simplified geometrical models are accurate at low making molds and then scanning them separately. Third, frequencies, they become increasingly inaccurate at after the various scans are spliced together, the number of higher frequencies. Because of the importance of high- elements in the model must be pruned to computation- frequency localization cues for proper externalization, ally tractable quantities without compromising spatial elevation perception, and front/back resolution of resolution. Finally, solution of the acoustical equations sound, simplified geometrical models are not suitable for requires significant computation. For all of these reasons, creating individualized HRTFs. this approach currently requires more effort and expense A more promising approach has been to use a three- than acoustical measurement of HRTFs. dimensional laser scan to produce an accurate geometri- The technique does suggest an alternative approach to cal representation of a head as a basis for computational determining individualized HRTFs. A deformable head acoustic simulation using finite-element modeling (FEM) model could be fashioned from finite elements and para- or boundary-element modeling (BEM) (Kahana et al., meterized with a set of anthropometric measurements. 1998, 1999). With this method, HRTFs can be deter- After making head measurements of a particular subject mined computationally with the same accuracy as acousti- and plugging these into the model, the model head would cal measurements, even at high frequencies. Using a “morph” into a close approximation of the subject’s head. 15,000-element model of the head and ear (Figure 3), At that point, the computational acoustics procedure Kahana demonstrated computation of HRTFs that match could be used to determine individualized HRTFs for the acoustical measurements very precisely up to 15 kHz. subject. Ideally, the subject’s measurements could be WINTER 2004 41

determined from images using computer vision tech- is necessary to achieve a sense of space. niques. The goal would be a system that could automat- Multichannel audio systems, such as the current ically determine individualized HRTFs based on a few 5.1 surround systems, have continued the trend of digital images of the subject’s head and ears. The chal- adding channels around the listener to improve spatial lenges will be to develop a head model that can morph to reproduction. 5.1 systems have left, center, and right fit any head, to obtain a sufficiently accurate ear shape, frontal speakers, left and right surround speakers posi- and to develop ways to estimate the parameters from tioned to the sides of the listener, and a subwoofer to images of the subject. reproduce low frequencies. Because 5.1 systems were designed for cinema sound, the focus is on accurate Cross-talk–Cancelled Audio frontal reproduction so that movie dialogue is spatially Binaural audio can be delivered to a listener over con- aligned with images of the actors speaking. The sur- ventional stereo loudspeakers, but each loudspeaker round speakers are used for off-screen sounds or uncor- (unlike headphones) creates significant “cross-talk” to related ambient effects. The trend in multichannel the opposite ear. The cross-talk can be cancelled by pre- audio is to add more speaker channels to improve the processing the speaker signals (called cross-talk can- accuracy of on-screen sounds and provide additional cellers) with the inverse of the 2 × 2 matrix of transfer locations for off-screen sounds. As increasing numbers functions from the speakers to the ears. Cross-talk can- of speakers are added at the perimeter of the listening cellers use a model of the head to anticipate what cross- space, it becomes possible to reconstruct arbitrary talk will occur, then add an out-of-phase cancellation sound fields within the space, a technology called signal to the opposite channel. Thus, the cross-talk is wave-field synthesis. acoustically cancelled at the listener’s ears. If the head responses of the listener are known, and if the listener’s head remains fixed, an individualized cross-talk cancellation system can be designed that works extremely well. Stereophonic audio systems Non-individualized systems are effective only up to 6 kHz and then only when the listener’s position is reproduce a set of locations known (Gardner, 1998). However, despite poor high- between the speakers. frequency performance, cross-talk–cancelled audio is capable of producing stunning, well externalized, virtual sounds to the sides of the listener using frontally placed loudspeakers. As a result of the listener’s pinna cues, the Ultrasonic Audio sounds are well externalized. The sounds are shifted to Ultrasonics can be used to produce highly directional the side as a result of the dominance of low-frequency audible sound beams. This technology is based on phys- time-delay cues in lateral localization; the cross-talk can- ical properties of air, particularly that air becomes a cellation works effectively at low frequencies to provide nonlinear medium at high sound pressures. Hence, it is this cue. possible to transmit two high-intensity ultrasonic tones, say at 100 kHz and 101 kHz, and produce an Multichannel Audio audible 1 kHz tone as a result of the intermodulation The first audio reproduction systems were mono- between the two ultrasonic tones. However, the phonic, reproducing a single audio signal through one demodulated signal will be significantly distorted, so transducer. Stereophonic audio systems, recording and the audio must be preprocessed to reduce the distortion reproducing two independent channels of audio, sound after demodulation (Pompei, 1999). Although this much more realistic. With two loudspeakers, it is pos- technology is impressive, it cannot reproduce low- sible to position a sound source at either speaker or to frequency sounds effectively, and it has lower fidelity position sounds between the speakers by sending a pro- than standard loudspeakers. portion of the sound to each speaker. Stereo has a great advantage over mono because it reproduces a set of Summary locations between the speakers. Also with stereo, Binaural audio has the potential to reproduce sound uncorrelated signals can be sent to the two ears, which that is indistinguishable from sounds in the real world. The 42 BRIDGE

However, the playback must be individualized to each of the Audio Engineering Society’s 16th International listener’s head response. This is currently possible by Conference. New York: Audio Engineering Society. making acoustical measurements or by making geomet- Kistler, D.J., and F.L. Wightman. 1992. A model of head- rical scans and applying computational acoustic model- related transfer functions based on principal components ing. A practical means of individualizing head responses analysis and minimum-phase reconstruction. Journal of the has yet to be developed. Acoustical Society of America 91(3): 1637–1647. Lopez-Poveda, E.A., and R. Meddis. 1996. A physical References model of sound diffraction and reflections in the human Algazi, V.R., and R.O. Duda. 2002. Approximating the head- concha. Journal of the Acoustical Society of America related transfer function using simple geometric models of 100(5): 3248–3259. the head and torso. Journal of the Acoustical Society of Middlebrooks, J.C., E.A. Macpherson, and Z.A. Onsan. 2000. America 112(5): 2053–2064. Psychophysical customization of directional transfer func- Brown, C.P., and R.O. Duda. 1997. An efficient HRTF tions for virtual sound localization. Journal of the Acousti- model for 3-D sound. Pp. 298–301 in Proceedings of the cal Society of America 108(6): 3088–3091. IEEE Workshop on Applications of Signal Processing to Pompei, F.J. 1999. The use of airborne ultrasonics for gener- Audio and Acoustics. New York: IEEE. ating audible sound beams. Journal of the Audio Engi- Gardner, W.G. 1998. 3-D Audio Using Loudspeakers. neering Society 47(9): 726–731. Boston, Mass.: Kluwer Academic Publishers. Wenzel, E.M., M. Arruda, D.J. Kistler, and F.L. Wightman. Hartmann, W.M., and A. Wittenberg. 1996. On the exter- 1993. Localization using nonindividualized head-related nalization of sound images. Journal of the Acoustical Soci- transfer functions. Journal of the Acoustical Society of ety of America 99(6): 3678–3688. America 94(1): 111–123. Kahana, Y., P.A. Nelson, and M. Petyt. 1998. Boundary ele- Wightman, F.L., and D.J. Kistler. 1989a. Headphone simula- ment simulation of HRTFs and sound fields produced tion of free-field listening I: stimulus synthesis. Journal of by virtual acoustic imaging. Proceedings of the Audio the Acoustical Society of America 85(2): 858–867. Engineering Society’s 105th Convention. Preprint 4817, Wightman, F.L., and D.J. Kistler. 1989b. Headphone simula- unpaginated. tion of free-field listening II: psychophysical validation. Kahana, Y., P.A. Nelson, M. Petyt, and S. Choi. 1999. Journal of the Acoustical Society of America 85(2): Numerical modeling of the transfer functions of a dummy- 868–878. head and of the external ear. Pp. 330–334 in Proceedings WINTER 2004 43

NAE News and Notes

NAE Newsmakers

Frances E. Allen, IBM Fellow future of scientific exploration and communications.” The medal was Emerita, IBM Thomas J. Watson invention in the United States. The presented to Dr. Lanzerotti at the Research Center, was presented the Lemelson-MIT Lifetime Achieve- 35th COSPAR Scientific Assembly Anita Borg Award for Technical ment Award includes a cash prize in Paris, France, on July 19, 2004. Leadership by the Anita Borg Insti- of $100,000. Gordon E. Moore, chairman tute for Women and Technology at Nick Holonyak Jr., John emeritus, Intel Corporation, received a banquet on October 7, 2004. Bardeen Chair Professor of Electri- the Society of Chemical Industry W. Gene Corley, senior vice pres- cal and Computer Engineering and (SCI) America Section 2004 Perkin ident, Construction Technology Physics, University of Illinois, won Medal, on September 14, 2004. The Laboratories Inc., received the the prestigious Lemelson-MIT Perkin Medal is the highest honor ASCE President’s Award for distin- Prize for 2004 for his invention given for outstanding applied chem- guished leadership and service to the of the first practical LED (light- istry in the United States. Dr. United States. The award, which emitting diode), the first visible- Moore, one of the pioneers of the was established to honor George spectrum semiconductor laser, and semiconductor industry, was honored Washington, was presented to Dr. the household dimmer switch. His for helping to create the world’s first Corley in recognition of his service discoveries have had a major impact integrated circuit, the world’s first as the ASCE representative on on the lighting industry, global com- microprocessor, and several other the building-performance assess- munications, and consumer prod- cornerstone products. On the same ment teams organized to investigate ucts. The Lemelson Prize, which day, SCI honored Dr. Moore, widely how well the Alfred P. Murrah Fed- includes a cash award of $500,000, known for “Moore’s law,” by eral Building stood up to the 1995 is the world’s largest single cash prize announcing a new medal named for bombing and the performance of the for invention and is awarded annu- him. The SCI Gordon E. Moore World Trade Center in response to ally to an individual who demon- Medal will be awarded for innova- the terrorist attacks in 2001. strates remarkable inventiveness tion by an industrial scientist under Edith M. Flanigen, retired fellow, and creativity and who inspires 45 years of age. UOP LLC, won the Lemelson-MIT others. Dr. Holonyak received the Two NAE members have been Lifetime Achievement Award in award on April 23 in a ceremony at elected to the National Academies April 2004 for four decades of pio- the National Academy of Sciences. Institute of Medicine (IOM). neering work in chemistry and mate- Louis J. Lanzerotti, consultant, Frances H. Arnold, IBM Fellow rials science that has resulted in more Bell Laboratories, Lucent Technolo- Emerita, IBM Thomas J. Watson than 100 patents and revolutionized gies, and distinguished research pro- Research Center, and Edward H. molecular sieve materials (porous fessor, New Jersey Institute of Kaplan, William N. and Marie A. crystals that can separate molecules Technology, received the Commit- Beach Professor of Management on the basis of size). Her inventions tee on Space Research (COSPAR) Sciences and professor of public have helped make the petroleum- William Nordberg Medal in recog- health, Yale School of Management, refinement process cleaner, safer, and nition of his “exceptional contribu- were elected to the IOM on Octo- more efficient. As part of the award, tions to the practical application of ber 18, 2004. Dr. Flanigen will participate in a fundamental research in space congressional briefing about the physics to space and ground-based The 44 BRIDGE

NAE President Honored by ASME

President Wm. A. Wulf was pre- and appreciation of engineers and engineering. The medal was presented with the American Society engineering. Dr. Wulf was recog- sented during the 2004 Interna- of Mechanical Engineering’s Ralph nized for his many achievements tional Mechanical Engineering Coats Roe Medal on November 17. as NAE president, particularly Congress and RD&D Expo in The medal honors individuals who the initiative to survey activities Anaheim, California. have raised public understanding designed to raise public awareness of

2004 Annual Meeting

NAE members, foreign associates, addressed the group on the question attend the meeting, so his award was and guests gathered in Washington, of outsourcing and other problems accepted by NAE member Howard D.C., this October for the 2004 NAE confronting engineers in the twenty- Brenner, who also read his accep- Annual Meeting. The meeting first century (see p. 47). The induc- tance speech (see p. 51). began on Saturday, October 2, with tion of the NAE Class of 2004 The 2004 Arthur M. Bueche the orientation of new members. followed President Wulf’s address. Award was presented to John B. This was followed by the NAE The program continued with the Slaughter, president and CEO of Council Dinner honoring the 75 new presentation of the 2004 Founders the National Action Council for members and 11 foreign associates. Award to Eli Ruckenstein, Distin- Minorities in Engineering, “for sup- The public session on Sunday, guished Professor of Chemical port of engineering research and October 3, was opened by NAE Engineering at the State University education within the National Sci- Chair Craig R. Barrett, who called of New York, Buffalo, who was rec- ence Foundation, many contribu- for more government funding for ognized “for leadership in moderniz- tions to the development of science R&D to keep the United States coming research and development in and technology policy, and lifelong petitive in the global economy (see key areas of chemical engineering.” dedication to increasing diversity in p. 45). President Wm. A. Wulf then Dr. Ruckenstein was not able to the disciplines of science and engineering.” Dr. Slaughter’s remarks are reprinted on p. 53. Following the presentation of the awards, the recipient of the 2004 Gordon Prize, Frank S. Barnes, lectured on “Opportunities and Challenges for Engineering Educa- tion as Seen from an Interdiscipli- nary Telecommunications Master of Science” (p. 55). The Gordon Prize, which is awarded for new modalities and experiments in education that develop effective engineering leaders, was presented to Dr. Barnes, Distinguished Pro- fessor in the Department of Elec- trical and Computer Engineering, University of Colorado at Boulder, Dr. Melody Moore was presented with the Gilbreth Lecture honor for “Frontiers of Human-Computer Interaction: Direct at an awards ceremony in Wash- Brain Interfaces.” (Left to right) Craig Barrett, Melody M. Moore, Wm. A. Wulf. ington, D.C., last February. WINTER 2004 45

computer interaction. The second Kenneth Flamm, Dean Rusk Chair speaker, Dr. Alan J. Russell, director in International Affairs, Lyndon B. of the McGowan Institute for Johnson School of Public Affairs, Regenerative Medicine at the Uni- University of Texas at Austin; versity of Pittsburgh, spoke about Alfonso Velosa, associate director, using biotechnology to detect and Gartner, Inc.; Delcie R. Durham, counteract chemical weapons. program director, Division of Design, At the Annual Business Session Manufacture and Industrial Innova- on Monday morning, members were tion, National Science Foundation; asked to vote on proposed amend- Lawrence J. Rhoades, president and ments to the NAE Articles of Orga- chief executive officer, Extrude nization and the NAE bylaws. Hone Corporation; and Sidney Members who were not present had Perkowitz, Charles Howard Candler been given an opportunity to vote by Professor, Department of Physics, mail. The amendments to both doc- Emory University. On Monday after- Dr. Alan Russell delivered the Gilbreth Lecture on uments were approved. (The revised noon, members and foreign associ- “Using Biotechnology to Detect and Counteract Chem- documents can be found on the ates participated in NAE section ical Weapons.” NAE website, http://www.nae.edu.) meetings at the Keck Center. The business session was followed The final event of the meeting After a break, the two recipients by a symposium entitled “A Century was the annual reception and dinner of the Lillian M. Gilbreth Lecture- of Innovation in Manufacturing.” dance, held for the first time at the ships, which recognize outstanding Speakers included Pamela A. Drew, Andrew W. Mellon Auditorium. young engineers, presented talks. vice president and deputy general Entertainment was provided by the The first speaker, Dr. Melody manager, Airborne Intelligence, Sur- Gross National Product and music Moore, assistant professor in the veillance and Reconnaissance, Inte- was provided by the Radio King Computer Information Systems grated Defense Systems, The Boeing Orchestra. Department at Georgia State Uni- Company; James Padilla, chief oper- The next annual meeting is sched- versity, spoke on frontiers of human- ating officer, Ford Motor Company; uled for October 9 and 10, 2005.

Chairman’s Remarks

ing—a celebration on two fronts. before. In the last 10 years, roughly First, we’re celebrating the success of half of the world’s population has the new inductees to the National joined the free economic system. Academy of Engineering. Second, India, China, Russia, and the East- in a larger sense, we’re celebrating ern European countries have a com- engineering as a whole and the con- bined population of approximately tributions of engineering to the suc- three billion people, roughly half cesses of the United States—the the world’s population. Almost success of the U.S. economy and the overnight in economic terms, half of continued increases in capability the world’s population has suddenly Craig R. Barrett around the world. become available to purchase goods This celebration comes at an and services from the rest of the These remarks were delivered interesting time. As we look around world and contribute their labors October 3, 2004, at the NAE the world today, a significant change and expertise to the global eco- Annual Meeting. is taking place. In fact, it’s a once- nomic system. It’s a pleasure to be here today to in-human-history change, some- Bringing three billion people into celebrate the success of engineer- thing that has never happened the free economic system is bound to The 46 BRIDGE cause dislocations, competition, and three billion. We should expect in R&D, the greater the generation fundamental changes in the way the some strong competition going for- of new ideas and new concepts for system operates. We’ve seen a little ward for engineering-based jobs and goods, services, and businesses. bit of that in the national debate knowledge-based jobs. Investment in R&D has driven the here in the United States, which has The potential response of the established economies around the centered on phrases such as off- United States to these new circum- world today. shoring; outsourcing, the movement stances could be shaped by engi- But if you look at U.S. invest- of jobs out of the United States; and neering fundamentals. First, any ment, specifically in the physical competition for jobs. If any of you competition for economic growth, sciences and engineering, particu- watch Lou Dobbs on CNN Tonight, any competition for standard of liv- larly in basic R&D, which is the you’ve seen him call me, and others, ing, typically goes with the educa- domain of the U.S. government the Benedict Arnold CEOs of the tional level of the workforce. The (primarily the National Science world for moving jobs offshore. Lou average education of a population is Foundation, U.S. Department of often forgets to mention that my a good predictor of the standard of Defense, and U.S. Department of company does 70 percent of its busi- living of the country. If you want Energy) and track that investment ness outside the United States. the highest standard of living in the over the past 20 or 25 years, you find I have yet to figure out how we’re world, as the United States, West- a surprising result. Investment has supposed to do 70 percent of our ern Europe, and Japan currently been basically flat in terms of business overseas but not have any enjoy, you have to assume that you absolute dollars of today. It’s been at foreign employees. need an educated workforce that is about $5 billion for the last 25 years The interesting thing about this second to none. with very little variation. change, though, is that the coun- There’s been a lot of discussion on I am consistently amazed by that tries that have just come into the education in the United States, statistic. I can put it in terms of two free economic system come with specifically K through 12 education, comparisons for you. First, I work rich educational heritages. Russia, and, frankly, if you captured all of the for Intel Corporation. Our R&D China, India, and the Eastern hot air that’s been expended on this budget, for just one company— European countries have excellent topic and funneled it down turbines, mostly for product-oriented applied universities, excellent university we’d probably be energy indepen- R&D rather than basic research—is graduates, and knowledge-based dent. We also have discussions approximately $5 billion a year, workers who can do just about any on No Child Left Behind. Senator which is exactly the total invest- job anyone else in the world can do. Glenn ran a commission a few years ment by the U.S. government in Even if you discount 90 percent of ago that issued a report, Before It’s Too basic R&D. those three billion people as agrar- Late, which basically focused on how The other comparative statistic is ian and uneducated, that still leaves to interest children in K through 12 the amount of money the U.S. gov- 300 million who are educated. in mathematics and science and on ernment spends on agricultural sub- That’s equivalent to the U.S. work- improving the teaching of math- sidies. Last time I looked, agriculture force and the Western European ematics and science, and hence engi- was the industry of the nineteenth workforce. There’s bound to be neering, in the United States. But I century, whereas engineering and competition. think there is probably enough dia- technology represent the industries You can look at another aspect of logue on education already. of the twenty-first century. The the situation. Consider the impact The second engineering metric amount the United States govern- of Taiwan with a population of you can apply to economic compet- ment spends on agricultural subsidies fewer than 25 million or the impact itiveness is research and develop- each year, direct agricultural subsi- of South Korea with a population ment (R&D). If you do a statistical dies, not food stamps, is roughly $25 of fewer than 50 million. Look at study on the standard of living rela- billion. I am continually perplexed their impact on the world’s eco- tive to the percentage of gross that we spend five times as much sup- nomic system, and then make a domestic product invested in R&D, porting industries of the nineteenth rough extrapolation, as any good you find there’s a very strong corre- century as we do investing in the engineer would do, of the impact of lation. The higher the investment industries of the twenty-first century. WINTER 2004 47

So I leave you with one thought are investing heavily in knowledge- mentions investment in R&D as a this morning as we celebrate engi- based industries and universities. way to ensure our future economic neering accomplishments and the The basic research laboratory in the health and economic well-being. I’d ideals of the National Academy of United States is not an industrial love us all to raise our voices and ask Engineering and the new inductees. research laboratory; it is the collec- the U.S. government, Congress, the If we want to be successful in the tion of tier-one and tier-two univer- President, and the presidential can- future, we have to make an invest- sity research laboratories. didates about investing in the ment. There is no way around that. I encourage all of us to raise our future. In fact, there is no future Competition has increased in voices on one topic. I desperately unless we invest today. I hope you the world—to a level that has hope that when the two presidential will join me and Bill Wulf and the never been seen before. Many candidates meet to talk about rest of the Academy in raising our other countries around the world domestic issues at least one of them voices in that direction.

President’s Address

our society, not just for engineering. website (www.nae.edu), with a link So before I turn to outsourcing per at the top level. I encourage you to se, let me put it in context. engage in the discussion. First, everyone who has looked Third, I refer you to the 20 Great- into the problem agrees that the data est Achievements Project. For the on outsourcing are poor! Last year, benefit of new members and guests, at an NAE “summit” on the engi- this is a project we undertook in neering workforce, an incredible 1999 to come up with a list of the range of opinions were expressed— 20 greatest engineering achieve- from shortage to oversupply—and ments of the twentieth century; each opinion was supported by the “greatest” was defined in terms of Wm. A. Wulf selective use of data. A Rand report impact on our quality of life. The released this June makes the same full list is available on our website, These remarks were delivered point, namely, that the data are and a “coffee table” book is for sale October 3, 2004, at the NAE inadequate. As a result, the situa- by the National Academies Press. Annual Meeting. tion is now rife with hype, the sort of The important point for this dis- It is an immense honor to wel- situation the media love to amplify. cussion is how profoundly engi- come once again our new members Let’s try to avoid that. neering has affected society in the and foreign associates. This talk Second, the physical arrangement last 100 years—arguably more than each year gives me an opportunity to today, that is, me at the podium and any other single thing. Remember, reflect on problems and opportuni- you in the audience, is not con- in 1900, few people had electricity ties for engineering. In reviewing ducive to a dialogue. But dialogue is or telephones or automobiles. the possibilities this year, I realized what we need, specifically a dialogue There were no airplanes; the aver- that there is an elephant in the liv- about what we are going to do (or not age life span in the United States ing room, and I had better not do). It’s too easy to fob this problem was 46, whereas it is now 77 (it has ignore it. I am referring to the out- off onto politicians. As the most been estimated that clean water sourcing (or offshoring) of engineer- respected assembly of engineers in added 20 of those 31 years); 50 per- ing jobs to India and China and the United States, NAE should be cent of the U.S. population lived on other places. The outsourcing issue front and center on this issue, which farms, whereas now only 2 percent is connected to other issues that, is critical to our profession. Today, I do because of the increase in pro- taken together, form a “mosaic,” an am setting a context for that dia- ductivity brought about by mechan- image of troubling proportions for logue, a “discussion database” on our ization; there were no antibiotics, The 48 BRIDGE radios, TVs, Internet, and so on. the mid-1990s; it is back up to about been lower than in the workforce as Fourth, I am not an economist— 75,000 in the latest data. To put this a whole! Does that tell us some- whether that’s a good thing or not in context, the number is more than thing has changed structurally? We we’ll have to see—and this is clearly 200,000 per year in China and must be careful in interpreting this an economic issue. So I won’t pre- India, and about one million per because of the crazy economic tend that I can see all of the policy year worldwide. Thus, new U.S. boom/bust we’ve experienced. It’s levers that may be applicable here. engineers account for only about also worth noting that 6 percent In that context, I will tell you 7.5 percent of the world total. unemployment used to be consid- some of the things I know and then Fifth, I know we were importing ered full employment and that less move on to some of the things I engineers on H1B visas, about than 6 percent was considered to believe but can’t prove. I will then 20,000 to 25,000 per year at the cause inflation. Is the situation suggest some things we ought to be peak. I have seen this figure re- today an anomaly, or was the situa- doing and some linkages to other ported as 195,000 in some places, tion in the 1990s an anomaly? It’s issues that form the “mosaic” I but that is the total for all profes- hard to tell! referred to earlier. sions, not just engineering. This is Finally, I know we need to keep the sort of hype that annoys me. things in perspective. Six percent Some Things I Know Sixth, I know we are creating unemployment among engineers is First, I know that the plural of engineering jobs in India and China; about 120,000 individuals. That anecdote is not data! The popular one need only drive down a road in should be compared to the press likes to tell stories about indi- Bangalore to see that. The question 500,000—half a million—Ameri- viduals: “John Jones says he spent is to what extent these jobs are cans who lose their jobs each week. four years being educated for a three- replacing jobs in the United States. In a healthy economy, jobs “churn,” year engineering career, but now he Surely they are to some degree, but I and that’s a good thing because is leaving engineering . . . .” That’s have seen no reliable data on num- that’s how things improve, how we not data. The story is meant to sug- bers. There are reports of good and lose low-end jobs and acquire high- gest that the problem is rampant and bad experiences with outsourcing, end jobs. to make you feel personally threat- and there are also cases of “insourc- Here’s an example. In 1970, there ened. But, in fact, it provides no ing,” in the pharmaceutical industry, were 421,000 switchboard operators basis for knowing anything. for example. supporting 9.8 billion phone calls. Second, I know that conflicting Seventh, starting salaries for new Today there are 78,000 operators definitions of “engineer” by the bachelor-level engineers are not (less than one-fifth as many) sup- National Science Foundation moving up significantly (which porting 98 billion calls (10 times as (NSF) and the Bureau of Labor Sta- would imply a shortage) nor down many). Should we have protected tistics confuse the situation. In my significantly (which would imply an those jobs? I don’t think so. view, the term “computer science” oversupply). They are 1.5 to 2 times Although we can have empathy for further confuses things (even the starting salaries for new B.A.s, those who lost their jobs, and we though it is called a science, I believe and the average engineering salary perhaps have a societal respon- it is [mostly] an engineering disci- is comparable to a new lawyer’s sibility to retrain them, consumers pline). We need to be very careful starting salary. This suggests to me are the big winners—a 40-fold to use consistent, comparable data. that employers see value in U.S. decrease in the cost of telephony Third, I know much of the hard engineers. and a host of new services. data we have is reported two years Eighth, engineering unemploy- late. That really matters in the cur- ment is up from 0.5 percent to Some Things I Believe rent context because of the “dot around 6 percent—at least in First, I believe that engineers are com” boom/bust. some fields. That is comparable to not telephone operators! U.S. pros- Fourth, I know that the number unemployment in the rest of the perity, security, and health depend of new engineers per year in the workforce—which is unusual and on technology created by engineers, United States declined from about worrisome because historically and that dependence is evident 80,000 in 1985 to about 65,000 in engineering unemployment has in the 20 great achievements I WINTER 2004 49

mentioned earlier. I also believe 10,000 miles and 12 times zones however, that the total number of that this dependence will increase. away, then they will outsource—and undergraduates has increased in the And this dependence on engineer- they should! The problem is the intervening 20 years. So, although ing fundamentally changes the nation’s access to engineering talent, the absolute number is up somewhat, nature of the discussion, from empa- and the individual manager in an it is still a decreasing fraction of the thy for individuals to concern for individual company is not responsi- total number of undergraduates. our economy, our security, our ble for solving that problem. But Second, the percentage of under- health, the environment, etc. those are all big “ifs,” and there is graduates in engineering (between Second, I strongly believe two some evidence that each has been 4 and 5 percent) is the second low- things that are superficially incom- wrong in specific cases. est in the developed world. By way patible. One is that human capital What bothers me most is the of contrast, the percentage is 12 per- is precious—when we lose an engi- long-term protectionist approach cent in the United Kingdom and neer we diminish that stock. The advocated by some, even some pro- more than 40 percent in China. other is that leaving engineering is fessional societies. Some form of Third, as NAE members already not, in itself, a failure of either the protection might or might not be know from my newsletter, I am individual or the profession. An appropriate in the short term, but deeply concerned about the increas- engineering education is great it is not a long-term solution in ing difficulty of students and senior preparation for many careers. In my view. The world is changing— scientists and engineers getting fact, I wish we had more engineers or, perhaps more accurately, has visas to enter the United States. in other professions; from where I sit changed. The familiar and com- Very briefly, in the wake of 9/11, these days, I especially wish we had fortable world of mass production the U.S. government instituted new more engineers in politics! manufacturing by predominantly policies for issuing visas, which was Third, I believe that the aggregate nationally owned and operated both inevitable and appropriate in data and discipline-specific data are companies is gone. In the long some ways. However, I believe the quite different. That matters term, we have to prepare our engi- way this has been done has been because engineering is changing, neers to operate in this new envi- counterproductive. and we need to be able to disaggre- ronment in a way that the “ifs” I In the past, the best and brightest gate the trends. listed are not true. We must also in the world came to the United Fourth, I don’t believe in a pro- have empathy for those who are dis- States to study, and many remained tectionist approach to solving the placed from their jobs and help here to contribute their talents to problem of access to engineering tal- them prepare for new ones. improving our way of life. In the ent. Protectionism might be OK in past, senior and respected scholars the short run, but it’s not in the The Mosaic came to visit U.S. universities, to long-term interest of the country. It The proposed protectionist lecture, and to share their knowl- would not have been good for the approach as a long-term solution to edge. In the past, major interna- country to protect those telephone the problem of outsourcing is part of tional conferences were hosted in operators’ jobs I mentioned earlier, the mosaic I mentioned earlier, a the United States, giving our scien- for example. We need to think of mosaic that presents an image of tists and engineers privileged access positive, constructive alternatives to increasingly short-term thinking to the most recent developments ensure the nation’s access to the and a lack of balance. Let me sug- around the world. In the past, the engineering talent it needs. gest some other pieces of the mosaic image of the United States was of a Finally, I believe that, for the most that relate to engineering. free and open society—a land of part, managers make rational deci- First, the absolute number of engi- opportunity. sions. If they can get comparable neering graduates has declined. As I We have not completely elimi- talent at one-fifth the cost in India, said earlier, we graduated about nated those “in the past” items, but and if the start-up cost is small, and 85,000 in the mid-1980s; the num- we are coming perilously close! In if the cost stays small, and if the ber dropped to about 65,000 in the an attempt to make ourselves safer, productivity per unit cost is high mid-1990s, but has recently risen to we have both prevented the world’s enough, and if they can manage from about 75,000. One must remember, best and brightest from attending The 50 BRIDGE our universities and alienated them increase at NIH was bigger than the they simply do not have the knowl- and senior scientists and engineers entire NSF budget. As I grow older, edge they need to function in a in the process. I do not begrudge NIH a nickel of technological society. Fourth, after three decades of that. But I think the imbalance I could describe more pieces in my steady progress in engineering for reflects a lack of appreciation of how mosaic, but I will get too depressed. women and underrepresented much our society—even medical minorities, in the early 1990s, the technology—depends on the con- The Whole Picture percentage of women and under- tributions of the physical sciences Each piece in my mosaic tends to represented minorities declined. It and engineering. I also think the be treated as a separate problem— then rebounded a bit . . . but has growing imbalance reflects a short- and taken alone is, perhaps, not all essentially been flat for more than a term focus on the “disease du jour” that important a problem. But they decade now. and a failure to appreciate the all have one property in common, Fifth, for most of our history, impact of long-term increases in short-term thinking. None of them higher education was considered a knowledge and skills in all of the sci- addresses the long-term ramifica- public good, that is, it was consid- ences on our quality of life, includ- tions of the overall problem. For ered good for the country to increase ing the eradication of disease. me, viewing them in this overall the number of educated citizens. Finally, we live in the most tech- context makes them all seem urgent. Before becoming president of NAE, nologically dependent society that Some of these problems we can I was a professor at the University of has ever existed. Yet the vast major- do something about and some Virginia, which was founded by ity of our population is technologi- we can’t—at least not directly. For Thomas Jefferson, who believed cally illiterate—not dumb, just example, we can help make engi- that we could not have a democracy uninformed. I mentioned Thomas neering more attractive to U.S. without an educated citizenry. In Jefferson’s opinion about the need for students. NAE has established the the middle of the nineteenth cen- an informed citizenry—well, I think Public Understanding of Engi- tury, our national leaders created the he would be worried. Sitting, as I do, neering Program to create a better land grant colleges, because they felt at the nexus of science, engineering, (and more accurate) image of engi- we could not have prosperity with- and public policy, I am worried! neering. We can also do something out an educated workforce. Until Despite our efforts here at the about engineering education to recently, states have supported state National Academies, I see policy improve the curriculum and peda- universities, provided scholarships, makers every day expounding on gogy. It is a disgrace, for example, and kept tuitions low to make national and homeland security, that only about half of the students higher education accessible to all. energy, and dozens of other tech- who enter engineering programs But something has changed! We nical issues about which they finish them. The students who no longer seem to consider higher haven’t a clue! I interact with jour- leave are not poor students; in education a public good, but a pri- nalists who, in my view, miss the fact, in terms of GPA, grades in vate one. State support for universi- real issues in a story because they math and science, SAT scores, rank ties is dropping like a stone, tuitions have no knowledge of its technical in high school, they are indistin- are rising, and student loans have dimensions. I put up with advo- guishable from those who stay. replaced scholarships in many states. cates of all stripes proposing tech- They are being driven out by our All of these trends reflect a largely nologically nonsensical solutions to curriculum and pedagogy. And that unstated view that shifting costs to very real problems. I worry about is entirely fixable! students is OK because only students people having to make personal We have little direct leverage benefit from higher education. It’s decisions—whether or not to dis- over the visa situation, but we have the “only” in that last sentence that arm an air bag, for example—about direct and personal knowledge of worries me. which they have fragmentary infor- the benefits foreign students and Sixth, funding at NIH has dou- mation and no disciplined thought scholars bring to the United States. bled to almost $30 billion; at process on which to rely. Mostly, As an organization, NAE is trying to the same time, funding at NSF is these are good, intelligent people convince policy makers of that. In $4.25 billion. The last yearly who want to do the right thing, but addition, each and every one of us WINTER 2004 51

can write to our representatives and expect the other 95 percent of stu- into depth on them individually, I demand action. dents to become technically literate? see them as pieces of a mosaic. And We can do much the same with I believe engineering schools must the composite image of the mosaic respect to other issues, like the accept the responsibility for the tech- is one of disinvestment in the future. imbalance in funding between the nological literacy of all students on Each piece of the mosaic raises a life sciences and the physical sci- campus. We cannot force them to topic of concern to engineers, and ences and engineering. And each of take engineering courses, of course, directly or indirectly, we can do us, in our companies or schools, can but we can at least make the courses something about them. Engineers take ownership of the diversity issue. that are available accessible to the have had a profound impact on soci- I want to make a special plea on average liberal arts major. In fact, I ety through the products, processes, the issue of technological literacy. think we have an obligation to do so. and infrastructure we developed. Engineering schools tend to see Now it is time for us to contribute in themselves as professional schools, Conclusion another way. and they only offer courses for engi- Although I have touched on a neering majors. So just how do we broad range of topics and not gone

Remarks by Eli Ruckenstein, 2004 Founders Award Recipient

The early days of my career coincided with the early days of communism in Romania. I was not a Party member, and I did not adjust easily to the very political tone of the times—which will not surprise many of you. I worked in a scientific community that had few if any connections to the outside world, and I would never have imagined, even in my most ambitious dreams, that I would find myself experiencing a moment like this, addressing you on this occasion. Over the course of my career I worked on a broad range of problems, many of which are conven- tionally associated with subfields of (Left to right) Wm. A. Wulf, Eli Ruckenstein, Craig Barrett, Christopher Magee. physics, chemistry, and engineering. The fact that I am here addressing The 2004 Founders Award was I feel honored, humbled, and at you on this special occasion under- presented to Eli Ruckenstein, Dis- the same time overwhelmed to scores the interdisciplinary nature of tinguished Professor of Chemical receive this year’s Founders Award, our field. Engineering research is a Engineering, State University of which puts me in very distinguished synthesis, which implies learning New York, Buffalo, for “leader- company. whatever experimental or theoreti- ship in modernizing research and I was born in a small town in cal techniques are required to solve development in key areas of chem- northern Romania in an atmosphere the problem at hand, many times ical engineering.” These remarks that might remind you of the novels forays into distant disciplines. were delivered October 3, 2004, of Gogol or Turgenev, and I was the My own way of doing research was during the NAE Annual Meeting. first one in my family to go to college. to a large extent defined by the fact The 52 BRIDGE that I am basically self-taught and In the area of catalysis, my stu- taught me, inspired me, and stimu- that I don’t belong to any school. I dents and I studied theoretically lated me. I would like to share this was educated during a special time in and experimentally the stability of award with them. the history of Romania, my native small metallic clusters on catalytic I consider myself a very lucky man country, immediately after the Sec- supports. The other area of cataly- who has been surrounded by many ond World War. A tradition in pure sis I was fascinated with was the guardian angels. I would not have mathematics had existed for some kinetics of selectivity of supported survived my early days as a young time, and physics and chemistry catalysts. I also formulated a theory assistant professor at the Polytechnic benefited from the return of a few for the mechanism of oxidation by University in Bucharest without Pro- distinguished researchers who had mixed oxides. fessor Bratu, who protected me from been trained in the West. However, In the area of colloids, we devel- my own inability to fit into the polit- Romania had no tradition in modern oped a hydrodynamics of colloidal ically driven academic environment engineering, and much of my educa- particles that accounted for double- of communist Romania, where I was tion consisted of studying the avail- layer and van der Waals interactions. considered a dangerous reactionary. able original literature on my own. We introduced the concept of I also want to thank Jim Davis, My early work on heat and mass “interaction force boundary layer.” Bill Gill, and the Chemical Engi- transfer—which didn’t get to the We also proposed thermodynamic neering Department at Clarkson West until 1958—owes much to the theories of surfactant aggregation, University. They are responsible for encouragement and understanding microemulsions, and liquid crystals. my initial move to the United of my teacher and mentor, Professor I feel happiest about two contribu- States, and I owe them all a debt Bratu. In spite of uncertainties tions: (1) the reformulation of the of gratitude. under the Communist regime and classic theory of double-layer forces I also owe a debt of gratitude to the fact that I had no formal author- to include in a unified way the inter- Art Metzner and my colleagues ity, I managed to attract a group of play between double-layer hydration at the University of Delaware who somewhat younger colleagues with forces and the hydration of ions; and welcomed me and my family whom I cooperated on a variety of (2) a unified kinetic approach to warmly, made us feel at home, and problems, from mass and heat trans- nucleation and growth. helped me through the years of fer to kinetics of gene expression to Rather than boring you with adjustment to the American way of interfacial phenomena and thermo- descriptions of other theoretical life and the American academic dynamics of small systems. efforts, I will mention some of our system. I will never forget their When I arrived in the United more recent technological innova- moral support for me and my wife States at the age of 45, everything I tions: the development of a solid- during the two and one-half years had learned about the West was from solution catalyst for CO2 reforming we struggled to get our then-teenage a limited number of journals and an of methane; the preparation of some children out of Romania. even more limited number of con- interesting compounds for the stor- I also want to acknowledge tacts with researchers from outside age of H2; and the preparation of a George Lee, Bill Gill, Ralph Yang, Romania. I was immediately struck, paste with high thermal conductiv- Carl Lund, and my colleagues at Buf- overwhelmed, by the amount and ity, which is now used in all IBM falo for providing the resources and breadth of information that became computers. In addition, from con- helping create the supportive atmos- available to me overnight. At that centrated emulsions we developed phere that made my last 25 years the point, I really started a new career. various technologies for preparing most productive, enjoyable, and Since then, my work has extended polymers, conductive polymers, and rewarding of my career. in roughly five different directions: membranes for separation processes. Someone who deserves a special catalysis, colloids, separation of pro- My career has been and remains a mention is Howard Brenner, who teins, polymers, and material sci- great source of satisfaction to me. was instrumental in my getting my ence. Don’t worry, I’m not going to For this, I must acknowledge my stu- first position in the United States. I give a lecture on each of these direc- dents, postdocs, and collaborators am deeply thankful to Howard for tions. I just want to highlight the who, through their hard work, his help, encouragement, and friend- contributions closest to my heart. patience, and dedication, have ship over the years. WINTER 2004 53

Finally, I would like to thank my have had, and I certainly would not energizes me and encourages me to family, especially the person who be here today. continue searching. I started my has been the greatest influence in Science has been my life, and I career in those dark days in Roma- my life, my wife Velina, for her con- have been lucky to have the oppor- nia with little but ambition and stant support, love, and selfless ded- tunity to wake up every morning youthful exuberance on my side. I ication during our 56 years together. excited about my work and curious now live in a free country, and I still Without her, I would not have the about the next adventure. On my have the ambition—and on a good wonderful children I have, I would next birthday, I will turn 80, and I day a little exuberance. not have as many friends as I have, still hope to find that intangible Thank you again for this honor, I would not have had the career I Holy Grail. Your recognition tonight and thank you for your attention.

Remarks by John Brooks Slaughter, 2004 Bueche Award Recipient

strength; I dedicate this award to her. I had the privilege of getting to know Art Bueche during the last few years of his life. He was extremely helpful and supportive when I took on the responsibilities of the National Science Founda- tion, and he was a wise and gener- ous mentor. I cannot begin to express my humility as I receive this award recognizing his contributions as a statesman and an advocate for science and technology. It is also with a sense of humility that I join the ranks of distinguished 2004 Bueche Award presented to John B. Slaughter. (Left to right) Christopher Magee, Wm. A. Wulf, John B. engineers and public servants who Slaughter, Craig Barrett. received the Arthur M. Bueche Award in the past. I am truly hon- The 2004 Bueche Award was I wish to express my appreciation ored to have my name added to presented to John Brooks Slaugh- to Dean Terry King and his associates the list of luminaries from our profes- ter, president and CEO, National from the College of Engineering at sion who preceded me. Although I Action Council for Minorities in my alma mater, Kansas State Univer- hardly deserve to be in their company, Engineering, for “support of engi- sity. Their persistence and tenacity I am grateful to the National Acad- neering research and education made this award to me possible. I also emy of Engineering for this honor. within the National Science Foun- want to thank everyone who wrote I am not joking when I say I was dation, many contributions to the letters and supported my candidacy the first black engineer I ever met. development of science and tech- so earnestly. Most of all I want to African Americans and Latinos in nology policy, and lifelong dedica- thank my wife, Bernice, and my chil- Topeka, Kansas, in the 1940s and tion to increasing diversity in the dren, John Jr. and Jacqueline, whose 1950s were hardly encouraged to disciplines of science and engineer- sacrifices, support, and love have achieve in science or engineering. ing.” These remarks were deliv- sustained me throughout my profes- Nor was Topeka a city with much ered on October 3, 2004, during sional career. Bernice has been my that could be characterized as scien- the NAE Annual Meeting. cornerstone, my fountain of inspira- tific or technological, with the tion, and my unending source of exceptions of Menninger’s for the The 54 BRIDGE former and the Santa Fe Railroad for supported and encouraged by a ded- participation of any of our country’s the latter. Children in Topeka were icated faculty and staff, and are suc- most valuable resource—our youth.” unlikely to have neighbors or uncles ceeding. Although much remains Although I am encouraged by the who were scientists or engineers to to be done to eliminate disparities number of students in engineering inspire them to take courses in math in access and opportunities, the from historically underrepresented and science or introduce them to same thing is happening on many minorities, I remain discouraged by the mysteries of nature and campuses across the country. the small number of minority faculty machines. Consequently, it was My organization, the National members in science and engineer- uncommon for youngsters, espe- Action Council for Minorities in ing. The argument often advanced cially minority youngsters, to want Engineering (NACME), is working by colleges and universities that to become engineers. My hardwork- to increase the number of educa- there aren’t enough minority Ph.D.s ing, loving parents kindled and kept tional institutions committed to in science and engineering in the alive my spark of interest in engi- improving their capacity to identify, pipeline is not as tenable as it was a neering. They kept it from being enroll, educate, retain, and graduate few years ago. Unfortunately, only snuffed out by the indifference and minority engineering students. In about 1 percent of engineering fac- low expectations of a public school 2003, NACME joined with the ulty nationwide are African Ameri- system whose leaders were con- National Academy of Engineering, can or Latino, a situation that has sumed with defending at all costs the National Academy of Sciences, not improved in the past 20 years, the tradition of segregated schools. MIT, Stanford, IBM, and DuPont in even though the number of black This year we have been reminded of filing an amicus brief in the Univer- and brown Ph.D.s in science and the demise of legally sanctioned seg- sity of Michigan affirmative-action engineering increases every year and regated schools as we celebrate the case before the U.S. Supreme Court. more and more of them are available 50th anniversary of Brown vs. the The statement expressed our con- and fully prepared for faculty Board of Education of Topeka. viction that “access to the educa- appointments. My eyes were opened to the tional opportunities available in our Nevertheless, I am hopeful—no, opportunities and excitement of nation’s selective universities is I am confident—that the American engineering at Kansas State Uni- essential if underrepresented minori- higher education community will versity. In 1956, I was the only ties are to have an opportunity to accelerate its efforts to create a more African American graduate in engi- contribute to strengthening Ameri- inclusive and pluralistic environ- neering at the university. On a ca’s scientific and technological ment. Then, and only then, can it recent visit to the campus, I was capabilities.” The statement went assist our nation in living up to its pleasantly surprised to meet a large on to say, “We can no longer afford promise and ascend to a higher and group of African American and the potential loss of creativity, pro- nobler plateau, a place where both Latino engineering students who ductivity, and talent that results from excellence and equity reside. are enjoying their experiences, are policies and practices that inhibit the Thank you! WINTER 2004 55

Opportunities and Challenges for Engineering Education as Seen from an Interdisciplinary Telecommunications Master’s Degree Program

and some serious problems in the the program had to be open to way the current social structure of essentially any student with a bach- engineering schools limits the diver- elor’s degree. This forced us to sity and quality of the programs we develop the program that had very can offer our students. minimum prerequisites and took advantage of the students’ maturity. The Interdisciplinary Telecom- With the limit of one to two years munications Master’s Degree for a full-time student to finish the Program at the University of program and considering the Colorado at Boulder breadth of subject matter to be cov- In the late 1960s and early 1970s, ered, the courses could not require Frank S. Barnes many students considered engineer- four years of prerequisites in any ing “bad” and responsible for many subject. The required math back- The 2004 Gordon Prize was pre- of the world’s ills. Enrollments were ground, for example, was limited to sented in February 2004 to Frank down, and I was looking for a new algebra, trigonometry, and calculus. S. Barnes, Distinguished Profes- way to attract students to electrical To add depth, we required a thesis sor, University of Colorado at engineering. In a conversation with so that the student had some depth Boulder, for innovations in engi- Dr. John Richardson, I learned of a of knowledge in at least one subject neering education. This lecture need for people who not only under- and had shown an ability to write a was delivered on October 3, 2004, stood the limits of communications coherent document. during the NAE Annual Meeting. technology, but also understood the In 1971 the program started with Today I wish to describe an inter- economic and political realities of a single student with an under- disciplinary master’s degree program how the Federal Communications graduate degree in international in telecommunications that Profes- Commission (FCC) and others were relations. That student is now an sor George Codding and I started determining the shape of the indus- attorney with a practice in tele- back in 1970. The program was try. John also pointed out to me that communications. Over the next designed to meet a significant Professor George Codding, in the 30 years, the program grew to a peak worldwide need for students who Political Science Department, was a enrollment of about 500 students— not only knew the technology of world expert on the International half of whom were participating telecommunications, but also had Telegraph Union (ITU). via distance education. About half acquired a good understanding of The result of George and my dis- of the students have engineering the legal, political, and economic cussions was a proposal for a new degrees, and half have undergrad- aspects of the industry. The pro- M.S. degree funded in part by the uate degrees in a wide variety of gram does not fit the normal mold National Science Foundation that other disciplines. The program now for an engineering program, but it included business, political science, offers more than 20 courses, approx- has enabled more than 2,000 stu- sociology, and electrical engineer- imately half of which are technical dents to play significant roles in the ing. The program had several new and half of which cover economics, operation, management, and growth features. First, it was the first pro- business, law, and policy. The cur- of the telecommunications industry gram that I know of that had a rent requirements are given in over the last 34 years. It has also serious technical content for elec- Table 1 below; brief course descrip- served as a model for many other trical engineers that also included tions are given in Appendix I. programs both in the United States law and policy as a fundamental The electives include a wide and around the world. part of the program. Second, none range of topics, from wireless and With this program as a model, I of us wanted to teach a service optical fibers through protocols will describe some new opportunities course for more than a few years, so and security to standards and The 56 BRIDGE

TABLE 1 program in telecommunications. First, many students do not know at • The M.S. degree in telecommunications in the School of Engineering and Applied age 18, when they leave high school, Sciences at the University of Colorado at Boulder requires 36–37 master’s credits. that they will need a technical back- •Two courses: TLEN 5300 and 5836 are designed to improve the proficiency of ground to access many of the jobs students in math, writing, and presentations skills. These courses are prerequisites that could lead to satisfying careers but may be waived by the faculty advisor. These credits are not included in the required total for the degree. for them. Even after they graduate from college, it may take several Degree requirements include: more years for them to learn that 1. Four technical courses for a total of 12 credits, including TLEN 5310 (Telecommu- they need a technical background. nications Systems) and TLEN 5330 (Data Communications 1) and two additional But by then, they may not be able to 3-credit courses. take four years to start over as a 2. Two economics or business courses for a total of 6 credits, including TLEN 5010 freshman in engineering, math, or (Telecommunications Finance 1) and one additional business or economics course. physics. However, we have found 3. One policy and law course for a total of 3 credits. that they can learn, in just one or 4. One programming course for a total of 3 credits. two years, much of what they need 5. Two telecommunications seminars for a total of 2 credits. to operate and manage major communications systems. These stu- Two options for final requirements: dents often bring verbal and people 1. Two research seminars for a total of 2 credits. skills with them that many engi- 2. Three additional 3-credit courses electives for a total of 9 credits. neering students do not have, and or these skills often lead them into top management positions. 3. One elective course for a total of 3 credits. Many of the toughest problems 4. TLEN 5700 research course and 6 thesis credits (may substitute 3-credit research engineers face today involve dealing methods course and 4 thesis credits). with the public to implement technical solutions to societal problems. telecommunications law. These Cox with a nontechnical back- Government bodies, including the options make it possible for students ground who, more than 20 years FCC, the ITU, and state public util- to develop strengths in their respec- later, is using material he learned in ities commissions, set the rules by tive areas of interest. class on characteristics of communi- which competitors can compete. Courses such as the required cations systems; a young lady in These rules, in turn, often determine course for some programming can Mexico who took my course in opti- which technologies are used and be waived if the student already has cal fibers who now holds a position developed and who gets rich or goes a background in that area. In gen- corresponding to an FCC commis- broke. For example, look at the cur- eral, I advise students with non- sioner and who is one of a very few rent growth in mobile telephones. technical backgrounds to take engineers in Mexico with some This technology was held up for technical courses beyond the background in policy and econom- more than 20 years in the United requirements—especially the lab. I ics; a recent graduate who has been States because the bandwidth was advise students with engineering implementing a telecommunica- allocated to UHF TV. Another backgrounds to take additional tions system that allows doctors at example is the current rules on the business and policy courses. the Harvard Medical School to universal service fund relating to In my view, this program has been prescribe drugs directly to nurses in subsidies for rural telephones and remarkably successful in making it the slums of Lima, Peru, and has cut voice-over IP. Another class of prob- possible for a wide variety of stu- the error rate in drug delivery from lems shows up in locating cell phone dents to pursue successful careers in 17 percent to less than 5 percent. base stations in places where there fields that would not otherwise have It is useful to reflect on some of are public concerns about possible been open to them. Examples the things that we have learned in health effects, aesthetics, and possi- include: a senior vice president at the 34 years of teaching this M.S. ble reductions in property values. WINTER 2004 57

Similar problems occur in the power without decreasing the quality of upper-classmen to review and under- industry related to who builds trans- the technical education already stand the fundamentals of classes mission lines in a competitive envi- offered. As it is, many technical they have already taken. Teaching ronment and where power plants are concepts students should know can- could be carried all the way up located. In civil engineering, there not be covered for lack of time. through the senior year, with fifth- are related problems in managing And students often do not master year students and graduate students river basins, sewer systems, and traf- the material as completely as we helping teach upper-class courses. fic. In all of these cases, engineers think they should. There will be fluctuations in quality, need to know more than the tech- The key to adding new material is as always. However, seminars on nology if they are going to take on twofold: (1) improving our choice teaching, how people learn, ethics, leadership positions. of materials so students leave our and sexual harassment based on program with a solid technical and problems encountered in actual Some New Directions for liberal arts foundation and the abil- teaching experiences might well be Engineering Education ity to acquire new skills rapidly on worked into the curriculum. Based on our experience with the the job; and (2) improving the effi- A second approach is for some telecommunications program, I sug- ciency with which the students mas- classes to be taught jointly across gest that we have the ability to open ter the material presented in class. continental boundaries. Our stu- new opportunities for our engineer- Each faculty will have to make its dents are going to have to compete ing undergraduates and M.S. stu- own choices about how to do this. in a global economy, and more and dents by providing them with a However, I would like to suggest more design teams are going to be background in public policy and an some options that I hope we will be geographically dispersed. A number overview of major worldwide prob- able to implement, at least in part, of schools already team students lems in infrastructure. Engineers at the University of Colorado. from schools in Europe and the need more than technical compe- A first step is motivating students United States who must work tence to satisfy the needs for energy, to become deeply involved in a sub- together to solve a common design communications systems, and con- ject to the extent that they will problem. At the University of Col- trol of environmental pollution. spend more time on it than it takes orado, we have two start-up courses, Courses on risk analysis, environ- to earn a passing grade. This can be one on advanced aircraft design and mental law, resources and popula- done in many ways—by demanding another on embedded systems. tion growth, and security would all more with tough tests and frequent These courses will be jointly taught be good electives or aspects of the quizzes or by getting students in Europe and at the University of social science requirements. Many involved so that they are self- Colorado with students in both major challenges for engineers in motivated by interest in the subject places receiving some lectures from the future will require solving polit- or by peer pressure. each school. Challenges in working ical and economic problems in One possible approach is to get across cultures and time zones addition to technical problems. students involved in teaching other should help prepare students to Technology will be only a part of students on a systematic basis, work on worldwide design teams the problem, and often the easy because you learn a subject more after they graduate. part. For example, building a thoroughly when you try to teach it. We are also developing labs that nuclear power plant is much easier Because students do not want to be can be operated over the Internet so than persuading people to let you embarrassed in front of their peers, that students can do at least part of build it in their “backyard.” they prepare more thoroughly for their lab work from anywhere and It is also much easier to get fa- teaching; they also learn organiza- at times of their choosing. Many of culty agreement that a new course tional and communications skills. the experiments in the telecommu- would be valuable for some of their Thus, we might have sophomores nications lab are being made avail- students than it is to remove a or juniors running help sections able this way (the use of routers, course from the curriculum. The for the freshman calculus classes. etc.), and a lab in optical communi- trick in adding any new material This would give the freshmen more cations is being jointly developed at to the curriculum is fitting it in personal attention and force the CU and the University of Houston. The 58 BRIDGE

These labs will have the advantage of department consider important. 3. Using student evaluations to increasing the hours available for Often the time taken to develop a determine raises and promotions using expensive computer-controlled course for another school or pro- is often destructive. Student rat- equipment. In addition, no school gram is considered a waste of ings are frequently a measure only can either afford, or find, all of time. Finally, students in these of how well a course matches the the faculty they would like to courses are seldom in a position to student’s expectations. If a course have. Using telecommunications contribute to Ph.D.-level research is much more demanding than to share faculty and import course in the home discipline. Thus, expected or if one gives tests with material may be one way to working on an interdisciplinary low exam scores (independent of strengthen programs. program at either the undergradu- the grades at the end of the term), The next step in improving our ate or master’s level may be haz- one has a high probability of teaching efficiency will require a ardous to an assistant professor’s receiving low student ratings. better understanding of how stu- promotion and tenure. Students are seldom in a position dents learn. This understanding is to evaluate the importance of the 2. A second major hazard is the likely to be based on an improved material until well after they grad- amount of time it takes to raise understanding of how information is uate, and the choice of material is research funding. With current stored and processed in the brain, as one of an instructor’s most impor- funding rates for grants from NSF well as by acquiring information tant functions. In my view, stu- and NIH running at 3 to 20 per- from psychology and educational dent ratings can be used in about cent, 80 to 97 percent of the fac- theory. For example, a Russian pro- 10 percent of cases to improve ulty are wasting their time. gram that builds a profile of a stu- instruction, but they should have Imagine going to class on the first dent’s learning style by looking not little effect on promotion, tenure, day and telling students that, on only at the kinds of wrong answers and raises. In too many cases, the average, 80 to 97 percent of the student gives but also the time instructors are less demanding them are going to fail the course! delay in their decision making than they should be, thus sending To me this suggests that many caught my attention when the Rus- the wrong message about how faculty would be better off doing sians claimed they could reduce the courses should be taught. the research they can with lim- time required to learn a course like ited or no funding or spending geography by a factor of three over their time on improving the Summary standard results. courses they teach. We were fortunate to develop an Problems in Developing New The audience here today is interdisciplinary program in Educational Programs made up largely of those who telecommunications at the right have succeeded in raising the time that has enabled many techni- 1. Educational institutions and fac- money they needed to do impor- cal and nontechnical students to ulties are extremely conservative tant things. In most cases, this pursue successful careers that would and it is very hard to initiate a has been done at a cost of lots of not have been available to them program that does not fit into the time. It is not unusual for suc- otherwise. I believe that this pro- current departmental structure. It cessful senior faculty members to gram can serve as a model for the is even more difficult to keep it spend 50 percent of their time development of programs in other going, even if the students are writing grant proposals and fields that will also allow students to very successful, if the program reports or attending contractor take leadership roles in solving covers a broad range of material meetings just to keep their pro- major infrastructure problems and does not include a hot grams funded. This suggests that throughout the world. research topic that is well funded. we have done a poor job of sys- In closing, I would like to thank In addition, faculty may have a tems engineering in distributing the National Academy of Engineer- difficult time getting tenure if research money and that some of ing, Bernard Gordon, and Bill Wulf they are not doing research that the leadership in this room should for giving me the opportunity to fits into a standard area that make a serious effort to improve speak to you this afternoon. I also senior members in their home the system. want to express my deep appreciation WINTER 2004 59

to the many students and faculty and international aspects. Includes Satellite Communications members who have made our pro- telephone, cable, satellite, cellular, Discusses fundamental concepts gram a success. Without their and other wireless systems and the of communication satellites includ- accomplishments, there would have Internet. ing orbital mechanics, spacecraft been nothing for me to report. and earth station system configura- Telecommunications Theory and tion, link budgets, propagation Appendix I Applications issues, modulation and multiplexing Provides necessary mathematical techniques, multiple access schemes Telecommunications Economics and engineering background to (FDMA, TDMA, CDMA), orbit Introduces students to a range of study telecommunications, includ- selection, error control coding, and microeconomic principles and mod- ing physical units, numbering sys- satellite network architecture. els used to undertake economic and tems, trigonometric functions, sine financial analyses of telecommuni- waves, attenuation, logarithms, IP Routing cations and e-commerce networks. indices, decibels, complex numbers, This course breaks IP routing calculus, elementary probability, technologies into two fundamental International Telecommunications and power and circuit analysis. pieces: an in-depth study of Interior Strategy and then Exterior Gateway Proto- Telecommunications Systems Covers issues involved in devel- cols (IGPs and EGPs). oping a business plan for a telecom- Provides an in-depth look at munications investment in markets basic telecom terminology and Data Communications 2 outside the United States, including concepts, introductions to voice Expands on the topics associated estimating revenues and costs, and data networks, signaling, and with Internet protocols, processes, sources of data, network costing modulation/multiplexing. and selected applications. Intro- issues, and interconnection. duces Internet-related protocols and Data Communications 1 applications, including the applica- Telecommunications Business Strategy An up-to-date technical survey of tion layer (HTTP, FTP, SMTP, Covers concepts, strategies, and data and computer communications DNS, and sockets), the transport practical implementation of market- starting with the basics and includ- layer (UDP, TCP), routing (RIP, oriented business strategy in the ing details about Wireless, MAN, OSPF, BGP), IP version 6, multi- telecom industry with plenty of real- and WAN systems and standards media (RTP), network security world examples. that include T-carrier, SONET, (IPSec), and network management HDLC, Frame Relay, and ATM. (SNMP). Standardization and Standards Wars Provides students with a tool set Telephony Multimedia Networking of economics, business, and legal Covers fundamental require- Focuses on how multimedia ser- principles involved in the tech- ments for voice communication and vices and networking technology nology standardization process and processing, multiplexing, synchro- are co-evolving to support conver- covers topics including the role nization, switch architecture, and gence, with Voice-over-IP devel- of various standards models (e.g. subscriber access. Transport and oped in detail. open vs. closed, proprietary vs. non- access technologies discussed proprietary, and de facto vs. de jure), include ISDN, frame relay, asyn- Advanced Telecommunications Lab as well as a survey of international chronous transfer mode, ADSL, Students work in teams to learn standard-setting organizations and SONET, and signaling system no. 7. the fundamental techniques of sig- their respective roles. Packet speech services, protocols, nal transmission and impairment and operation over frame relay, measurement, voice and data Telecommunications Law ATM, and Internet protocol are switching, systems administration, Examines laws governing tele- also considered. and the fundamental functions of communications industries, includ- data networking and services. ing federal and state regulation The 60 BRIDGE

Signaling Systems Network Security Lab Unix, C, C++ Develops an understanding of Lab work covering system hard- This course is recommended for modern signaling protocols and dif- ening, firewalls, intrusion detec- students who anticipate a need for ferences among these protocols, tion, vulnerability assessment, and knowledge of both UNIX/Linux including an appreciation of actually investigation. and C programming in their future implementing signaling protocols in careers. the Internet Protocol environment. Foundations of Computer and Network Security Applied Network Security Wireless and Cellular Introduces students to computer Technical discussion of threats, Communications and network security, the branch of vulnerabilities, detection, and pre- Covers miniaturization, clever computer science that studies meth- vention is presented, including use of radio spectrum, the expanded ods and techniques to protect criti- issues such as cryptography, fire- roles of the cellular phone and cal information resources. walls, network protocols, intrusion pager, PDAs, the popularity of the detection, security architecture, IEEE802.11 networks, the new Spectrum Policy and Management security policy, forensic investiga- breed of low earth orbiting satellites, The radio spectrum is a natural tion, privacy, and the law. and even the Global Positioning resource that is increasingly crucial Systems (GPS). to the economic and social well- Law for the Information Age being of all people. This course This seminar addresses the Telecommunications Seminar focuses on how this resource is being numerous legal issues that arise in National and local speakers give managed in the face of escalating relation to the Internet, in parti- lectures on technical, business, pol- demand, rapidly changing technol- cular issues raised by telecom- icy, and regulatory issues in telecom- ogy and increasing globalization. munications regulation, intellec- munications. tual property, and antitrust and Privacy, Security, and Digital Rights constitutional law, highlighting the Capstone 1 Management continuing challenges posed by Capstone 1 develops basic quan- Introduces students to the laws technological change for economic titative and qualitative research that regulate the basic technologies regulation. techniques. of the Internet and the management of information in the digital age. Research Techniques Network Management and Provides both quantitative and Operations Capstone 2 qualitative techniques for use in Covers both technological and Requires teams of 3 or 4 students capstone projects, master’s theses, managerial aspects in network man- to select a topic, research and docu- and Ph.D. dissertations. agement through an examination ment this topic through the guid- of the capabilities and needs of ance of a faculty member, and Video and Technology organizational network and systems present their findings to their fellow Students will obtain an in-depth management, including network students in the form of a 20-minute knowledge of the specifics of various management standards such as defense (including time for ques- technical aspects of television and SNMP, RMON, and CMIP. tions and discussion) and a support- visual human perception and how ing paper. image transmission and delivery fit Wireless LANs into the telecommunications indus- Discusses packet wireless data Telecommunications Application try market mix. communication with an emphasis Programming on wireless LANs. Students learn server-side, object- Leadership and Management oriented programming as a cross- Provides a background in leader- platform application development ship philosophy, principles, and environment. practices to enable new managers to develop skills in leading and WINTER 2004 61

managing, including organizational Microeconomics (Telecom) Telecommunications Finance 2 leadership, leadership styles, leader- Introduces economic theories of This course introduces a rela- ship tools, roles of effective leaders, consumer choice, production, firm tively new methodology, real change management, and business and market organization, and gen- options, to evaluate investments ethics. Same as EMEN 5050. eral equilibrium and applies eco- and projects to account for uncer- nomic models to telecommunica- tainty, with application to telecom- Optical Communications tions demand, costs, pricing and munications investments. Lectures, readings, and projects in strategic interaction, and network the field of optical communications externalities. Independent Study systems theory and practice, includ- An opportunity to study an area ing optical fiber properties, optical IP Routing Design of the student’s interests under the transmitters and receivers, Raman Applies IP routing theory directly guidance of a faculty member. and Erbium optical amplifiers, mod- to design-based networking prob- ulation, optical link design, and lems and real-world examples incor- Thesis wavelength division multiplexing porating physical design, logical A master’s thesis is a substantial optical networks (undersea, long design, financial analysis, and labo- body of original scholarly work by haul, metro, local loop, and LAN). ratory implementation. an individual student.

Call for Nominations for 2005–2006 Awards

For more than 40 years, NAE has involved in determining U.S. Bernard M. Gordon Prize for Inno- recognized outstanding engineers science and technology policy, pro- vation in Engineering and Technol- who have made a lifelong commit- moting U.S. technological develop- ogy Education is a cash prize of ment to advancing the human ment, and improving relations $500,000, split equally between the condition through engineering between industries, government, recipient and his or her institution achievement and/or innovation in and universities. for the continuation of the award- engineering and technology educa- The Charles Stark Draper Prize winning program. The Gordon tion. NAE awards five prizes worth and Bernard M. Gordon Prize are Prize honors technology educators more than $2.5 million (biennially) presented annually, and the Fritz J. whose innovative programs have for achievements in engineering and Dolores H. Russ Prize is pre- strengthened the engineering work- and engineering education—the sented biennially, in February during force by cultivating students’ com- Founders Award, Arthur M. Bueche the National Engineers Week cele- munication skills, creativity, and Award, Charles Stark Draper Prize, bration at the NAE Annual Awards teamwork. Fritz J. and Dolores H. Russ Prize, Dinner and Presentation Ceremony. NAE takes pride in identifying and Bernard M. Gordon Prize. The Charles Stark Draper Prize is a and saluting the best in engineering The Founders Award and Arthur $500,000 cash prize honoring an (see our website, www.nae.edu/ M. Bueche Award are presented engineer whose contributions have awards, for a list of previous award annually in October during the significantly improved the quality of recipients). We invite you to help us NAE Annual Meeting. The life, made it easier for people to live continue the tradition of honoring Founders Award is presented to an freely and comfortably, or provided individuals or teams by nominating NAE member or foreign associate access to information. The Fritz J. outstanding colleagues, professors, who has exemplified the ideals and and Dolores H. Russ Prize is a peers, and industry leaders for the principles of NAE through profes- $500,000 cash prize honoring upcoming prize cycle. Nominations sional, educational, and personal achievements in bioengineering for the 2005 Founders and Bueche achievement and accomplishment. worldwide that have improved the awards and the 2006 Draper and The Bueche Award is awarded to an human condition or had a signifi- Gordon prizes will be accepted Janu- engineer who has been actively cant impact on society. The ary 4, 2005, through April 8, 2005. The 62 BRIDGE

Members and foreign associates will from Leila Rao at (202) 334-1237 or to: NAE Awards, National Acad- receive nomination materials by mail. [email protected] or download them emy of Engineering, 500 Fifth Street, Nonmembers who wish to submit from our website, www.nae.edu/ N.W., NAS 308, Washington, DC nominations can obtain materials awards. Mail or fax your nomination 20001. Fax: (202) 334-1595.

Report of the Home Secretary

in NRC and NAE study task • After careful consideration of the forces. Efforts will continue to effect of the number of references identify members with expertise for nominees (currently a mini- and interests in particular studies. mum of three and a maximum of four) on the probability of election, • The approximate distribution of the impact on the peer committee members among business, acade- of fewer than four references, and mia, and other sectors is 5/4/1, the opportunity, already in place, which is considered appropriate. for one reference per nomination •Available data indicate that to be based on study of a nominee’s efforts to encourage the nomina- accomplishments rather than on W. Dale Compton tion of underrepresented minori- personal acquaintance with the ties and women have been nominee’s work, the MPC unan- The proposed changes in the effective. Rapid changes in the imously concluded that the mini- NAE Articles of Organization were demographics of the profession are mum number of references should approved by a vote of 798 to 40, certain to continue, however, and not be changed. with 1 abstention. The proposed will require continued efforts to changes in the NAE Bylaws were The NAE Council has asked the encourage nominations of minori- approved by a vote of 740 to 11, current MPC to examine the fol- ties and women, as well as mem- with 1 abstention. lowing issues: bers of other ethnic categories. The Membership Policy Commit- • The ultimate size of the Academy tee (MPC) has concluded a two-year • Replacement nominations should and actions that should be taken assessment of membership proce- be in the hands of a “nominating to achieve it. dures and presented its conclusions group” made up of the section to the NAE Council on May 13, chair, the section vice chair, and • Actions that can be taken to 2004. The recommendations that the peer committee chair from the increase the number of nomina- were accepted are summarized below: previous election cycle. Replace- tions of individuals from the busi- ment nominations will be vetted ness sector. • Election to NAE membership by the home secretary in consulta- should continue to be based on • Review of the experiences of the tion with the president prior to accomplishment, rather than other ad hoc Peer Committee 13 during submission to the NAE Council. criteria, such as future promise. the 2002 and 2003 elections to NAE should continue to develop • Data on the sections show that determine if the process, or a mod- its young engineer program and small sections are indeed growing, ification of it, would be useful for similar programs to provide visibil- although perhaps not as fast as identifying emerging technologies. ity for new, promising technologies some would prefer. Experience and technologists. has shown that management’s encouragement of nominations of • NAE should continue to encour- strong candidates is critical to sec- W. Dale Compton age NAE members to participate tion growth. Home Secretary WINTER 2004 63

New Program Officer for the Committee on Engineering Education

Universities as a corporate and foun- (now Philadelphia University) and dation relations consultant with the his master’s degree from the Insti- National Science Foundation. In tute of Textile Technology. that position, he was responsible for CEE is engaged in numerous the operational management of an activities, including the multiphase alliance of public and private funders Engineer of 2020 Project. The of precollege and undergraduate Phase I report, which focused on education with the goal of identify- defining the characteristics of ing and disseminating effective fund- desired student outcomes, has been ing strategies. immensely popular and is now in its Mr. Taber’s prior positions second printing. This past summer Richard Taber include senior research engineer a Phase II workshop was held on with the Institute of Textile Tech- defining implementation strategies Richard Taber joins NAE as the nology and process engineer and for achieving the desired student new program officer for the Com- production manager with Milliken outcomes identified in the Phase I mittee on Engineering Education & Company. He received his bach- report. The report of the Phase II (CEE). Mr. Taber joins us after three elor’s degree from the Philadelphia workshop is scheduled for publica- years with Oak Ridge Associated College of Textiles and Science tion in early 2005.

Education Research Scholar Joins CASEE

Dr. Barbara Lovitts has joined the Sciences at the University of Mary- Upon joining NAE, Dr. Lovitts Center for the Advancement of land, College Park, where she was a stepped down as a member of the Scholarship on Engineering Educa- research associate in charge of activ- National Academy of Sciences tion (CASEE) as a senior program ities of two Alfred P. Sloan Founda- Committee on Assessment for NIH officer. She is leading a project to tion-funded studies, including Minority Research Training Pro- develop a prototype database of edu- Making the Implicit Explicit: A Blue- grams. Dr. Lovitts is the author of cational practices in engineering and print for Creating Performance Expec- the widely cited Leaving the Ivory science education. Dr. Lovitts is a tations and Assessing the Outcomes of Tower: The Causes and Consequences distinguished researcher, author, and Doctoral Education (Stylus, 2006). of Departure from Doctoral Study invited speaker with expertise in Dr. Lovitts’ experience includes (Rowman and Littlefield, 2001). quantitative and qualitative educa- senior research analyst with the She received her doctoral degree in tion research, project and program American Institutes of Research and sociology from the University of evaluation, and grant program man- program director in the Research on Maryland, College Park. agement. She joins NAE from the Learning and Teaching Program of College of Behavioral and Social the National Science Foundation. The 64 BRIDGE

CASEE Fall Student Intern

Sarah Hunt, a Ph.D. candidate research but might benefit from the in ecological and environmental findings of such research. anthropology at the University of Ms. Hunt earned her bachelor’s Georgia, is interning this fall with degree in biology from Earlham Col- the Center for the Advancement of lege, where she was admitted to Phi Scholarship on Engineering Educa- Beta Kappa. She hopes her intern- tion (CASEE). Ms. Hunt’s primary ship with the National Academies project for the fall term is identify- will provide experience to advance ing literature on changes in student her long-term career goals, which learning outcomes as a result of include becoming a faculty member academic or co-curricular inter- and researcher on educational activ- Sarah Hunt ventions. This literature will then ities that span anthropology, engi- be incorporated into a prototype neering, and environmental studies, database for use by engineering with a focus on the nexus of tech- faculty who may have little experi- nology, innovation, and policy. ence with engineering education

Janice Tsai, NAE Intern

Rutgers, Ms. Tsai studied the practice EngineerGirl! project by developing of politics, agenda-setting, lobbying, funding proposals to support the and communications in govern- addition of interactive web content. ment. She completed her B.A. in As a Christine Mirzayan Gradu- mathematical methods in the social ate fellow, Ms. Tsai says her experi- sciences and political science at ence at NAE “was invaluable in Northwestern University. Prior to exposing me to the intersections of her return to graduate school, she engineering and policy.” Her career spent two years in Chicago working goals are to help develop and guide at IBM designing network architec- technology policy through research, Janice Tsai tures for web hosting systems. quantitative analysis, and experi- During her three-month intern- ence, especially in the areas of intel- ship at NAE, Ms. Tsai worked with lectual freedom and copyrights, Janice Tsai will earn a master’s the Committee on Diversity in the information security, and collabora- degree in library and information Engineering Workforce on the tive research. She will pursue a mas- science at Rutgers University next Diversity Innovation Initiative to ter’s or Ph.D. in public policy spring. As an Eagleton fellow at find ways to increase diversity in beginning next fall. the Eagleton Institute of Politics at engineering. She also assisted the WINTER 2004 65

Tenth U.S. Frontiers of Engineering Symposium

This year’s dinner speaker was begun planning for the next U.S. Alexander Singer, a film director FOE meeting, to be held September who has served on two National 22–24, 2005, at GE Global Research Research Council committees, one Labs in Nisakayuna, New York. on the convergence of telecommuni- Funding for the 2004 symposium cations, computers, and entertain- was provided by the Air Force Office ment and one on opportunities for of Scientific Research, DARPA, collaboration between the defense DOD (DDR&E-Research), NASA, and entertainment research commu- Eastman Kodak, Microsoft Corpora- nities. His talk, “Unlikely Partners: tion, Cummins Inc., ATOFINA DARPA and Me,” touched on Chemicals Inc., Air Products and numerous issues at the intersection Chemicals Inc., and Dr. Ruth M. of entertainment and engineering. Davis and other individual donors. He reminded Frontiers participants NAE has been hosting an annual that their work is as creative as the U.S. FOE symposium since 1995. work of artists. FOE also hosts bilateral programs Laura Ray (Dartmouth) demonstrates the design Pablo P. Debenedetti, Class of with Germany and Japan. FOE rationale for scalable mobile robots for use on the 1950 Professor, Department of meetings bring together outstanding Antarctic plateau. Chemical Engineering, Princeton engineers from industry, academia, University, chaired the organizing and government at a relatively early NAE hosted the tenth U.S. Fron- committee and the symposium. The point in their careers (all partici- tiers of Engineering (FOE) Sympo- 2005 organizing committee, also pants are 30 to 45 years old) and pro- sium at the Beckman Center in chaired by Dr. Debenedetti, has vide an opportunity for them to Irvine, California, September 9–11. learn about developments, tech- This milestone meeting was at- niques, and approaches at the fore- tended by approximately 110 engi- front of fields other than their own. neers, and talks covered topics in As engineering becomes increasing- engineering for extreme environ- ly interdisciplinary, interactions with ments, designer materials, multi- people in other disciplines are criti- scale modeling, and engineering and cal to researchers on the cutting edge entertainment. In February 2005, of new concepts and technologies. NAE will publish a symposium vol- The meeting also facilitates the ume containing extended sum- establishment of contacts and col- maries of the presentations. laborations among the next genera- On the second afternoon of the tion of engineering leaders. symposium, representatives of the For more information about the Defense Advanced Research Projects symposium series or to nominate an Agency (DARPA), the U.S. Depart- outstanding engineer to participate ment of Defense (DOD) Office of in future FOE symposia, contact Basic Science, and National Aero- Janet Hunziker at the NAE Program nautics and Space Administration Office at (202) 334-1571 or by (NASA) gave presentations to the Greg Carman (UCLA) describes thin-film active mate- e-mail at [email protected]. group about their missions and work. rials at the U.S. Frontiers of Engineering Symposium. The 66 BRIDGE

Calendar of Meetings and Events

2004 February 4 NAE Membership Policy February 24 NAE Finance and Budget December 1–2 Invention Session, Committee on Committee Meeting Committee Conference Call Diversity in the Engineering Irvine, California March 16 NRC Executive Committee Workforce February 7–8 NRC Governing Board Meeting Meeting December 4 NAE Committee on Membership Irvine, California NRC Governing Board Meeting Meeting February 9–10 NAE Council Meeting March 22 NAE Regional Meeting December 9 News and Terrorism: Irvine, California University of Southern California, Communicating in a Crisis February 10 NAE National Meeting Los Angeles Workshop Irvine, California April 21 NAE Regional Meeting Miami, Florida February 16 NRC Executive Committee University of Minnesota, December 16 Committee on Engineering Meeting Minneapolis Education Meeting NRC Governing Board Meeting All meetings are held in the Academies Building, 2005 February 20–26 National Engineers Week Washington, D.C., unless otherwise noted. For January 12 NRC Executive Committee February 21 NAE Awards Dinner Ceremony information about regional meetings, please Meeting and Presentation contact Sonja Atkinson at [email protected] or NRC Governing Board Meeting (202) 334-3677.

In Memoriam

JOHN H. ARGYRIS, 87, professor died on August 29, 2001. Dr. Chest- elected to NAE in 1970 for contri- emeritus, University of Stuttgart, nut was elected to NAE in 1974 for butions to the design and engin- died on April 2, 2004. Dr. Argyris contributions to the theory and eering of the Mercury and Apollo was elected a foreign associate of practice of control systems and for spacecraft. NAE in 1986 for pioneering and systems engineering. continuing contributions in com- WILLIS M. HAWKINS, 90, con- puter mechanics over a period of FLOYD L. CULLER, 81, president sultant, Lockheed Martin Corpora- more than 30 years. emeritus, Electric Power Research tion, died on September 28, 2004. Institute Inc., died on September 29, Dr. Hawkins was elected to NAE in RICHARD M. CARLSON, 79, 2004. Mr. Culler was elected to 1966 for the design and develop- retired chief, Advanced Systems NAE in 1974 for contributions to ment of aircraft, missile, and space Research and Analysis Office, U.S. the development of nuclear power. systems. Army Aviation and Troop Com- mand, and consultant, died on July BOB O. EVANS, 77, partner, BERTRAM WOLFE, 77, retired 12, 2004. Dr. Carlson was elected to Rocket Ventures, died on Septem- vice president and general manager, NAE in 1990 for his contributions ber 2, 2004. Mr. Evans was elected GE Nuclear Energy, and consultant, to the application of composite to NAE in 1970 for contributions to died on September 6, 2004. Dr. materials to operational helicopters. the development of nuclear power. Wolfe was elected to NAE in 1980 for contributions to the develop- HAROLD CHESTNUT, 83, MAXIME A. FAGET, 83, retired ment of advanced nuclear concepts retired consultant, GE Corporate director, Space Industries Inc., died and projects, particularly fast flux Research and Development Center, on October 9, 2004. Dr. Faget was reactors. WINTER 2004 67

Publications of Interest

The following reports have been management; risk assessment; risk work toward reconciling inconsis- published recently by the National management; and linking risk tencies in national water-related Academy of Engineering or the assessment and risk management. policies and create a body to coordi- National Research Council. Unless Paper, $35.00. nate water policies among federal otherwise noted, all publications are agencies with water-management for sale (prepaid) from the National Air Quality Management in the United responsibilities. The committee also Academies Press (NAP), 500 Fifth States. According to this report, recommends that the federal “Prin- Street, N.W., Lockbox 285, Wash- the implementation of air quality ciples and Guidelines” document, ington, DC 20055. For more infor- regulations should be less bureau- which was last revised in 1983, be mation or to place an order, contact cratic—with more emphasis on revised and updated. Paper, $28.50. NAP online at results and less on process—and or by phone at (888) 624-8373. should protect ecosystems as well Army Science and Technology for (Note: Prices quoted by NAP are sub- as people. The report recommends Homeland Security: Report 2—C4ISR. ject to change without notice. Online that the Environmental Protection Shortly after the events of Septem- orders receive a 20 percent discount. Agency target groups of pollutants ber 11, 2001, the U.S. Army asked Please add $4.50 for shipping and han- rather than individual pollutants the National Research Council dling for the first book and $0.95 for and that revised or new regulations to conduct a series of studies on each additional book. Add applicable address air pollution that travels how the Army could use science sales tax or GST if you live in CA, across state and international and technology to meet its home- DC, FL, MD, MO, TX, or Canada.) borders. This will require better land defense obligations. The first tracking of emissions to assess report, Science and Technology for Accident Precursor Analysis and Man- which populations are at the high- Army Homeland Security—Report 1, agement: Reducing Technological Risk est risk of health problems from included a survey of a broad range through Diligence. In the aftermath of pollution and to measure the effi- of technologies and recommended catastrophes, it is common to find cacy of pollution-control strategies. applying Future Force technologies prior indicators, missed signals, and Hardcover, $49.95. to homeland security wherever dismissed alerts, that is, precursors possible. The committee concluded that might have signaled impending Analytical Methods and Approaches for that the Army should play a major danger and even averted the cata- Water Resources Project Planning. role in providing emergency com- strophes. This report of an NAE This review of analytical and plan- mand, control, communications, workshop in July 2003 documents ning methods used by the U.S. computers, intelligence, surveil- industrial and academic approaches Army Corps of Engineers (USACE) lance, and reconnaissance (C4ISR) to detecting, analyzing, and benefit- focuses on the steps outlined in the capabilities and that the tech- ing from accident precursors and federal “Principles and Guidelines” nology and architecture for home- examines public-sector and private- document and in the USACE land security C4ISR was com- sector roles in the collection and “Planning Guidance Notebook.” patible with the technology of the use of precursor information. The The study concludes that clear, Army’s Future Force. This second report includes an analysis, findings, comprehensive planning studies are report focuses on how C4ISR can and recommendations by the often hindered by inconsistent leg- facilitate the Army’s efforts to assist authoring committee and 11 indi- islation and limited interagency col- the U.S. Department of Homeland vidually authored background laboration. The report recommends Security and emergency responders papers on four topics: the oppor- that the administration and Con- during or after catastrophic events. tunity of precursor analysis and gress, in cooperation with the states, Paper, $36.50. The 68 BRIDGE

California Agricultural Research Priori- years to come. This report reviews than provide a comprehensive list ties: Pierce’s Disease. The glassy- the U.S. Department of Energy of research topics or a taxonomy of winged sharpshooter is one of the (DOE) Methane Hydrate Research research areas, the committee recent invasive pests afflicting Cali- and Development Program, the pro- describes key ideas at the core of fornia agriculture. The sharpshooter ject selection process, and projects CS, but does not define boundaries. transmits a bacterial pathogen that funded to date. Key recommenda- This volume also includes some two causes Pierce’s disease, which impairs tions include focusing research on dozen essays on particular aspects of production of wine, table, and raisin seven high-priority areas; ensuring CS research and their results. grapes in California. The report rec- scientific oversight in the selection, Paper, $35.00. ommends improving the process and initiation, monitoring, and assess- setting priorities for research on ment of major projects funded by Emerging Technologies and Ethical Pierce’s disease funded by state agen- DOE; funding of fellowships by DOE Issues in Engineering: Papers from a cies and wine industry groups. to increase its contribution to edu- Workshop, October 14–15, 2003. A Research should be focused on iden- cation and training; and providing group of distinguished engineers tifying feasible options for control- project applicants with instructions and ethicists gathered for an NAE ling the spread of the disease and and guidelines that outline the workshop to discuss the respon- designing sustainable approaches requirements for timely, full disclo- sible development and uses of that are adaptable and affordable sure of project results and spell out emerging technologies. Presenta- over the long term. Areas of research the consequences of noncompliance. tions were focused on four areas of that should be pursued more intense- Paper, $43.00. engineering—sustainability, nano- ly include: the genetic makeup of technology, neurotechnology, and the pathogen that triggers Pierce’s Climate Data Records from Environmen- energy—and the ethical issues disease; mechanisms for making tal Satellites: Interim Report. This engineers and society as whole are grapes resistant to the disease; the report identifies the necessary ele- likely to face. Several approaches introduction of predators to the ments in a program to create high- to tackling ethical issues were dis- sharpshooter; and the management quality climate data from satellites. cussed, including: analyzing the of crop planting to avoid spread of The study includes a review of past factual, conceptual, application, the disease. Paper, $35.00. attempts to create climate data and moral aspects of an issue; assess- records; suggests steps for generating the risks and responsibilities of Charting the Future of Methane Hydrate ing, re-analyzing, and storing satel- a particular course of action; and Research in the United States. lite climate data; and points out the considering theories of ethics or Methane hydrate is a natural form of importance of partnering among codes of ethics developed by engi- clathrate—a chemical substance in agencies, academia, and industry. neering societies. which one molecule forms a lattice The National Oceanic and Atmos- Ethics can be integrated into around a “guest” molecule with pheric Administration will use this engineering education for students chemical bonding. In methane report—the first in a two-part and professionals, either as an hydrate, the guest molecule is study—to draft an implementation aspect of courses in the engineering methane and the lattice consists of plan for creating climate data curriculum or as components of water that forms an ice-like solid. records. Paper, $27.25. engineering projects to be exam- Methane hydrate has become the ined along with research findings. focus of international attention Computer Science: Reflections on the Engineering practice workshops can because of its vast potential for Field, Reflections from the Field. This also be effective for confronting human use worldwide. If methane report describes the intellectual ethical issues, particularly when the can be produced from hydrate, a rea- character of computer science (CS) participants include experienced sonable assumption given that there in accessible form and through engineers and young engineers, or are no obvious technical or engi- examples. The object of the report even engineering students. Because neering roadblocks to commercial is to prepare readers for what the of the enormous impact of engi- production, the nation’s natural gas future might hold and to inspire CS neering on the environment, health supply could be extended for many researchers to help create it. Rather care, and quality of life, the explicit WINTER 2004 69

consideration of ethics should be an of scientific knowledge, including water temperatures are highest. integral part of engineering educa- measures of uncertainty, into effec- These periods of low flows, high tion, as well as engineering practice. tive decision support systems and demand, and high temperatures are Paper, $36.00. resources.” Paper, $30.00. critical for juvenile salmon migrating downstream through the Columbia The Hydrogen Economy: Opportunities, Indicators for Waterborne Pathogens. River hydropower system. Although Costs, Barriers, and R&D Needs. The With recent and forecasted advances the impact on salmon of individual announcement of a hydrogen fuel in microbiology, molecular biology, water withdrawals may be small, the initiative in the president’s 2003 and analytical chemistry, a reassess- cumulative effects of numerous with- State of the Union speech stimu- ment is in order of the current prac- drawals affect Columbia River flows lated interest in the potential of tice of relying predominantly or and thus increase risks to salmon. hydrogen in the nation’s long-term exclusively on traditional bacterial The report concludes that, although energy future. Prior to that time, the indicators to detect and identify all our knowledge of the migratory U.S. Department of Energy (DOE) types of waterborne pathogens. This behavior and physiology of salmon is had asked the National Research report concludes that indicator imperfect, decision makers should Council to conduct a study of the approaches will be required for the acknowledge this and be willing to technical issues involved in the foreseeable future because monitor- take action in the face of uncertain- hydrogen economy to assist in the ing for the complete spectrum of ties. The report recommends that development of its hydrogen R&D microorganisms in water is not prac- the state of Washington, Canada, program. This assessment includes tical or even feasible and because other basin states, and tribal groups the current status of technology; cost many known pathogens are difficult establish a basin-wide forum to estimates; the status of CO2 emis- to detect directly and reliably in consider future water withdrawal sions; distribution, storage, and end water samples. This comprehensive application permits. If the state of use; and the DOE RD&D program. report recommends the develop- Washington issues additional permits The report assesses hydrogen as a ment and use of a “tool box” for water withdrawals, those permits fuel for the future and identifies approach by the Environmental should include provisions for curtail- challenges that must be overcome Protection Agency and others for ing withdrawals in critical high- to realize that future. Topics covered assessing microbial water quality; demand periods. Hardcover, $35.00. include hydrogen end-use tech- this would mean available indicator nologies, transportation, hydrogen- organisms (and/or pathogens in some Meeting the Energy Needs of Future production technologies, and trans- cases) and detection method(s) Warriors. The central feature of itional issues for hydrogen-fueled could be matched to the require- combat soldiers’ equipment is vehicles. Paper, $32.00. ments of a particular situation. The increasingly sophisticated sensing, report also recommends that a communications, and related elec- Implementing Climate and Global Change phased, three-level monitoring tronics for use on the battlefield. Research: A Review of the Final U.S. Cli- framework be developed to support The most critical factor for their use mate Change Science Program Strategic the selection of indicators and indi- is the development of power supply Plan. The report reviews a draft cator approaches. Paper, $54.00. systems capable of operating those strategic plan from the U.S. Climate electronics effectively for up to Change Science Program, which was Managing the Columbia River: Instream 72 hours. The Army requested that established in 2002 to coordinate Flows, Water Withdrawals, and Salmon the National Research Council U.S. research on climate change and Survival. Flows of the Columbia review the state of the art in power global change. The U.S. Climate River, although modified substan- systems and recommend techno- Change Science Program combines tially during the twentieth century, logies that could support the rapid the decade-old Global Change still vary considerably from season development of power systems for Research Program and the new Cli- to season and year to year. The future soldier systems. This report mate Change Research Initiative. smallest flows tend to occur during includes assessments of techno- The primary goal of the program is to the summer when the demand for logical options for meeting different “measurably improve the integration irrigation water is highest and when power-level requirements, power- The 70 BRIDGE system designs, and soldier-energy about the quality of the patents a science-oriented, apolitical forum sinks. Future design concepts, par- issued (whether they meet statutory for leaders in the chemical sciences ticularly low-power systems, are standards of novelty, utility, non- to discuss chemically related issues also described. The recommenda- obviousness, and adequate written affecting government, industry, and tions are focused on technology description), the resources available universities. Paper, $18.00. development and system design. to the U.S. Patent and Trademark Paper, $23.00. Office to keep pace with changes Retooling Manufacturing: Bridging and the volume of applications, fea- Design, Materials, and Production. As Naval Forces’ Defense Capabilities tures of U.S. law that limit the dis- the U.S. Department of Defense against Chemical and Biological Warfare semination of information in patents (DOD) continues to pursue the Threats. U.S. naval forces must be and that raise the cost and uncer- development of the future warrior prepared to respond to a broad array tainty of litigation over patent valid- system, the difficulty of moving of threats, including chemical and ity and infringement, access to complex technologies rapidly from biological warfare (CW and BW) patented research technologies for design to manufacturing is becom- threats. To help review its current basic noncommercial research, and ing a major concern. Communica- state of preparedness, the Chief of redundancies and inconsistencies tion gaps between designers and Naval Operations asked the Na- among national patent systems that manufacturers have hindered the tional Research Council to assess the raise the cost of global intellectual rapid development of new products U.S. Navy’s defense capabilities property protection. Paper, $35.00. for future military operations. DOD against CW and BW threats and the asked the National Research Coun- development of methods of sensing Preparing Chemists and Chemical Engi- cil to develop a framework for and analyzing chemical and biolog- neers for a Globally Oriented Work- “bridging” these gaps through data ical agents, withstanding or avoiding force: A Workshop Report to the management, modeling, and simula- exposure to such agents, and dealing Chemical Sciences Roundtable. Global- tion. This report provides a frame- with contamination under a broad ization—the flow of people, goods, work for virtual design and range of operational conditions. services, capital, and technology manufacturing and an assessment of This report provides an overview of across international borders—signif- necessary tools; an analysis of the potential threats and an evaluation icantly affects the chemistry and economic dimensions of the prob- of the Navy’s operations, nonmedical chemical engineering professions. lem; a discussion of barriers to virtu- programs, and medical countermea- Chemical companies are looking for al design and manufacturing in the sures designed to confront CW and new ideas, a trained workforce, and DOD acquisition process; recom- BW threats. The report also includes new market opportunities, regardless mendations for addressing these general and specific findings and rec- of geographic location. During an issues; and a description of research ommendations based on these assess- October 2003 workshop, leaders in needs. Paper, $28.50. ments. Paper, $28.00. chemistry and chemical engineering from industry, academia, govern- River Basins and Coastal Systems Plan- A Patent System for the 21st Century. ment, and private funding organiza- ning within the U.S. Army Corps of Engi- This report focuses on how well the tions came together to explore the neers. The emphasis of U.S. Army patent system encourages research, implications of the global research Corps of Engineers (USACE) water innovation, and the dissemination environment for the chemistry and resources projects is changing. Tra- of knowledge and how well it is chemical engineering workforce. ditional civil works construction is being adapted to rapid technological The workshop presentations identi- giving way to maintenance, opera- and economic changes. The panel fied deficiencies in the current U.S. tional, habitat restoration and reha- concludes that the system has shown educational system and pointed out bilitation, and recreational projects. admirable flexibility in accommo- the need for the United States to Planning water projects requires an dating new technologies and reflect- create and sustain a globally aware integrated systems approach cap- ing the importance of intangible workforce in the near future. The able of balancing all relevant issues capital of all sorts. Nevertheless, goal of the workshop was to inform and identifying unintended con- there is reason to be concerned the Chemical Sciences Roundtable, sequences or cumulative effects. WINTER 2004 71

Congress requested that the Na- USACE’s project process and guide- individuals and scientific communi- tional Research Council review lines (e.g., mandatory federal cost ties. The committee recommends USACE’s planning, design, opera- sharing or required assessments for that the U.S. scientific community tion, and evaluation processes in the stating how new projects will fulfill and other agencies use the IPY to context of the nation’s river basins USACE’s commitment to environ- initiate environmental change and and coastal systems. This report mental stewardship). Paper, $39.00. variability in polar regions; explore provides detailed recommendations scientific frontiers, ranging from the for improvements in planning and A Vision for the International Polar Year molecular to the planetary scale; management practices for river 2007–2008. In recognition of the and engage the public (and ulti- basins and coastal systems, includ- International Polar Year (IPY), the mately advance general scientific ing, but not limited to: modifica- United States plans to present an literacy) by providing information tion and updating of planning overview of potential science about polar regions in the global sys- guidance documents and the incor- themes, enabling new technologies, tem. Paper, $28.50. poration of new requirements into and public outreach programs for

The Periodicals BRIDGE Postage Paid (USPS 551-240)

NATIONAL ACADEMY OF ENGINEERING 2101 Constitution Avenue, NW Washington, DC 20418