Winter 2005

The BRIDGE LINKING ENGINEERING AND SOCIETY

Large-Scale Activity-Recognition Systems Matthai Philipose Science and Engineering Research That Values the Planet Arne Jacobson and Daniel M. Kammen The Promise of Synthetic Biology Jay Keasling Agent-Based Modeling as a Decision-Making Tool Zoltán Toroczkai and Stephen Eubank Fuel Cells: Current Status and Future Challenges Stuart B. Adler Organic Semiconductors for Low-Cost Solar Cells Chiatzun Goh and Michael D. McGehee

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. 35, No. 4, Winter 2005 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. © 2005 by the National Academy of Sciences. All rights reserved.

Erratum In the table on p. 8 in the September issue, the names of the cities for Microsoft and Oracle should have been Beijing, Bangalore, and Hyderabad.

A complete copy 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 35, Number 4 • Winter 2005 BRIDGE LINKING ENGINEERING AND SOCIETY

Editor’s Note 3 Cutting-Edge Research in Engineering Pablo G. Debenedetti

Features 5 Large-Scale Activity-Recognition Systems Matthai Philipose An activity-recognition system could improve the lives of the elderly and infirm. 11 Science and Engineering Research That Values the Planet Arne Jacobson and Daniel M. Kammen Ecological stewardship will be the guiding scientific principle for new avenues of inquiry. 18 The Promise of Synthetic Biology Jay Keasling Synthetic biologists may soon design and build engineered biological systems. 22 Agent-Based Modeling as a Decision-Making Tool Zoltán Toroczkai and Stephen Eubank Control over agent-based systems can be achieved via modeling tools. 28 Fuel Cells: Current Status and Future Challenges Stuart B. Adler Fuel cells of the future will be based on solid electrolytes. 33 Organic Semiconductors for Low-Cost Solar Cells Chiatzun Goh and Michael D. McGehee The world will need access to 30TW of power without releasing carbon into the atmosphere.

NAE News and Notes 40 NAE Newsmakers 40 2005 Annual Meeting 42 Chairman’s Remarks 44 President’s Address 48 2005 NAE Founders Award Acceptance Remarks 50 2005 Bueche Award Acceptance Remarks

(continued on next page) The BRIDGE

52 Gordon Prize Lecture 56 Christine Mirzayan Science and Technology Policy Graduate Fellows 57 ExxonMobil Scholar-in-Residence 57 CASEE Scholar-in-Residence 58 U.S. Frontiers of Engineering Holds 2005 Meeting at GE Global Research Center 60 NAE Hosts Japan-America Frontiers of Engineering Symposium 61 Leading Philanthropists Inducted into Einstein Society 62 Calendar of Meetings and Events 63 In Memoriam

64 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. Ralph J. Cicerone 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. Ralph J. Cicerone and Dr. Wm. A. Wulf are chair and vice chair, respec- tively, of the National Research Council. www.national-academies.org WINTER 2005 3 Editor’s Note

Cutting-Edge Research Garrick Louis and Amy Smith organized the session in Engineering on engineering for developing communities, which The U.S. Frontiers of Engi- opened with a presentation by Kurt Kornbluth. He neering (FOE) Symposium, described the DISASCARE Wheelchair Center in a yearly event sponsored by Zambia, a project that makes use of locally available NAE, brings together some resources to address technological needs. Daniele Lan- 100 outstanding young engi- tagne highlighted engineering inputs to the Centers for neers (ages 30 to 45) from Disease Control and Prevention Safe Water System academia, industry, and gov- (SWS) Program. This initiative provides safe drinking ernment laboratories for water to people with no access to infrastructure-treated Pablo G. Debenedetti is the Class three days of sharing ideas water through point-of-use water chlorination, water of 1950 Professor of Engineering and learning about cutting- storage in safe containers, and education to improve and Applied Science at Princeton edge research on a broad hygiene and water practices. Julie Beth Zimmerman University and an NAE member. range of engineering topics. explained how engineers could use the principles of The competitively selected green engineering as a design protocol for promoting emerging engineering leaders who attend FOE symposia sustainability. She argued that a design framework that represent a wide spectrum of backgrounds, interests, and incorporates sustainability factors as performance crite- talents, and the event offers them a unique opportunity ria could advance the goals of prosperity and a health- to learn about the latest research in engineering areas ier environment. Daniel Kammen, who gave the last other than their own. Six papers based on this year’s presentation in the session, focused on sustainability presentations are included in this issue of The Bridge. science. In his paper (p. 11), he argues for the devel- The eleventh FOE Symposium, held on Septem- opment of a science and engineering research agenda in ber 22–24, 2005, at the General Electric Global which the preservation of natural and social systems Research Center in Niskayuna, New York, encom- plays a central role. passed four themes: ID and verification technologies, Luis Amaral and Kelvin Lee organized the session on engineering for developing communities, engineering the engineering of complex systems. In a brief opening complex systems, and energy resources for the future. talk, Luis Amaral identified the distinguishing character- The session on ID and verification technologies was istic of complex systems as the emergence of behavior chaired by Visvanathan Ramesh. The first speaker, not foreseeable by the system designer. In Alessandro Peter Belhumeur, addressed the challenge of face recog- Vespignani’s presentation, he showed how the computa- nition as a computational pattern-recognition or tional study of complex networks, such as the World Wide machine-learning problem. He described the chal- Web, can provide valuable insights into infrastructure lenges that lie ahead, such as handling voluntary design, epidemiology, and social science. Jay Keasling, changes in facial expression and natural outdoor whose paper is printed in this issue (p. 18), described syn- lighting. Jonathon Phillips focused on the design of thetic biology, that is, the design and construction of new objective biometrics for assessing face- and fingerprint- biological entities, such as cells, enzymes, and genetic cir- recognition problems and independent evaluations. cuits. The exciting possibilities in this emerging field Matthai Philipose, whose paper appears in this issue include engineered bacteria for the production of anti- (p. 5), described the development of computing systems malaria drugs. The final presentation by Zoltán Toroczkai that can observe, understand, and act on physical described modeling of agent-based systems, collectives of human activity. As an example of the potential bene- living entities, as distinguished from collectives of inan- fits of this technology, he described how it might be imate constituents, such as multiparticle systems. He used in caring for the elderly. Rapid progress is being illustrated how agent-based modeling can be used for made in this field based on a family of sensors based on planning urban transportation and for analyzing disease- radio frequency identification. spread scenarios. His paper appears on p. 22. The 4 BRIDGE

The subject of the fourth session, chaired by John interests and activities. The get-acquainted sessions Vohs, was energy resources for the future. John Reinker turned out to be highly interactive and very educational. discussed the current and future electrical energy picture The dinner speaker, a traditional highlight of FOE in the United States. He also addressed the challenges programs, was Shirley Jackson, president of Rensselaer of satisfying world energy demand and the roles of Polytechnic Institute. In the course of her distinguished fossil fuels, nuclear, wind, solar, hydroelectric, and bio- career, Dr. Jackson has held leadership positions in gov- mass technologies in meeting those challenges. Sunita ernment, industrial research, and academia. She is an Satyapal provided an overview of research and develop- NAE member and a fellow of the American Physical ment activities in the U.S. Department of Energy Hydro- Society and American Academy of Arts and Sciences. gen Program. She described how improved theoretical Her truly inspirational speech, entitled “Engineering for modeling, high-throughput screening techniques, and a New World,” addressed the need for a larger, more understanding at the nanoscale have impacted the dis- diverse workforce in science and engineering, the chal- covery and optimization of materials to meet the targets lenges and opportunities involved in meeting world for commercially viable vehicular hydrogen storage sys- energy demand, and the emergence of new economic tems. Stuart Adler described the current status and powers, such as China and India. future challenges of fuel cell technologies (p. 28). He For the past three years, it has been my privilege to compared polymer-electrolyte and solid-oxide fuel cells chair the FOE organizing committee, which selects the and described the significant technical challenges that speakers and topics for the symposium. I thank all the must be overcome for these technologies to become session chairs, speakers, and participants of the 2003, commercially viable. Michael McGehee addressed the 2004, and 2005 meetings for maintaining the high qual- need for alternative technologies to meet worldwide ity that has become the standard of FOE symposia. I am energy demand while minimizing adverse environmental also grateful to Lance Davis, NAE Executive Officer, and consequences. He described organic semiconductors, Janet Hunziker, NAE Program Officer, for their invalu- which can be dissolved in common solvents and sprayed able contributions to the planning, organization, and or painted onto substrates, as promising candidates for successful implementation of these unique meetings. I the development of solar cells that would not only wish the incoming chair of the organizing committee, be environmentally benign, but would also be economi- Dr. Julia Phillips, Director of the Physical, Chemical and cally competitive (p. 33). Nano Sciences Center at Sandia National Laboratories, The technical talks were followed by extended, lively every success. Q&A sessions with enthusiastic participation by the I know of no meeting as interdisciplinary, diverse, and audience. The program this year also featured 90-minute stimulating as FOE, and I hope that the six papers get-acquainted sessions, during which attendees were included in this issue convey some of the excitement we divided into nine groups. Each person had been asked experienced in Niskayuna in September. before the meeting to prepare a transparency describing his or her work and was given two minutes to introduce him/herself and explain his/her technical work. The bal- ance of the time was devoted to discussions on research An activity-recognition system could improve the lives of the elderly and infirm.

Large-Scale Activity- Recognition Systems

Matthai Philipose

Building computing systems that can observe, understand, and act on day- to-day physical human activity has long been a goal of computing research. Such systems could have profound conceptual and practical implications. Because the ability to reason and act based on activity is a central aspect of human intelligence, from a conceptual point of view, such a system could improve computational models of intelligence. More tangibly, machines Matthai Philipose is a researcher at that can reason about human activity could be useful in aspects of life that the Intel Research Laboratory in are currently considered outside the domain of machines. Seattle, where he leads the Sys- Monitoring human activity is a basic aspect of reasoning about activity. In fact, monitoring is something we all do—parents monitor children, adults tem for Human Activity Recognition monitor elderly parents, managers monitor teams, nurses monitor patients, and Prediction (SHARP) Project. and trainers monitor trainees; people following medication regimens, diets, recipes, or directions monitor themselves. Besides being ubiquitous, however, monitoring can also be tedious and expensive. In some situations, such as caregiver-caretaker and manager- worker relationships, only dedicated, trained human monitors can make detailed observations of behavior. However, such extensive observation causes fatigue in observers and resentment in those being observed. The constant involvement of humans also makes monitoring expensive. Tasks that are ubiquitous, tedious, and expensive are usually perfect can- didates for automation. Machines do not mind doing tedious work, and The 6 BRIDGE expensive problems motivate corporations to build The system would not just improve the quality of data machines. In fact, given the demographics of our soci- collected, but (because it provides constant monitoring) ety, systems that notify family members automatically might also be able to provide proactive intervention and when elderly relatives trigger simple alarms, such as other assistance. Figure 1 shows a prototype form of the falling, not turning off the stove, or not turning off hot Caregiver’s Assistant. Actual forms include activities water, are now commercially available. However, com- in 23 categories, such as “housework” and “hygiene,” pared to a live-in family member who can monitor an which instantiate to tens of thousands of activities, such elder’s competence in thousands of day-to-day activities, as “cleaning a bathtub” and “brushing teeth.” these systems barely scratch the surface. In this paper, I Thus, an activity-recognition system that could track describe a concrete application for a monitoring system thousands of activities in non-laboratory conditions with broad activity-recognition capabilities, identify a would remove a substantial burden from human moni- crucial missing ingredient in existing activity “recogniz- tors. Professional caregivers could, at any time, be pro- ers,” and describe how a new class of sensors, combined vided with a version of this form with potentially with emerging work in statistical reasoning, promises to troublesome areas highlighted. If a nurse were given this advance the state of the art by providing this ingredient. form before a visit, for instance, he or she could make better preparations for the visit and could focus on The Caregiver’s Assistant the most important issues during the visit. A study of Caring for the elderly, roughly one hundred professional caregivers around the either as a professional country has shown that such a system would be useful, caregiver or as a family at least for caregivers. member, is a common bur- den in most societies. Discriminating among Activities Gerontologists have devel- The process of recognizing mundane physical activities oped a detailed list of activ- can be understood as mapping from raw data gathered by ities, called the activities of sensors to a label denoting an activity. Figure 2 shows daily living (ADLs), and how traditional mapping systems are structured. Feature metrics for scoring per- selection modules typically work on high-dimensional, formance of crucial day-to- high-frequency data coming directly from sensors (such as day tasks, such as cooking, cameras, microphones, and accelerometers) to identify dressing, toileting, and relatively small numbers of semantically higher level fea- FIGURE 1 Prototype of electronic socializing, which are cen- tures, such as objects in images, phonemes in audio Activities of Daily Living Form with tral to a person’s well-being. streams, and motions in accelerometer data. Symbolic check mark added by the electronic An elder’s ADL score is inference modules reason about the relationship between Caregiver’s Assistant. accepted as an indicator of these features and activities in a variety of ways. The rea- his or her cognitive health. soning may include identifying ongoing activities, Professional caregivers in the United States are often detecting anomalies in the execution of activities, required to fill in ADL forms each time they visit their and performing actions to help achieve the goal of patients. Unfortunately, although the data they collect the activities. are used as a basis for making resourcing decisions, such Both feature selection and inference techniques have as Medicaid payments, the data are often inaccurate been investigated extensively, and depending on the because (1) they are often based on interviews with feature, researchers can draw on large bodies of work. In elders who may have strong motives for misrepresenting the computer vision community alone, extensive work the facts and (2) because the data-collection window is has been done on objects, faces, automobiles, gestures, narrow relative to the period being evaluated. Given and edges and motion flows, each of which has a dedi- increasing constraints on caregivers’ time, purely man- cated sub-community of researchers. Thus, once fea- ual data collection seems unsustainable in the long run. tures for an activity-recognition system have been The Caregiver’s Assistant system is intended to fill selected, a very large number of model representations out large parts of the ADL form automatically based on and inference techniques are available. These tech- data collected from the elder’s home on a 24/7 basis. niques differ in several ways, such as whether they WINTER 2005 7

distinguish between thousands of activities. For Discriminating among Activities instance, objects used during activities have long been thought to be crucial discriminators. However, existing activities object-recognition and tracking systems tend not to work very well when applied to a large variety of objects Symbolic Inference in unstructured environments (Sanders et al., 2002). Modules Activity-recognition systems based on tracking objects, therefore, tend to be customized for particular environ- ments and objects, which limits their utility as general- purpose, day-to-day activity recognizers. Given that high-level features producing each customized detector is a research task, the goal of general-purpose recognition has, not surpris- Feature Selection ingly, not been reached. Modules A new class of small, wireless sensors seems likely to provide a practical means of detecting objects used in many day-to-day activities (Philipose et al., 2004; Tapia raw sensor data et al., 2004). Given a stream of objects, recent work has sensors shown that even simple symbolic inference techniques are sufficient for tracking the progress of these activities. FIGURE 2 A typical activity-recognition system. Detecting Object Use with Radio Frequency support statistical, higher order, or temporal reasoning; Identification Tag Sensors the degree to which they learn and the amount of A passive radio frequency identification (RFID) tag human intervention they require to learn; and the effi- (Figure 3a) is a postage-stamp-sized, wireless, battery-free ciency with which they process various kinds of features, transponder that, when interrogated (via radio) by an especially higher dimensional features. In Figure 2, the ambient reader, returns a unique identifier (Finkenzeller, variety of feature selections and inference algorithms is 2003). Each tag consists of an antenna, some protocol indicated by stacks of boxes. logic, and optional nonvolatile memory. RFID tags use Despite the profusion of options, no activity infer- the energy of the interrogating signal to return a 64-bit encing system capable of recognizing large numbers of to 128-bit identifier unique to each tag, and when applic- day-to-day activities in natural environments has been able, data stored in on-tag memory. Short-range tags, developed. A key underlying problem is that no exist- which are inductively coupled, have a range of 2 to ing combination of sensors and feature selector has been 30 cm; long-range backscatter-based tags have a range of shown to detect robustly the features necessary to 1 to 10 m. Tags are available off the shelf for less than

ab

FIGURE 3 a. RFID tags. b. Tagged toothbrush and toothpaste. The 8 BRIDGE

ab

FIGURE 4 a. Close-up of an iBracelet with a quarter for comparison. b. An iBracelet in use.

50 cents each. Short-range readers cost a few hundred Long-range tags, however, do not require the explicit dollars; long-range readers cost a few thousand dollars. If cooperation of those being monitored. Readers in the current trends continue, there will be a steep drop in the corner of a room can detect tags anywhere in that room. price of both tags and readers in the next few years. Unfortunately, because a conventional RFID tag simply If an RFID tag is attached to an object (Figure 3b) reports the presence of tagged objects in the reader’s field, and the tag is detected in the vicinity of a reader, we can and not their use, long-range tags cannot tell us when infer that the attached object is also present. Given objects are being used either. Long-range tags simply list their object-tracking abilities, RFID-based systems are all tagged objects in the room they are monitoring. currently being seriously considered for commercial However, each of these modalities can be re-engineered applications, such as supply-chain management and to detect object use unobtrusively. Figure 4 shows how asset tracking. Existing uses include livestock tracking, the short-range RFID reader can be adapted to become theft protection in the retail sector, and facilities man- an unobtrusive sensor of object use (Figure 4a). Essen- agement. The promise of a viable RFID system for tially, the RFID reader is a radio with built-in processor, tracking the presence of large numbers of objects sug- nonvolatile memory, and a power supply integrated into gests that it might be the basis of a system for tracking a single bracelet called the iBracelet (Fishkin et al., objects used by people whose activities we wish to mon- forthcoming). The antenna of the RFID reader is built itor. Because a sensor can be attached to each object, we into the rim of the bracelet. When turned on, the have, in principle at least, an “ultra-dense” deployment bracelet scans for tags at 1 Hz at a range of 20–30 cm. of sensors that could allow each tagged object to Any object, such as the water pitcher in Figure 4b, that “report” when it is in use. has a tag within 10 to 15 cm of its grasping surface, can However, neither short-range nor long-range RFID therefore be identified as having been touched. The data systems, as conventionally designed, are quite up to the can either be stored on board (for later offloading task of detecting object use in a way that would be use- through a data port) or immediately radioed off board. ful for tracking activity. Short-range RFID readers are The bracelet can currently read for 30 hours between typically bulky hand-held units (similar to bar-code charges when storing data locally, and roughly 10 hours readers) that must be intentionally “swiped” on tags. when transmitting data. Clearly, it is not practical to expect a person whose Careful placement of tags on objects can reduce false activities are being tracked (whether an elder or a med- negative rates (i.e., tags being missed). However, given ical student) to carry a scanner and swipe tagged objects the range of the bracelet, “accidental” swipes of objects in the middle of day-to-day tasks. are unavoidable. Therefore, the statistical framework WINTER 2005 9

that processes the data must be able to cope with these false “hits.” Early studies indicate that an iBracelet equipped with inexpensive inductively coupled tags are a practical means of detecting object touch, and there- fore object use. Some people may consider wearing a bracelet an unacceptable requirement, however. In these cases, wireless identification and sensing platforms (WISPs) may be a useful way of detecting object use (Philipose et al., 2005). WISPs, essentially long-range RFID tags with integrated sensors, use incident energy from dis- tant readers not only to return a unique identifier, but also to power the onboard sensor and communicate the current value of the sensor to the reader. For activity- inferencing applications, so-called α-WISPs, which FIGURE 5 WISPs: a. Schematic drawing. b. A single α-WISP. c. A WISP on a have integrated accelerometers and are about the size of coffee mug. a large Band-Aid™, are attached to objects being tracked (Figure 5). When a tagged object is used, more will be seen in an observation window. In the figure, the often than not the accelerometer is triggered and the first step (corresponding to boiling tea) takes five min- ambient reader notified. utes on average; in each one-second window, there is a A single room, which may contain hundreds of tagged 40-, 20- and 30-percent chance respectively of a kettle, objects (most of them inactive at any given time), can stove, or faucet being used. Experiments in a real home be monitored by a single RFID reader. A complication with 14 subjects, each performing a randomly selected with WISPs is that the explicit correspondence between subset of 66 different activities selected from ADL the person using the object and the object being used is forms, and using activity models constructed by hand to lost. Thus, higher-level inference software may be nec- classify the resulting data automatically, have yielded essary to track the correspondence implicitly. higher than 70 percent (and often close to 90 percent) accuracy in activity detection. Inference Systems Although the models are simple, it is still impractical Given the sequence of objects detected by RFID-based to model tens of thousands of activities by hand. How- sensors, the job of the inference system is to infer the type ever, because the features to be recognized are English of activity. The inference system relies on a model that words that represent objects and the label to be mapped translates from observations (in this case, the objects to is an English phrase (such as “making tea”), the seen) to the activity label. Recent work has shown that process of building a model is essentially translating even very simple statistical models of activities are suffi- cient to distinguish between Inference Systems on Object-Use Data dozens of activities per- formed in a real home t µ = 5 min t=µ 2 min t=µ 30 sec (Philipose et al., 2004). Figure 6 shows a model for making tea, in which each activity is represented as a linear sequence of steps. Each step has a specified 0.4 0.2 0.30.6 0.2 0.4 0.4 average duration, a set of objects likely to be seen in kettle stove faucet cup teabag milk sugar that step, and the probabil- ity that one of these objects FIGURE 6 A simple probabilistic model for making tea. The 10 BRIDGE probabilistically from English phrases to words. Recent Rea and Ken Fishkin. The work on WISPs was done work based on this observation has successfully, com- with Joshua Smith and the WISP team. The work on pletely automatically, extracted translations using word mining models was done with Mike Perkowitz, Danny co-occurrence statistics from text corpora, such as the Wyatt, and Tanzeem Choudhury. Inference techniques Web (Wyatt et al., 2005). If one million Web pages were developed jointly with Dieter Fox, Henry Kautz, mention “making tea” and 600,000 of them mention and Don Patterson. “faucet,” these systems accept 60 percent as the rough probability that a faucet is used when making tea. These References crude “common-sense” models can be used as a basis for Finkenzeller, K. 2003. RFID Handbook, 2nd ed. New York: building customized models for each person by applying John Wiley & Sons. machine-learning techniques to data generated by that Fishkin, K., M. Philipose, and A. Rea. Forthcoming. Hands- person. Experiments on the data set just described have On RFID: Wireless Wearables for Detecting Use of Objects. shown that these completely automatically learned Proceedings of the International Symposium on Wearable models can recognize activities correctly roughly 50 per- Computing, Osaka, Japan, October 18–20, 2005. Wash- cent of the time. Analyses of these corpus-based tech- ington, D.C.: IEEE Computer Society. niques have also provided indirect evidence that Philipose, M., K.P. Fishkin, M. Perkowitz, D.J. Patterson, D. object-based models should be sufficient to discriminate Fox, H. Kautz, and D. Hähnel. 2004. Inferring activities between thousands of activities. from interactions with objects. Pervasive Computing (October-December):50–57. Available online at: Conclusions http://seattleweb.intel-research.net/people/matthai/pubs/ Monitoring day-to-day physical activity is a tedious pervasive_sharp_04.pdf. and expensive task now performed by human monitors. Philipose, M., J. Smith, B. Jiang, A. Mamishev, S. Roy, and K. Automated monitoring has the potential of improving Sundara-Rajan. 2005. Battery-free wireless identification the lives of many people, both monitors and those and sensing. Pervasive Computing 4(1): 37–45. being monitored. Traditional approaches to activity Sanders, B.C.S., R.C. Nelson, and R Sukthankar. 2002. The recognition have not been successful at monitoring OD Theory of TOD: The Use and Limits of Temporal large numbers of day-to-day activities in unstructured Information for Object Discovery. Pp. 777–784 in Pro- environments, partly because they were unable to iden- ceedings of the Eighteenth National Conference on Artifi- tify reliably sufficiently discriminative high-level fea- cial Intelligence. Menlo Park, Calif.: AAAI Press. tures. A new family of sensors, based on RFID, is able Tapia, E.M., S.S. Intille, and K. Larson. 2004. Activity to identify most of the objects used in activities simply Recognition in the Home Setting Using Simple and Ubiq- and accurately, and even simple statistical models can uitous Sensors. Pp. 158–175 in Proceedings of PERVA- classify large numbers of activities with reasonable SIVE 2004, vol. LNCS 3001, edited by A. Ferscha and F. accuracy. In addition, these models are simple enough Mattern. Berlin/Heidelberg: Springer-Verlag. Available that they can extract automatically from massive text online at: http://courses.media.mit.edu/2004fall/mas622j/ corpora, such as the Web, and can be customized for 04.projects/home/TapiaIntilleLarson04.pdf. observed data. Wyatt, D., M. Philipose, and T. Choudhury. 2005. Unsuper- vised Activity Recognition Using Automatically Mined Acknowledgments Common Sense. Pp. 21–27 in Proceedings of the Twenti- This paper describes work done by the author jointly eth National Conference on Artificial Intelligence and the with the SHARP group at Intel Research Seattle and Seventeenth Innovative Applications of Artificial Intelli- with researchers at the University of Washington. gence Conference. Menlo Park, Calif.: AAAI Press. Specifically, work on the iBracelet was done with Adam Ecological stewardship will be the guiding scientific principle for new avenues of inquiry.

Science and Engineering Research That Values the Planet

Arne Jacobson and Daniel M. Kammen

The recognition that human activity is transforming the planet, both in intended and dramatically unintended ways, has led to the development of a new field of research—sustainability science. Widely discussed essays (e.g., Arne Jacobson Clarke, 2002; Kates et al., 2001; Kennedy, 2003; McMichael et al., 2003; Swart et al., 2002), special issues of premier journals (NAS, 2003), and exten- sive websites (FSTS, 2005) are now devoted to defining sustainability and identifying useful modes and topics for research. Building on this foundation, we now have a tremendous opportunity to advance a new global scientific research paradigm—the generation and implementation of sustainability sci- ence. One important lesson emerges very clearly from this body of work— only by posing the question of sustainability explicitly and, where necessary, repairing the damage humans have caused to the biosphere, can we begin to understand how humans can prosper without degrading the planet. In a seminal treatise on science policy, Vannevar Bush (1945) wrote that, “applied research invariably drives out pure [research],” to the detriment, in Daniel M. Kammen his view, of the national capacity for innovation. The subsequent separation

Arne Jacobson is an assistant professor in the Environmental Resources Engineering Department and a research sci- entist at the Schatz Energy Research Center, Humboldt State University. Daniel M. Kammen is the Class of 1935 Distinguished Professor of Energy in the Energy and Resources Group, Goldman School of Public Policy, and Depart- ment of Nuclear Engineering at the University of California, Berkeley. He is also founding director of the Renew- able and Appropriate Energy Laboratory and co-director of the Berkeley Institute of the Environment. The 12 BRIDGE of basic and applied research shaped the evolution of stewardship as a guiding scientific principle, entirely science and engineering research for decades and was a new avenues of inquiry are possible. point of departure for E.F. Schumacher (1973) and the At this moment in history, this message has the “appropriate technology” movement, a precursor of sus- potential to transform research careers and make sus- tainability science that involved identifying important tainability a theme that researchers, public officials, and but neglected issues for scientific study. This approach, civil society can all embrace. The World Conference dubbed “mundane science,” (Kammen and Dove, on Physics and Sustainable Development, held in Dur- 1997), involves projects that combine pragmatic and ban, South Africa, in October and November 2005, pro- goal-oriented applied research with potential advances vided a forum for showcasing opportunities for the in basic science (Stokes, 1997). The growing recogni- co-evolution of basic research and social advances tion of the value of supporting interdisciplinary research (SAIP, 2005). and the emergence of sustainability science are contin- Currently attention, debate, and a trans-Atlantic uations of the intellectual evolution of the interaction division are focused on how to provide meaningful, between science and society. long-term aid and assistance to Africa. To highlight a The scientific recognition of the reality of global potential solution, we present two cases of sustainable environmental change (Hansen et al., 2005), the polit- science, engineering, and action in developing nations ical awareness of the need to act now to address green- that advance both science and sustainable human and house gas emissions (Kennedy, 2005), and the ecological communities. increasing disparities between the lives of the poor and the wealthy provide an opportunity for galvanizing The Energy-Health-Ecology Nexus global action to place sustainability science at the fore- Household use of solid fuels is one of the leading front of educational, research, and career-development causes of death and disease in developing countries agendas. The next step toward putting sustainable sci- throughout the world—particularly among women ence into action is recognizing that, with ecological and children (Smith et al., 2004). Over the past decade, a series of studies has been conducted of pro- US EPA standard grams to design and dis- 0.15 seminate more efficient,

Ceramic Wood Stoves safer household stoves and 3-Stone Fire to develop and implement Charcoal sustainable forestry and ARI Probability of 0.1 fuel (often charcoal) pro- respiratory duction practices in Africa. symptoms As Figure 1 shows, com- bined attention to both stove and forestry programs 0.05 can lead to dramatic simul-

ALR I taneous improvements in human health, ecological sustainability, and local 0 economic development 04,2,000 000 6,000 8,000 (Kammen, 1995). Average daily particulate exposure (µg/m 3 ) The Kenya study showed that transitions from wood FIGURE 1 The exposure-response graph from a six year, 500 person, exposure and stove intervention study in Kenya. The ver- and dung fuels burned in tical axis shows the percentage of time subjects participating in biweekly health examinations exhibited ARI or acute lower respi- simple stoves to charcoal µ 3 burned in improved stoves ratory illness (ALRI) symptoms. The EPA particulate exposure standard of 200 g/m for PM10 (particles with diameters of less than 10 microns) is indicated by the dotted vertical line, which forms a lower bound for the exposure range observed in the Kenya reduced the frequency of project. The stove and fuel combinations indicate exposure ranges. Adapted from Ezzati and Kammen, 2001. acute respiratory infections WINTER 2005 13

(ARI) by a factor of two. This is a tremendous impact on ARI, the most common illnesses reported in medi-cal exams in sub- Saharan Africa. Compara- tively simple materials and design modifications to household stoves are now known not only to improve energy efficiency, but also to reduce particulate and greenhouse gas emissions (Bailis et al., 2005). These benefits can be achieved at exceptionally low cost, just a few dollars FIGURE 2 Sales of solar modules from 1987 to 2001 in Kenya showing the dramatic increase in sales of amorphous silicon (a-Si) solar cells. The average system size is less than 25 Wp, and current annual sales exceed 30,000 individual solar elec- per life saved, and have the tric home systems. A substantial fraction of crystalline silicon (c-Si) module sales are for institutional systems that are funded added benefit of mitigating primarily through donor aid programs. Sources: ESDA, 2003; Hankins, 2000; Hankins and Bess, 1994. atmospheric carbon, at just a few dollars per ton of car- emergence and growth of the Kenyan solar market bon (Ezzati and Kammen, 2002). By contrast, carbon (Hankins, 2000; Jacobson, 2004). today trades for roughly $30/ton on the London A key aspect of these advances involved minimizing exchange, a price that reflects only the impact of green- the initial light-induced Staebler-Wronski degradation of house gases. By making the dissemination and use of a-Si modules, a poorly understood materials issue with sig- improved cookstoves a component of a comprehensive nificant implications for low-cost solar cells. The power Africa-assistance strategy, both local health and devel- output of a-Si solar modules typically decreases by 15 to opment needs and global environmental protection 40 percent during the first few months of exposure to solar could be addressed with great economic efficiency. radiation due to Staebler-Wronski degradation. Better The project in Kenya led to a number of unanticipated quality brands have lower degradation levels (Staebler advances in “basic science.” The high pollution concen- and Wronski, 1977; Su et al., 2002), and after the initial trations observed in rural African homes—as much as period of degradation, the power output stabilizes. Fig- 100 times higher than those observed in the urban areas ure 3 shows degradation curves for two different brands of many industrialized nations—provided a laboratory for of a-Si modules, showing that the initial power output of examining the epidemiology of exposure-response in some brands drops significantly more than others. The a pollution regime that had not been studied before rated power of most reputable brands of a-Si PV modules (Ezzati and Kammen, 2001). These studies have greatly corresponds to the final, stabilized power output under extended the cutting-edge epidemiological work being standard test conditions of 1,000 W/m2 and 25°C. done largely in developed nations (Rich et al., 2005). A second important design issue has been the devel- opment of cost-effective sealant materials and methods Solar Electricity Markets in Developing Nations of preventing delamination. Water intrusion can lead Household solar photovoltaics (PV) have emerged as to outright module failure, and the actual power output the leading alternative to grid-based rural electrification of modules with significant delamination is often in many developing countries. In Kenya, 30,000 PV sys- reduced to less than 10 percent of the nameplate power tems are sold annually, making it a global leader, per rating. Figure 4 shows water-induced delamination in capita, in sales of residential renewable energy systems an a-Si module caused by low-quality seals. A number (Figure 2). Advances in amorphous silicon (a-Si) PV of a-Si manufacturers have developed highly effective technology, which led to the development of small, sealing techniques, but a few brands continue to have low-cost a-Si PV modules, played a critical role in the water-intrusion problems. The 14 BRIDGE

low-performing products from the market, as well as insights into institutional aspects of renewable en- ergy market development (Acker and Kammen, 1996; Duke et al., 2002; Jacobson, 2004; Kammen, 1995).

Making Sustainable Science the Norm The first step in making sustainable science the norm is to demonstrate that, once funding and a research/ action team have been assembled, these projects are no more difficult than tradi- FIGURE 3 Performance of two brands of a-Si solar modules during the first few months of exposure to solar radiation showing tional research projects. To substantial differences in light-induced Staebler-Wronski degradation for the two brands. The power output of the Brand C mod- be effective, however, pro- ule, although initially higher than its 14W power rating, drops far below its nameplate rating after several months in service. By jects must be neither exclu- contrast, the performance of the Brand B module stabilizes near the 12W rating. Note that these results are from 2000 and do sively in the academic or not reflect recent improvements for Brand C (shown in Figure 5). Source: Jacobson et al., 2000. laboratory setting, nor These advances have been important for the PV entirely in the sphere of nonprofit organizations or local industry as a whole, but have been especially significant governments. To take maximum advantage of both the for rural electrification with solar energy in developing emerging science and the implementation capacity for countries. In contrast to laboratory and commercial sustainability, we must demonstrate support in each of rivalries over which company produces the most the disciplines involved, both through actions and thermodynamically efficient solar cells, the firms that funding priorities. manufacture a-Si PV modules for markets in developing Second, we must make sustainability science a basic countries have focused on lower efficiency but signifi- precept of teaching in secondary schools, colleges, and cantly less expensive products (Green et al., 2005). The postgraduate studies. Pre-college students have already resulting 12 to 20W a-Si PV modules now available in demonstrated a tremendous aptitude for working in Kenya and elsewhere cost 50 percent less than compa- interdisciplinary areas. We must nurture and reward rable crystalline silicon (c-Si) PV modules, and are, by far, the best-selling solar products in the region. The dissemination of a-Si PV technology in Kenya has not, however, been without complications. In an extensive market survey (Figure 5), we found that, although most manufacturers produce high-quality products, one prominent brand performed well below its advertised levels. A previous study in 1999 showed a similar pattern, although for a different brand. Thus, the successful deployment of new technology requires market institutions that ensure quality and protect the public interest. The combination of technical studies of solar equipment performance and analyses of Kenyan market development, socio-cultural dynamics, and reg- FIGURE 4 Water-intrusion-related delamination in a Brand D a-Si PV module. The ulatory policy has led to progress toward eliminating actual power output of this 14W rated module was less than 1W. WINTER 2005 15

(2005) are models that could be adapted to the theme of sustainability science. The launch of Sputnik in 1957 initiated an unprece- dented mobilization of U.S. science and technology, a lesson in the power of a use-inspired drive to innovate. The Yale Environment Survey found overwhelming interest in energy and environmental sustainability (Yale University, 2005). Contrast that interest with the results of the 3rd International Mathematics and Sci- ence Study (TIMSS), in which American secondary school students ranked 19th out of 21 countries in both math and science (NRC, 1997). The TIMMS authors concluded that science and mathematics education in the United States lacked direction, vision, and motiva- tion. Sustainability science could give science, mathe- matics, and engineering education renewed meaning and immediacy, with paradigm-changing possibilities in both developed and developing nations. Third, we could establish sustainability awards— modeled after the Ashoka Innovators Awards (2005), the Ansari X Prize (X Prize Foundation, 2005) for the launch of a space vehicle, and the Ashden Awards (2005) for sustainable energy. These awards would bring together partners from developed and developing nations in academia, industry, civil society, and govern- ment and would encourage groups to take action on critical sustainability projects. Ideally, sustainability awards, jointly sponsored by private foundations and state or federal governments, would take advantage of the diversity of perspectives and skills that interdiscipli- nary, international teams would bring together. Finally, we must address the principal weakness in the economies of many poor nations—a lack of capacity to FIGURE 5 Average stabilized maximum power output results from 1999 and 2004–2005 for a-Si solar modules sold in Kenya. Aggregate test results for sev- compete in the global marketplace. Debt forgiveness for eral brands of c-Si modules are included for comparison. Note that, although most impoverished countries in Africa and elsewhere is laud- a-Si brands have power output levels similar to the more expensive c-Si modules, able (Sachs, 2005), but it has already been criticized by some brands perform well below their advertised power ratings. The 1999 test African leaders who have noted that aid alone is not a results are based on field measurements of 130 a-Si modules and 17 c-Si modules. panacea. Estimates of the percentage of overall eco- The 2004–2005 results involved 20 a-Si modules randomly selected from Kenyan nomic growth from innovation in science and technol- retail shops. The presence of low-performing brands has led to considerable acrimony ogy, virtually all in industrialized nations, are as high in the Kenyan solar industry, as indicated in the “Solar Scandal” advertisement from as 90 percent (Solow, 2000). Developing economies a local newspaper. Following the release in Kenya of the 2004–2005 results, the would be energized by dramatically increased invest- market presence of Brand D dropped. Source: Jacobson and Kammen, 2005. ment in indigenous innovation. A natural way to do that would be to reward investments in science and this interest with courses in junior high schools, high technology capacity for sustainable development with schools, and colleges on energy, the environment, additional debt relief or more favorable trade arrange- and the social drivers of resource degradation. In the ments. This is a perfect time for the G8 to adopt this United States, the Upward Bound Math-Science plan and assist all nations to invest in environmentally Program (DOEd, 2005) and Summer Science Program conscious innovation. The 16 BRIDGE

Acknowledgements People: The Kenya Case. Report for the Joint UNDP/World This work was supported by the Energy Foundation, Bank Energy Sector Management Assistance Programme the Class of 1935 of the University of California, Berke- (ESMAP). Washington, D.C.: ESMAP, World Bank. ley, the Schatz Energy Research Center, and a grant from Hansen, J., L. Nazarenko, R. Ruedy, M. Sato, J. Willis, A. Del the Gordon and Betty Moore Foundation. We thank Genio, D. Koch, A. Lacis, K. Lo, S. Menon, T. Novakov, J. Robert Bailis and Zia Mian for their insightful comments. Perlwitz, G. Russell, G. Schmidt, and N. Tausnev. 2005. Earth’s energy imbalance: confirmation and implications. References Science 308(5727): 1431–1435. Acker, R., and D.M. Kammen. 1996. The quiet (energy) rev- Jacobson, A. 2004. Connective Power: Solar Electrification olution: the diffusion of photovoltaic power systems in and Social Change in Kenya. Unpublished Ph.D. disserta- Kenya. Energy Policy 24(1): 81–111. tion, University of California, Berkeley. Ashden Awards. 2005. The Ashden Awards for Sustainabil- Jacobson, A., and D.M. Kammen. 2005. Engineering, insti- ity. Available online at: http://www.ashdenawards.org. tutions, and the public interest: evaluating product quality Ashoka. 2005. Ashoka Innovators Awards. Available online in the Kenyan solar photovoltaics industry. Submitted to at: http://www.ashoka.org. Energy Policy. Bailis, R., M. Ezzati, and D.M. Kammen. 2005. Mortality and Jacobson, A., R. Duke, D.M. Kammen, and M. Hankins. 2000. greenhouse gas impacts of biomass and petroleum energy Field Performance Measurements of Amorphous Silicon futures in Africa. Science 308(5718): 98–103. Photovoltaic Modules in Kenya. In ASES Annual and Bush, V. 1945. Science—The Endless Frontier: A Report to National Passive Solar Conference Proceedings Combined the President. Washington, D.C.: U.S. Government Print- 2000. Madison, Wis.: American Solar Energy Society. ing Office. Kammen, D.M. 1995. Cookstoves for the developing world. Clarke, T. 2002. Wanted: scientists for sustainability. Nature Scientific American 273(1): 72–75. 418(6900): 812–814. Kammen, D.M., and M.R. Dove. 1997. The virtues of mun- DOEd (U.S. Department of Education). 2005. Upward dane science. Environment 39(6): 10–15, 38–41. Bound Math-Science Program. Available online at: Kates, R.W., W.C. Clark, R. Corell, J.M. Hall, C.C. Jaeger, I. http://www.ed.gov/programs/triomathsci/index.html. Lowe, J.J. McCarthy, H.J. Schellnhuber, B. Bolin, N.M. Duke, R., A. Jacobson, and D.M. Kammen. 2002. Photo- Dickson, S. Faucheux, G.C. Gallopin, A. Grübler, B. Hunt- voltaic module quality in the Kenyan solar home systems ley, J. Jäger, N.S. Jodha, R.E. Kasperson, A. Mabogunje, P. market. Energy Policy 30(6): 477–499. Matson, H. Mooney, B. Moore III, T. O’Riordan, and U. ESDA (Energy for Sustainable Development, Africa). 2003. Svedin. 2001. Sustainability science. Science 292(5517): Study on PV Market Chains in East Africa. Report for the 641–642. World Bank. Nairobi, Kenya: ESDA. Kennedy, D. 2003. Editorial: Sustainability and the Com- Ezzati, M., and D.M. Kammen. 2001. Indoor air pollution mons. Science 302(5652): 1861. from biomass combustion as a risk factor for acute respira- Kennedy, D. 2005. Editorial: The Fight of the Decade. Sci- tory infections in Kenya: an exposure-response study. ence 308(5729): 1713. Lancet 358(9282): 619–624. McMichael, A.J., C.D. Butler, and C. Folke. 2003. New Ezzati, M., and D.M. Kammen. 2002. Evaluating the health visions for addressing sustainability. Science 302(5652): benefits of transitions in household energy technologies in 1919–1920. Kenya. Energy Policy 30(10): 815–826. NAS (National Academy of Sciences). 2003. Special sec- FSTS (Forum on Science and Technology for Sustainability). tion of the July 8 issue devoted to science in support of sus- 2005. Forum on Science and Technology for Sustainabili- tainability. Proceedings of the National Academy of ty. Available online at: http://sustsci.harvard.edu/. Sciences 100(14). Green, M.A., K. Emery, D.L. King, S. Igari, and W. Warta. NRC (National Research Council). 1997. A Splintered 2005. Short communication: solar cell efficiency tables Vision: An Investigation of U.S. Science and Mathematics (version 26). Progress in Photovoltaics: Research and Education, edited by W.H. Schmidt, S.A. Raizen, and C.C. Applications 13(5): 387–392. McKnight. Third International Mathematics and Science Hankins, M. 2000. Energy Services for the World’s Poor. Study. Washington, D.C.: National Academy Press. Washington, D.C.: ESMAP, World Bank. Rich, D.Q., J. Schwartz, M.A. Mittleman, M. Link, H. Hankins, M., and M. Bess. 1994. Photovoltaic Power to the Luttmann-Gibson, P.J. Catalano, F.E. Speizer, and D.W. WINTER 2005 17

Dockery. 2005. Association of short-term ambient air pol- Staebler, D.L., and C.R. Wronski. 1977. Reversible conduc- lution concentrations and ventricular arrhythmias. Amer- tivity changes in discharge-produced amorphous Si. ican Journal of Epidemiology 161(12): 1123–1132. Applied Physics Letters 31(4): 292–294. Sachs, J.D. 2005. Four easy pieces. New York Times, p. A15, Stokes, D. 1997. Pasteur’s Quadrant. Washington, D.C.: June 25, 2005. Brookings Institute Press. SAIP (South African Institute of Physics). 2005. World Su, T., P.C. Taylor, G. Ganguly, and D.E. Carlson. 2002. Conference on Physics and Sustainable Development, Direct role of hydrogen in the Staebler-Wronski effect in Durban, South Africa, 31 October to 2 November 2005. hydrogenated amorphous silicon. Physical Review Letters Available online at: http://www.saip.org.za/physics2005/ 89(1): 015502. WCPSD2005.html. Summer Science Program. 2005. The Summer Science Pro- Schumacher, E.F. 1973. Small Is Beautiful. New York: Harp- gram. Available online at: http://www.summerscience.org/ er and Row. home/index.php. Smith, K.R., S. Mehta, and M. Maeusezahl-Feuz. 2004. Swart, R., P. Raskin, J. Robinson, R. Kates, and W.C. Clark. Indoor Air Pollution from Household Use of Solid Fuels. 2002. Critical challenges for sustainability science. Sci- Pp. 1435–1493 in Comparative Quantification of Health ence 297(5589): 1994–1995. Risks: Global and Regional Burden of Disease Attributable X Prize Foundation. 2005. X Prize. Available online at: to Selected Major Risk Factors, edited by M. Ezzati, A.D. http://www.xprizefoundation.com. Lopez, A. Rodgers, and C.J.L. Murray. Geneva: World Yale University. 2005. Yale Center for Environmental Law Health Organization. and Policy. Available online at: http://www.yale.edu/ Solow, W.M. 2000. Growth Theory: An Exposition. Oxford, envirocenter/environmentalpoll.htm. U.K.: Oxford University Press. Synthetic biologists may soon design and build engineered biological systems.

The Promise of Synthetic Biology

Jay Keasling

It has been estimated that for every successful drug compound, 5,000 to 10,000 compounds must be introduced into the drug-discovery pipeline. On average, it takes $802 million and 10 to 15 years to develop a successful drug. Given this very low success rate and the incredibly high costs, drug compa- nies must introduce as many drug candidates into their pipelines as possible. Natural products have been important sources of drug leads; as much as Jay Keasling is director of the 60 percent of successful drugs are of natural origin (Cragg et al., 1997), and Berkeley Center for Synthetic Biol- some of the most potent natural products have been used as anticancer, ogy, University of California, antibacterial, and antifungal drugs. However, most natural products evolved for purposes other than the treatment of human disease. Thus, even though Berkeley, and Lawrence Berkeley they can sometimes function as human therapeutics, their pharmacological National Laboratory. properties may not be optimal. Furthermore, many are produced in minis- cule amounts in their native hosts, thus making them expensive to harvest. Organic chemistry methodologies are widely used to synthesize pharma- ceuticals (of natural origin or not) and functionalize pharmaceutically rele- vant natural products. With appropriate protection and deprotection steps, chiral centers and functionalities can be introduced into molecules with precision. With the advent of combinatorial chemical synthesis, researchers have been able to construct entire families of molecules substituted at several positions with several different substituents, thus allowing drug com- panies to fill drug-discovery pipelines with variations of promising leads. WINTER 2005 19

Despite the creation of complicated molecules made Although redesigned biological control systems have possible by advances in organic synthesis methodolo- been generally effective for their intended purposes gies, the performance of these molecules is hardly com- (controlling rather roughly the expression of a single parable to the ease, specificity, and “green-ness” of gene or a few genes), not surprisingly they are often enzymes. Indeed, many organic synthesis routes now inadequate for more complicated engineering tasks incorporate one or more enzymes to perform transfor- (e.g., controlling very large, heterologous, metabolic mations that are particularly difficult using nonenzy- pathways or signal transduction systems). In addition, matic routes. Furthermore, enzymes are now being used these borrowed “biological parts” retain many of the fea- for the in vitro, combinatorial functionalization of com- tures that were beneficial in their native forms but make plex molecules. The next logical step in the synthesis of chemotherapeutics is the use of enzymes for combinato- rial synthesis inside the cell, which would allow the pro- duction of drug candidates from inexpensive starting No standards have been materials and avoid the need for purification of enzymes, which may be necessary for in vitro synthesis. defined for building larger

Biological Engineering for the Synthesis of Drugs biological devices. Rich, versatile biological systems are ideally suited to solving some of the world’s most significant challenges, such as converting cheap, renewable resources into them difficult to use for purposes other than the ones for energy-rich molecules; producing high-quality, inex- which they evolved. Well characterized standard bio- pensive drugs to fight disease; detecting and destroying logical parts, and larger devices made from such parts, chemical or biological agents; and remediating polluted would make biological engineering more predictable sites. Over the years, significant strides have been made and enable the construction and integration of larger in engineering microorganisms to produce ethanol, bulk systems than is currently possible. chemicals, and valuable drugs from inexpensive starting In almost every other field of engineering, standards materials; to detect and degrade nerve agents as well as have been developed for building large integrated systems less toxic organic pollutants; and to accumulate metals by assembling components from various manufacturers. and reduce radionuclides. However, biologists and engineers have not yet defined However, meeting these biological engineering chal- standards for the parts that might allow them to build lenges requires long development times, largely because larger biological devices. The design and construction of of a lack of useful tools that would enable engineers to new devices (e.g., genetic-control systems) would benefit easily and predictably reprogram existing systems, let greatly from standards governing how various parts (e.g., alone build new enzymes, signal transduction pathways, regulatory proteins, promoters, ribosome binding sites) genetic circuits, and, eventually, whole cells. The ready should interact and be assembled. Setting standards would availability of these tools would drastically alter the also encourage manufacturing firms to develop parts. biotechnology industry, leading to less expensive phar- Biological engineering has been held back because maceuticals, renewable energy, and biological solutions many of the most effective biological parts (promoters, to problems that do not currently offer sufficient mone- genes, plasmids, etc.) have been patented and are avail- tary returns to justify the high cost of biological research. able only to companies that can afford the royalty Most of the biological engineering tools currently payments. This has not only increased the cost of drug available to scientists and engineers have not changed development, but also hampered the development of significantly since genetic engineering began in the new biological solutions to problems that may not have 1970s. Biologists still use natural, gene-expression con- significant monetary payoffs (basically, anything other trol systems (promoters with cognate repressors/activa- than drug development). Open-source biological parts, tors). The ability to place a single heterologous gene devices, and eventually whole cells would reduce the under the control of one of these native promoters and cost of engineering biological systems, make biological produce large quantities of a protein of interest is the engineering more predictable, and encourage the basis for the modern biotechnology industry. development of novel biological solutions to some of The 20 BRIDGE our most challenging problems. The development of to a better understanding, and new types, of biological open-source biological technology would improve components and systems. Third, natural living systems awareness of, and minimize possible future biological evolved to ensure their continued existence; they are risks, in the same way that open-source software tends not optimized for human understanding and intention. to promote a constructive and responsive community By thoughtfully redesigning natural living systems, it is of users and developers. possible simultaneously to test our current understand- ing and potentially implement engineered systems that Synthetic Biology are easier to interact with and study. Fourth, biology can Synthetic biology is the design and construction of be used as a technology, and biotechnology, broadly new biological entities, such as enzymes, genetic circuits, redefined, includes the engineering of integrated bio- and cells, or the redesign of existing biological systems. logical systems for the purposes of processing informa- The goal of synthetic biology, which builds on advances tion, producing energy, manufacturing chemicals, and in molecular, cellular, and systems biology, is to transform fabricating materials. biology in the same way that synthesis transformed Although the emergence of the discipline of synthetic chemistry and integrated circuit design transformed com- biology was motivated by these agendas, progress has only puting. The element that distinguishes synthetic biology been practical since the recent advent of two founda- from traditional molecular and cellular biology is the tional technologies, DNA sequencing, which has focus on (1) the design and construction of core compo- increased our understanding of the components and orga- nents (parts of enzymes, genetic circuits, metabolic path- nization of natural biological systems, and synthesis, ways, etc.) that can be modeled, understood, and which has enabled us to begin to test the designs of engineered to meet specific performance criteria, and (1) new, synthetic biological parts (Allert et al., 2004; (2) the assembly of these smaller parts and devices into Basu et al., 2004; Becskei and Serrano, 2000; Cane et al., larger integrated systems to solve specific problems. Just 2002; Datsenko and Wanner, 2000; De Luca and as engineers now design integrated circuits based on the Laflamme, 2001; Dwyer and Hellinga, 2004; Gardner and known physical properties of materials and then fabricate Collins, 2000; Gardner et al., 2000; Geerlings et al., 2001; functioning circuits and entire processors (with relatively Gerasimenko et al., 2002; Godfrin-Estevenon et al., high reliability), synthetic biologists will soon design and 2002; Guet et al., 2002; Kobayashi et al., 2004; McDaniel build engineered biological systems. et al., 1997) and (2) new biological systems (Bignell and Unlike many other areas of engineering, however, Thomas, 2001; Blake and Isaacs, 2004; Hughes and biology is nonlinear and less predictable, and much less Shanks, 2002; Iijima et al., 2004; Irmler et al., 2000; Judd is known about parts and how they interact. Hence, the et al., 2000; Kumar et al., 2004; Le Borgne et al., 2001; overwhelming physical details of natural biology (gene Martin et al., 2001, 2002, 2003; Okamoto et al., 2004). sequences, protein properties, biological systems) must Each of these examples demonstrates the incredible be organized and recast via a set of design rules that hide potential of synthetic biology, as well as the foundational information and manage complexity, thereby enabling scientific and engineering challenges that must be met for the engineering of multicomponent integrated biologi- the engineering of biology to become routine. cal systems. Only when this is accomplished will designs of significant scale be possible. References Synthetic biology arose from four different intellec- Allert, M., S.S. Rizk, L.L. Looger, and H.W. Hellinga. 2004. tual premises. The first is the scientific idea that a prac- Computational design of receptors for an organophosphate tical test of understanding is the ability to reconstitute surrogate of the nerve agent soman. Proceedings of the a functional system from its basic parts. Using synthetic National Academy of Sciences 101(21): 7907–7912. biology, scientists are testing models of how biology Basu, S., R. Mehreja, S. Thiberge, M.-T. Chen, and R. Weiss. works by building systems based on models and measur- 2004. Spatiotemporal control of gene expression with ing differences between expectations and observations. pulse-generating networks. Proceedings of the National Second, some consider biology an extension of chem- Academy of Sciences 101(17): 6355–6360. istry, and thus synthetic biology can be considered an Becskei, A., and L. Serrano. 2000. Engineering stability in extension of synthetic chemistry. Attempts to manipu- gene networks by autoregulation. Nature 405(6786): late living systems at the molecular level will likely lead 590–593. WINTER 2005 21

Bignell, C., and C.M. Thomas. 2001. The bacterial ParA- Hughes, E.H., and J.V. Shanks. 2002. Metabolic engineering ParB partitioning proteins. Journal of Biotechnology of plants for alkaloid production. Metabolic Engineering 91(1): 1–34. 4(1): 41–48. Blake, W.J., and F.J. Isaacs. 2004. Synthetic biology evolves. Iijima, Y., D.R. Gang, E. Fridman, E. Lewinsohn, and E. Pich- Trends in Biotechnology 22(7): 321–324. ersky. 2004. Characterization of geraniol synthase from the Cane, D.E., F. Kudo, K. Kinoshita, and C. Khosla. 2002. peltate glands of sweet basil. Plant Physiology 134(1): Precursor-directed biosynthesis: biochemical basis of the 370–379. remarkable selectivity of the erythromycin polyketide Irmler, S., G. Schroder, B. St. Pierre, N.P. Crouch, M. Hotze, synthase toward unsaturated triketides. Chemistry and J. Schmidt, D. Strack, U. Matern, and J. Schroder. 2000. Biology 9(1): 131–142. Indole alkaloid biosynthesis in Catharanthus roseus: new Cragg, G.M., D.J. Newman, and K.M. Snader. 1997. enzyme activities and identification of cytochrome P450 Natural products in drug discovery and development. CYP72A1 as secologanin synthase. The Plant Journal Journal of Natural Products 60(1): 52–60. 24(6): 797–804. Datsenko, K.A., and B.L. Wanner. 2000. One-step inactiva- Judd, E.M., M.T. Laub, and H.H. McAdams. 2000. Toggles tion of chromosomal genes in Escherichia coli K-12 using and oscillators: new genetic circuit designs. Bioessays PCR products. Proceedings of the National Academy of 22(6): 507–509. Sciences 97(12): 6640–6645. Kobayashi, H., M. Kaern, M. Araki, K. Chung, T.S. Gardner, De Luca, V., and P. Laflamme. 2001. The expanding universe C.R. Cantor, and J.J. Collins. 2004. Programmable cells: of alkaloid biosynthesis. Current Opinion in Plant Biology interfacing natural and engineered gene networks. Pro- 4(3): 225–233. ceedings of the National Academy of Sciences 101(22): Dwyer, M.A., and H.W. Hellinga. 2004. Periplasmic binding 8414–8419. proteins: a versatile superfamily for protein engineering. Kumar, P., C. Khosla, and Y. Tang. 2004. Manipulation and Current Opinion in Structural Biology 14(4): 495–504. analysis of polyketide synthases. Methods in Enzymology Gardner, T.S., C.R. Cantor, and J.J. Collins. 2000. Con- 388: 269–293. struction of a genetic toggle switch in Escherichia coli. Le Borgne, S., B. Palmeros, F. Bolivar, and G. Gosset. 2001. Nature 403(6767): 339–342. Improvement of the pBRINT-Ts plasmid family to obtain Gardner, T.S., and J.J. Collins. 2000. Neutralizing noise in marker-free chromosomal insertion of cloned DNA in gene networks. Nature 405(6786): 520–521. E. coli. Biotechniques 30(2): 252–254, 256. Geerlings, A., F.J. Redondo, A. Contin, J. Memelink, R. van Martin, V.J.J., C.D. Smolke, and J.D. Keasling. 2002. der Heijden, and R. Verpoorte. 2001. Biotransformation of Redesigning cells for production of complex organic mole- tryptamine and secologanin into plant terpenoid indole cules. ASM News 68: 336–343. alkaloids by transgenic yeast. Applied Microbiology Martin, V.J.J., D.J. Pitera, S.T. Withers, J.D. Newman, and Biotechnology 56(3-4): 420–424. J.D. Keasling. 2003. Engineering the mevalonate pathway Gerasimenko, I., Y. Sheludko, X. Ma, and J. Stockigt. 2002. in Escherichia coli for production of terpenoids. Nature Heterologous expression of a Rauvolfia cDNA encoding Biotechnology 21(7): 796–802. strictosidine glucosidase, a biosynthetic key to over 2000 Martin, V.J.J., Y. Yoshikuni, and J.D. Keasling. 2001. The in monoterpenoid indole alkaloids. European Journal of Bio- vivo synthesis of plant sesquiterpenes in Escherichia coli. chemistry 269(8): 2204–2213. Biotechnology and Bioengineering 75(5): 497–503. Godfrin-Estevenon, A.M., F. Pasta, and D. Lane. 2002. The McDaniel, R., C.M. Kao, S.J. Hwang, and C. Khosla. 1997. parAB gene products of Pseudomonas putida exhibit parti- Engineered intermodular and intramodular polyketide syn- tion activity in both P. putida and Escherichia coli. Mole- thase fusions. Chemistry and Biology 4(9): 667–674. cular Microbiology 43(1): 39–49. Okamoto, A., K. Tanaka, and I. Saito. 2004. DNA logic Guet, C.C., M.B. Elowitz, W. Hsing, and S. Leibler. 2002. gates. Journal of the American Chemical Society 126(30): Combinatorial synthesis of genetic networks. Science 9458–9463. 296(5572): 1466–1470. Control over agent-based systems can be achieved via modeling tools.

Agent-Based Modeling as a Decision-Making Tool

Zoltán Toroczkai and Stephen Eubank

Researchers have made considerable advances in the quantitative charac- terization, understanding, and control of nonliving systems. We are rather familiar with physical and chemical systems, ranging from elementary parti- Zoltán Toroczkai cles, atoms, and molecules to proteins, polymers, fluids, and solids. These sys- tems have interacting particles and well defined physical interactions, and their properties can be described by the known laws of physics and chemistry. Most important, given the same initial conditions, their behavior is repro- ducible (at least statistically). However, other types of ubiquitous systems are all around us, namely sys- tems that involve living entities (i.e., agents) about which we have hardly any quantitative understanding, either on an individual or collective level. In this paper, we refer to collectives of living entities as “agent-based systems” or “agent systems” to distinguish them from classical particle systems of inan- imate objects. Although intense efforts have been made to study these sys- tems, no generally accepted unifying framework has been found. Stephen Eubank Nevertheless, understanding, and ultimately controlling the behavior of agent systems, which have applications from biology to the social and political sciences to economics, is extremely important. Ultimately, a quantitative

Zoltán Toroczkai is deputy director of the Center for Nonlinear Studies at Los Alamos National Laboratory. Stephen Eubank is deputy director, Networks Dynamic and Simulation Science Laboratory, Virginia Bioinformatics Institute, Virginia Polytechnic Institute and State University. WINTER 2005 23

understanding can be a basis for designing agent systems, •A set of variables, z, describes the perceived state of the like robots or rovers that can perform tasks collectively environment, Z, which includes other agents, if there that would be prohibitive for humans. Examples include are any (e.g., level of congestion, state/quality of the deep-water rescue missions, minefield mapping, distrib- road, weather conditions, etc.). uted sensor networks (for civil and military uses), and • There is a set of allowable actions (output space), A rovers for extraterrestrial exploration. (swerve, brake, accelerate, etc.). Even though there is no unifying understanding of agent systems, some control over their behavior can be •A set of strategies, which are functions, s: (Z×X)→ A, achieved via agent-based modeling tools. The idea summon an action to a given circumstance, current behind agent-based modeling is rather simple—build a state of the agent, and history up to time, t. These are computer model of the agent system under observation “ways of reasoning” for the agent. One might think using a bottom-up approach by trying to mimic as much of strategies as behavioral input space. For example, detail as possible. This can be rather expensive, how- depending on age, background, and other factors, ever, because it requires (1) data collection, (2) model some drivers will brake and some will swerve to avoid building, (3) exploitation of the model and the collec- an accident. Social studies and surveys can supply tion of statistics, and (4) validation, which normally valuable statistical inputs, such as data showing that means comparing the output of the model with addi- agents with n years of driving experience between tional observations of the real system. ages a1 and a2 swerve f percent of the time and break The agent models described in this paper took about g percent of the time. nine years to develop at Los Alamos National Labora- • There is a set of utility variables, u∈U (e.g., time to tory. However, the framework for these models can be destination, number of accidents, number of speeding used to simulate many similar circumstances and to tickets, etc.). make predictions. •A multivariate objective function, F:U → Rm, might Properties of Agent Systems include constraints (“rules”) (e.g., the agent has to Agent-based systems are hard to describe and under- stay on the road). The analogous version in physics stand within a unified approach because they differ from is called action. The agent tries to optimize this objec- classical particle systems in at least two ways. First, an tive function (e.g., by minimizing the time to desti- agent is a complex entity that cannot be represented by nation, avoiding accidents, etc.). a simple function, such as a Hamiltonian function of a classical system (e.g., a spin system). Second, the inter- action topology, namely the rules by which particles interact with each other, is generally represented by a The framework for developed complex, dynamic graph (network), unlike the regular models can be used to lattices of crystalline solids or the continuous spaces of fluid dynamics. In many cases, the notion of “locality” simulate other agent systems. itself is elusive in agent-based networks; in social net- works, for example, the physical or spatial locality of agents may have little to do with social “distances” and Unlike particles in classical systems, agents usually interactions among them. To illustrate the complex have memory, which they can use to change/evolve structure of a “particle,” or agent, and its consequences strategies, a process called learning. Another important we can use traffic, namely people (agents) driving on a aspect of agents is that they can reason and plan, which highway, as an example. Keep in mind, however, that entail searching the choice tree and assigning weights the statements in this description are generally applica- and payoffs in light of what other agents might choose. ble to other agent systems. In realistic situations that involve hundreds of agents Agents have the following qualities: (such as markets or traffic), long-term planning and •A set of variables, x, describes the state of the agent reasoning are impossible because of the combina- (e.g., position on the road, speed, health of the driver, torial explosion of possibilities and also because not all etc.). The corresponding state space is X. of the information is available to any single agent. The 24 BRIDGE

Therefore, agents try to identify and exploit patterns in households that included observations of daily activity the responses of the surrounding environment to past patterns for each individual in the household. These actions and use these patterns to discriminate among activity patterns were then associated with synthetic strategies, reinforcing some and diminishing others (re- households with similar demographics. The locations inforcement learning). This leads to bounded-rationality- of various activities were estimated, taking into account like behavior and introduces de-correlations between observed land-use patterns, travel times, and the dollar strategies; for that reason, reinforcement learning actu- costs of transportation. ally makes statistical modeling plausible. For the intermediate timescale (Figure 2), routes and In the following sections, we briefly describe two large- trip chains were assigned to satisfy the activity requests. scale agent-based models developed at Los Alamos The estimated locations were fed into a routing algo- National Laboratory, a traffic simulator (TRANSIMS) rithm to find minimum-cost paths through the trans- and an epidemics simulator (EPISIMS). portation infrastructure consistent with constraints on mode choices (Barrett et al., 2001, 2002). For ex- ample, a constraint might be “walk to a transit stop, take transit to work using no more than two transfers and no more than one bus.” Finally, a very short time- scale (Figure 3) was used, associated with the actual execution of trip plans in the road network. This is done by a cellular automata simulation through a very detailed representation of the urban transportation network. The simulation, FIGURE 1 Large-scale view of the roadway network in Portland, Oregon. which resolved distances down to 7.5 meters and times down to 1 second, in TRANSIMS effect resolved the traffic congestion caused by everyone The transportation analysis and simulation system trying to execute plans simultaneously by providing (TRANSIMS) is an agent-based model of traffic in a updated estimates of time-dependent travel times for particular urban area (the first model was for Portland, each edge in the network, including the effects of con- Oregon). TRANSIMS conceptually decomposes the gestion. These estimates were fed to a router and loca- transportation planning task using three different time- tion estimation algorithms that produced new plans. scales. A large time scale associated with land use and The feedback process continued iteratively until demographic distribution (Figure 1) was used to create the system converged in a “quasi-steady state” in which activity categories for travelers (e.g., work, shopping, no agent could find a better path in the context of entertainment, school, etc.). Activity information typ- every other agent’s decisions. The resulting traffic pat- ically consisted of (1) requests that travelers be at a cer- terns compared well to observed traffic. Thus, the tain location at a specified time and (2) information on entire process estimated the demand on a transporta- travel modes available to the traveler. For the large tion network using census data, land-usage data, and timescale, a synthetic population was created and activity surveys.1 endowed with demographics matching the joint distri- butions in census data. Synthetic households were also 1 More information, including availability of the software, can be created based on survey data from several thousand obtained from http://transims.tsasa.lanl.gov. WINTER 2005 25

location node, l, the common intersection of the two time stamps told us the total time the two people spent in proximity to each other during the day, thus enabling us to determine the possibility of transmission of an air- borne infection. Using Portland as an example, there were about 1.6 million people nodes, 181,000 location nodes, and more than 6 million edges between them. This dynamic contact graph enabled us to simulate different disease- spread scenarios and test the sensitivities of epidemics to disease parameters, such as incubation period, person- FIGURE 2 Intermediate timescale view of locations and roads in downtown to-person infection rates, influence of age structure, Portland, Oregon. activity patterns, and so on. The epidemiological study tool thus generated, called EPISIMS, can be used as an aid to decision making and planning, for example, for an outbreak of smallpox. Based on the Portland data, we arrived at the follow- ing findings for the spread of smallpox.2 First, a person who has been vaccinated can be removed from the con- tact graph, along with his or her incident links. Thus, an efficient vaccination strategy removes the smallest subset of nodes, so that the resulting graph has many small disconnected pieces, which eliminates the spread of disease throughout the population. Unfortunately, the smallpox vaccine is not entirely harmless. In some people, it causes a disease called vaccinia that is some- FIGURE 3 The TRANSIMS very-small timescale (microsimulation) of downtown times fatal. Therefore, to minimize the incidence of Portland. vaccinia, mass vaccination (proposed by Kaplan et al., 2002) must be a last resort. EPISIMS After studying the projection of the bipartite graph Another application of the TRANSIMS model is in onto people nodes, however, we found very high expan- the field of epidemiology. Diseases, such as colds, flu, sion properties, and the only way to avoid mass spread smallpox, and SARS, are transmitted through the air of the disease would be to vaccinate everyone who had between two agents. These agents must either spend a 10 or more contacts during the day, which effectively long enough time in proximity to each other or be in meant mass vaccination. Ultimately, we found that vac- a building with a closed air-ventilation system to trans- cinating people who frequently took long-range trips mit the disease. Thus, we can assume that the majority across the city, corresponding to shortcuts in the net- of infections take place in locations, like offices, shop- work (Watts and Strogatz, 1998), made it possible for us ping malls, entertainment centers, and mass transit to use a more localized graph with a larger diameter. In units (metros, trams, etc.). case of an outbreak, this would allow for a ring strategy By tracking the people in our TRANSIMS virtual city, for quarantining and further vaccinations to stop the we generated a bipartite contact network, or graph, formed spread of disease. by two types of nodes—people nodes and location Finally, the crucial parameter in containing an epi- nodes. If a person, p, entered a location, l, an edge was demic is the delay in reaction time. If we assume that drawn between that person and the corresponding loca- sensors can perform an online analysis of pathogens tion node on the graph. The edge had an associated time in the air, the question is where they should be placed to stamp representing the union of distinct time intervals be most effective, that is, where they would capture the the person, p, was at the location, l, during the day. If two 2 For more details on EPISIMS, see Eubank et al., 2004. people nodes, p1 and p2, had an incident edge in the same The 26 BRIDGE

Acknowledgements a This work was done in collaboration with the TRANSIMS and EPISIMS teams. EPISIMS was sup- ported by the National Infrastructure Simulation and Analysis (NISAC) Program at the U.S. Department of Homeland Security. The TRANSIMS project was funded by the U.S. Department of Trans- portation. The work of Zoltán Toroczkai was sup- ported by the U.S. Depart- ment of Energy. b References Albert, R., and A.-L. Barabási. 2002. Statistical mechanics of complex networks. Review of Modern Physics 74(1): 47–97. Barrett, C., R. Jacob, and M.V. Marathe. 2001. Formal language-constrained path problems. Journal on Com- puting 30(3): 809–837. Barrett C., K. Bisset, R. Jacob, G. Konjevod, and M.V. Marathe. 2002. An Exper- FIGURE 4 Comparison of baseline cases (on the left), with targeted ( ) vaccination and ( ) quarantining strategies (right). a. b. imental Analysis of a Rout- The bars represent the number infected at each location, and the light color represents the fraction of infected people who are ing Algorithm for Realistic infectious. The inserts show the cumulative number of people infected or dead as a function of time, and for the targeted response, Transportation Networks. the number vaccinated and quarantined. Note the different scales between the inserts. Source: LANL, 2005. Technical Report No. onset of the outbreak. Due to a particular so-called scale- LA-UR-02-2427. Proceed- free property (Albert and Barabási, 2002) of the loca- ings of the European Symposium on Algorithms, Rome, tions projection of the bipartite network, one can Italy. Los Alamos, N.M.: Los Alamos National Laboratory. pinpoint a small set of locations (the so-called dominat- Barrett, C., S. Eubank, S.V. Anil Kumar, and M. Marathe. ing set, about 10 percent of all locations) that would 2004. Understanding Large-Scale Social and Infrastructure cover about 90 percent of the population and would thus Networks: A Simulation Based Approach. SIAM News, be optimal locations for detectors. The same locations March 2004. could be used for distribution purposes (e.g., of prophy- Barrett, C., S. Eubank, and J. Smith. 2005. If smallpox strikes lactics and supplies). Figure 4 shows the evolution of Portland. Scientific American 292(3): 54–61. epidemics after a covert introduction in a particular loca- Barrett, C., S. Eubank, and M. Marathe. Forthcoming. tion (at a university) when the disease is left to spread Modeling and Simulation of Large Biological, Information (left side) compared to using a targeted-contact tracing and Socio-Technical Systems: An Interaction-Based and quarantining strategy (right). Approach. In Interactive Computation: The New Paradigm, WINTER 2005 27

edited by D. Goldin, S. Smolka, and P. Wegner. New York: Kaplan, E., D. Craft, and L. Wein. 2002. Emergency response Springer Verlag. to a smallpox attack: the case for mass vaccination. Pro- Eubank, S., H. Guclu, V.S.A. Kumar, M.V. Marathe, A. Srini- ceedings of the National Academy of Sciences 99(16): vasan, Z. Toroczkai, and N. Wang. 2004. Modelling dis- 10935–10940. ease outbreaks in realistic urban social networks. Nature LANL (Los Alamos National Laboratory). 2005. Controlling 429(6988): 180–184. Smallpox: Strategies in a Virtual City Built from Empirical Halloran, M., I.M. Longini Jr., A. Nizam, and Y. Yang. 2002. Data. Available online at: http://episims.lanl.gov. Containing bioterrorist smallpox. Science 298(5597): Watts, D., and S. Strogatz. 1998. Collective dynamics of 1428–1432. small-world networks. Nature 393(6684): 440–442. Fuel cells of the future will be based on solid electrolytes.

Fuel Cells: Current Status and Future Challenges

Stuart B. Adler

Fuel cells, which convert chemical energy directly to electricity, are more efficient than current means of energy conversion. The question is where they might fit in the broad spectrum of energy choices. This paper briefly reviews and compares polymer-electrolyte fuels cells (PEFCs) and solid- oxide fuel cells (SOFCs) and then describes significant scientific chal- lenges that must be overcome before these technologies can become Stuart B. Adler is an assistant pro- commercially competitive. fessor in the Department of Chem- Fuel cells are not a new idea. Sir William Grove first demonstrated the ical Engineering at the University conversion of hydrogen to electricity using an acid-electrolyte fuel cell in 1839. However, turning this idea into a practical means of energy conver- of Washington. sion has proved to be elusive. A major technical and cost barrier has been implementation of liquid electrolytes, the basis for most commercial fuel cells (e.g., alkaline fuel cells, molten-carbonate fuel cells). In contrast, the fuel cells of greatest commercial interest today are based on solid electrolytes, which have benefited from recent advances in materials and manufacturing. For the purposes of discussion, we can divide solid-electrolyte fuel cells into two types: (1) PEFCs, often referred to as proton-exchange-membrane (or PEM) fuel cells; and (2) SOFCs. Figure 1 illustrates how these types of fuel cells function. A common justification for fuel cells has been environmental protection— the idea that fuel cells produce only water as a combustion by-product and WINTER 2005 29

thus are “zero emission” devices. However, it is difficult entirely on clean combustion of hydrogen or other to make the case for fuel cells based on this argument multisource fuels that does not include fuel cells. alone. Although fuel cells themselves produce only To understand the potential role of fuel cells, we must water, the production of hydrogen from hydrocarbons, instead consider their primary advantage—efficiency. such as oil or coal, involves the production of carbon In this regard, fuel cells are an enabling (rather than a dioxide (CO2) and requires the suppression of sulfur displacing) technology. They recover energy that is dioxide (SO2). Thus fuel cells merely transfer the envi- normally lost by the irreversible process of combustion. ronmental problem elsewhere. Thus, fuel cells offer a potential path toward overall In addition, numerous technologies are already avail- reduction of fuel consumption that combustion simply able that can eliminate SO2 and nitrogen oxides (NOX) cannot provide, even after many years of incremental from combustion. Widespread implementation of these improvements. technologies is simply a matter of cost and political will. By reducing the overall amount of CO2 produced per Thus, one can easily imagine an energy economy based kilowatt (kW) of usable power, increased efficiency may, ultimately, have environmental benefits as well. In addition, the required retooling of the fuel infrastructure

toward more generic, small-molecule fuels (e.g., H2, CO, CH4) might also lead to centralization of CO2 pro- duction, which would facilitate carbon sequestration and reduce the vulnerability of particular energy sectors to fluctuations in the supply of particular fuel sources (e.g., the dependence of gasoline prices on the avail- ability of oil from the Middle East).

Comparisons between PEFCs and SOFCs A primary factor influencing the trade-off between capital and efficiency in fuel cells is operating tempera- ture. SOFC stacks, which operate at temperatures rang- ing from 550°C to 900°C, produce high-quality waste heat that can be captured for increased efficiency, com- bined heat and power, or reformation of hydrocarbons (HCs). SOFC stacks tend to operate adiabatically wherein excess air is used as the primary coolant, and thus heat can be recovered from the SOFC exhaust. This feature has made SOFCs very attractive for the production of stationary power, where efficiency is of high importance relative to capital cost, and operation on reformed HCs is an advantage. Allowable capital FIGURE 1 Two types of solid electrolyte fuel cells. a. In a PEFC, a proton- costs for stationary power ($400/kW) are about 10 times conducting polymer membrane is exposed on one side to fuel (hydrogen) and on higher than for PEFCs in automotive applications the other to air. On the hydrogen side (anode), H2 gas is oxidized, and the pro- (DOE, 2004b). tons thus created migrate to the other side of the membrane (cathode), where O 2 By using thin-film ceramics supported on low-cost gas in the air is reduced to water. Some portion of the reversible work of the metal alloys, SOFC developers have reduced material net reaction is recovered as a voltage difference between cathode and anode, and manufacturing costs, lowered operating tempera- delivered to an external circuit by the flow of electrons. PEFCs typically operate tures, and significantly mitigated cell-degradation prob- at 80~100°C. b. In an SOFC, a ceramic oxygen ion conductor at elevated tem- peratures (500~1,000°C) serves as the electrolyte membrane. In this case, the lems. Figure 2 shows an example of a metal-supported cell based on a thin ceria electrolyte, capable of stable fuel (which can be a mixture of H2, CO, and/or hydrocarbons) is oxidized to H2O 2– power densities of ~500 mW/cm2 at 570°C (Brandon, and CO2 at the anode, while O2 is reduced to O at the cathode. In both types of fuel cells, cells are normally assembled into multicell stacks to increase system 2005). Systems based on this type of cell are nearing voltage and provide a means of distributing gases (fuel and air) evenly. efficiency and cost targets for use in homes (combined The 30 BRIDGE

Substantial progress has been made in increasing the power density of PEFCs (>1 kW/kg) (Gasteiger et al., 2005), as well as reducing the amount of platinum (Pt) catalyst to a level that is reasonable to recycle (< 15g/vehicle, three to four times the catalyst in a cat- alytic converter) (Cooper, 2004; Gasteiger et al., 2005). Based on these successes, several of the world’s biggest automakers, including General Motors, Ford, Daimler, and Honda (Figure 3), have built demonstration cars. Despite these significant advances, solid-electrolyte fuel cells have not yet achieved widespread penetration into the energy market for many reasons. In particular, fuel cell systems are still too costly to be competitive with existing technology at current energy prices. FIGURE 2 Example of a metal-supported, thin-film solid-oxide fuel cell capable of oper- Although this situation may change as fuel prices rise ation below 600°C. Photo courtesy of Ceres Power, Ltd., reproduced with permission. and capital costs come down with manufacturing improvements and economies of scale, fundamental heat and power) and auxiliary power units for trucks technological barriers must also be overcome before cost and aircraft. reductions are likely. Many of these technological hur- In contrast, PEFCs have historically been designed dles have been described in detail elsewhere (DOE, to operate isothermally, at or below 80°C. Low operat- 2004a). The discussion below focuses on areas of fun- ing temperatures have made them more suitable for damental research where breakthroughs might lead to small or mobile applications, for which capital cost significant technological advancements. requirements are much more stringent, pure hydrogen Material Properties by Design (H2) is assumed to be available, and the efficiencies of heat integration are less important. The most chal- Many of the materials used in SOFCs and PEFCs lenging market from a capital-cost perspective is today are similar to the ones used 25 years ago. Examples motive power (cars), for which allowable capital costs are estimated to be on the order of $35/kW (Garman, 2003). PEFCs are also generally thought to match the size, weight, and start-up constraints for primary power in automobiles.

FIGURE 4 Relationship between proton conductivity and relative humidity in the FIGURE 3 Sandy Spallino, first individual customer to purchase a PEFC-powered adjoining gas at various temperatures for PFSA and phosphoric-acid-doped poly- car, fills up her Honda FCX at one of many H2 refueling stations planned for Cali- benzimidazole (PBI). Curves are also shown for materials that would enable and fornia. Source: Honda press release, June 2005. be ideal for system simplification. Source: Gasteiger and Mathias, 2003. WINTER 2005 31

include the nickel (Ni)-cermet anode used in most in the PFSA ionomer; at present they are distributed SOFCs and the perfluorosulfonic acid (PFSA) mem- randomly throughout the electrode matrix. However, brane used as the electrolyte in most PEFCs (Dupont this type of nanostructural analysis, let alone control, is Nafion®). Despite numerous difficulties with these not possible today. materials, they are still considered state of the art As shown in Figure 5, one possible technique on because their unique combination of properties is still the horizon for SOFCs is focused-ion beam milling cou- unmatched. However, they also introduce fundamental pled with electron microscopy or other surface problems (Figure 4). In SOFCs, Ni-cermet has very poor analytical techniques, which may make it possible to sulfur tolerance, especially below 800°C, which makes it analyze and direct electrode microstructures in new ways unsuitable as a long-term SOFC anode (DOE, 2004a). (J. Wilson et al., 2005). Researchers have also recently PEFC developers have concluded, that to be successful in demonstrated solution impregnation of materials into an cars, the system must operate at 110~120°C, which electrolyte host matrix to obtain SOFC electrodes with introduces severe performance and degradation prob- improved hydrocarbon activity or O2 reduction perfor- lems for PFSA (Gasteiger and Mathias, 2003). To date, mance (Huang et al., 2005; McIntosh and Gorte, 2004). a trial and error approach has been used to search for new materials. However, further advances are likely to Understanding Electrode Degradation and require a directed design approach (Hickner et al., 2004) Other Degradation Processes and/or combinatorial methods (Kilner et al., 2005). The vast majority of work in the last ten years has been focused on improving fuel cell performance. How- Probing and Controlling ever, as the technology has now reached some perfor- Microstructure/Nanostructure mance targets, and as more cells and stacks have been Despite the technological advances in SOFC and tested for longer periods of time, long-term durability PEFC technology in the last ten years, our understand- has risen to the top of the list of performance targets. ing and design capability are mostly at the macroscopic/ For example, SOFC electrodes can be very sensitive to empirical level. The microstructure of a PEFC elec- chromia (Cr) poisoning (Simner and Stevenson, 2004). trode, for example, is still understood only in a very gen- Although electrode degradation has been positively eral sense; exactly how the catalyst, ionomer, and gas linked to Cr contamination from metal interconnects, it come together and affect performance is generally not is not clear why some electrode materials are more sen- well understood and thus not amenable to intelligent sitive than others or why seemingly similar electrodes design. For example, one proposed strategy for improv- tested by different groups degrade at different rates. The ing the catalyst in PEFC cathodes is to concentrate Pt answers to these questions require a much deeper mech- particles near the opening of the aqueous flow channel anistic and scientific understanding of electrode processes than we currently possess. Recent advances in microfabrication and diagnostics may significantly improve our ability to control and analyze electrode reactions (Adler, 2004; J.R. Wilson et al., in press). Recent work using nonlinear electro- chemical impedance spectroscopy to resolve SOFC cathode reaction mechanisms may eventually improve our ability to diagnose how and why electrodes degrade and guide the selection of new materials and fabrica- tions to mitigate degradation.

Outlook Fuel cells continue to face major technological hur- dles that may require many years of research and devel- FIGURE 5 3D reconstruction of pores inside a porous Ni-YSZ SOFC anode, based opment before they can be overcome. In addition, fuel on 2D FIB-SEM image slices. At left are cross-sectional image slices corresponding cells are not likely to be implemented in isolation. to the reconstructed 3D image. They must be part of a larger shift in fuel infrastructure The 32 BRIDGE and efficiency standards, which will require sustained on Science, 108th Congress, Serial Number 108-4, March political and economic pressure—and time. Finally, 5, 2003. Available online at: http://commdocs.house.gov/ like any technology, economy of scale will require a committees/science/hsy85417.000/hsy85417_0f.htm. natural maturation process over many years or decades Gasteiger, H.A., and M.F. Mathias. 2003. Fundamental (DeCicco, 2001). Research and Development Challenges in Polymer Elec- Taken together, these hurdles suggest that the wide- trolyte Fuel Cell Technology. Paper presented at the Work- spread adoption of fuel cell technology is not likely in shop on High Temperature PEM Fuel Cells, Pennsylvania the short term. Successful advancement of fuel cell State University, December 2003. technology will require a sustained, long-term com- Gasteiger, H.A., S.S. Kocha, B. Sompalli, and F.T. Wagner. mitment to fundamental research, commercial devel- 2005. Activity benchmarks and requirements for Pt, Pt- opment, and incremental market entry. alloy, and non-Pt oxygen reduction catalysts for PEMFCs. Applied Catalysis B: Environmental 56(1–2): 9–35. References Hickner, M.A., H. Ghassemi, Y.S. Kim, B.R. Einsla, and J.E. Adler, S.B. 2004. Factors governing oxygen reduction in McGrath. 2004. Alternative polymer systems for proton solid oxide fuel cell cathodes. Chemical Reviews 104(10): exchange membranes (PEMs). Chemical Reviews 4791–4844. 104(10): 4637–4678. Brandon, N.P. 2005. Metal Supported Solid Oxide Fuel Cells Huang, Y.Y., J.M. Vohs, and R.J. Gorte. 2005. Characteriza- for Operation at 500–600°C. Paper presented at SSI-15, tion of LSM-YSZ composites prepared by impregnation the International Conference on Solid State Ionics, Baden- methods. Journal of the Electrochemical Society 152(7): Baden, Germany, July 17–22, 2005. A1347–A1353. Cooper, J.S. 2004. Recyclability of Fuel Cell Power Trains. Kilner, J.A., J.C.H. Rossiny, and S. Fearn. 2005. Combina- Pp. 792–801 in Proceedings of the SAE 2004 World Con- torial Searching for Novel Mixed Conductors. Paper pre- gress. SAE Technical Paper 2004-01-1136. Warrendale, sented at SSI-15 International Conference on Solid State Pa.: SAE International. Ionics, Baden-Baden, Germany, July 17–22, 2005. DeCicco, J. 2001. Fuel Cell Commercialization Perspectives: McIntosh, S., and R.J. Gorte. 2004. Direct hydrocarbon sol- Market Context and Competing Technologies. Paper pre- id oxide fuel cells. Chemical Reviews 104(10): 4845–4865. sented at the 2001 Gordon Conference on Fuel Cells, Bris- Simner, S.P., and J.W. Stevenson. 2004. Cathode-Chromia tol, R.I., July 29–August 3, 2001. Interactions. Paper presented as part of the SECA 2004 DOE (U.S. Department of Energy). 2004a. Fuel Cell Hand- Annual Meeting and Core Program Review, Boston, book, 7th ed. National Energy Technology Laboratory, Massachusetts 2004. Morgantown, West Virginia. Washington, D.C.: Office of Wilson, J., W. Kobsiriphat, R. Mendoza, J. Hiller, D. Miller, K. Fossil Energy, U.S. Department of Energy. Thornton, P. Voorhees, S. Adler, and S. Barnett. 2005. DOE. 2004b. Fuel Cell Program Annual Report. National Three Dimensional Reconstruction of Solid Oxide Fuel Energy Technology Laboratory, Morgantown, West Vir- Cell Electrodes. Unpublished paper. ginia. Washington, D.C.: Office of Fossil Energy, U.S. Wilson, J.R., D.T. Schwartz, and S.B. Adler. In press. Non- Department of Energy. linear electrochemical impedance spectroscopy for mixed- Garman, D. 2003. Testimony by David K. Garmen, Assistant conducting SOFC cathodes. Electrochemica Acta. Secretary, Energy Efficiency and Renewable Energy, U.S. Available online at: http://dx.doi.org/doi:10.1016/ Department of Commerce. Hearing before the Committee j.electacta.2005.02.109. The world will need access to 30TW of power without releasing carbon into the atmosphere.

Organic Semiconductors for Low-Cost Solar Cells

Chiatzun Goh and Michael D. McGehee

Currently the world consumes an average of 13 terawatts (TW) of power. By the year 2050, as the population increases and the standard of living in developing countries improves, this amount is likely to increase to 30 TW. Chiatzun Goh If this power is provided by burning fossil fuels, the concentration of carbon dioxide in the atmosphere will more than double, causing substantial global warming, along with many other undesirable consequences. Therefore, one of the most important challenges facing engineers is finding a way to provide the world with 30 TW of power without releasing carbon into the atmos- phere. Although it is possible that this could be done by using carbon seques- tration along with fossil fuels or by greatly expanding nuclear power plants, it is clearly desirable that we develop renewable sources of energy. The sun deposits 120,000 TW of radiation on the surface of the earth, so there is clearly enough power available if an efficient means of harvesting solar energy can be developed. Only a very small fraction of power today is generated by solar cells, which Michael D. McGehee convert solar energy into electricity, because they are too expensive (Lewis and Crabtree, 2005). More than 95 percent of the solar cells in use today are made of crystalline silicon (c-Si). The efficiency of the most common panels

Chiatzun Goh is a graduate student in the Department of Materials Science and Engineering at Stanford University. Michael McGehee is an assistant professor of materials science and engineering at Stanford University. The 34 BRIDGE

is approximately 10 percent, and the cost is $350/m2. In Organic Semiconductors other words, the cost of the panels is $3.50/W of elec- Because organic semiconductors have different bond- tricity produced in peak sunlight. When you add in the ing systems from conventional, inorganic semiconduc- cost of installation, panel support, wiring, and DC to AC tors, they operate in a fundamentally different way. converters, the price rises to approximately $6/W. Over Conventional semiconductors are held together by the lifetime of a panel (approximately 30 years), the strong covalent bonds that extend three-dimensionally, average cost of the electricity generated is $0.3/kW-hr. resulting in electronic bands that give rise to its semi- By comparison, in most parts of the United States, elec- conducting properties. Organic materials have similar tricity costs about $0.06/kW-hr. Thus, it costs approxi- intramolecular covalent bonds but are held together mately five times as much for electricity from solar cells. only by weak intermolecular van der Waals interactions. If the cost of producing solar cells could be reduced by a The electronic wave function is thus strongly localized factor of 10, solar energy would be not only environ- to individual molecules, and the weak intermolecular mentally favorable, but also economically favorable. interactions instigate a narrow electronic bandwidth Although c-Si solar cells will naturally become formed in molecular solids. cheaper as economies of scale are realized, dicing and polishing wafers will always be somewhat expensive. Bonding Thus, it is desirable that we find a cheaper way to make The semiconducting nature of organic semiconductors solar cells. The ideal method of manufacturing would arises from the π electron bonds that exist when mole- be depositing patterned electrodes and semiconductors cules are fully conjugated (i.e., have alternating single on rolls of plastic or metal in roll-to-roll coating and double bonds). The weakly held π electrons are machines, similar to those used to make photographic responsible for all interesting optical and electronic tran- film or newspapers. Solar cells made this way would not sitions in organic semiconductors. The π to π* transi- only be cheaper, but could also be directly incorporated tions in organic semiconductors are typically in the range into roofing materials, thus reducing installation costs. of 1.4–2.5 eV, which overlaps well with the solar spec- Organic semiconductors that can be dissolved in com- trum and makes them very promising candidates as active mon solvents and sprayed or printed onto substrates are light absorbers in solar cells. A few examples of organic very promising candidates for this application. semiconductors used in solar cells are shown in Figure 1.

Excitons The main difference between organic semicon- ductors and inorganic semi- conductors as photovoltaic materials is that optical excitations of organic semi- conductors create bound electron-hole pairs (called excitons) that are not effec- tively split by the electric field (Gregg, 2003). To separate the bound elec- trons and holes, there must be a driving force to over- come the exciton-binding energy, typically 0.1– 0.4 eV. Excitons in organic semiconductors that are not split eventually recom- FIGURE 1 The chemical structures of four different organic semiconductors used in organic solar cells. bine either radiatively or WINTER 2005 35

nonradiatively, thereby reducing the quantum effici- dangling bonds at surfaces. Therefore, organic-organic ency of a solar cell. junctions or organic-metal junctions in organic solar In inorganic semiconductors the attraction between cells (interface states) do not act as potential charge- an electron-hole pair is less than the thermal energy kT. carrier recombination sites. Therefore, no additional driving force is required to gen- erate separated carriers. Research has shown that exci- 18 tons in organic semiconductors can be efficiently split at 5x10 100 Integrated Spectral Photon Flux (%) )

a heterojunction of two materials with dissimilar elec- -1 4x1018 80 nm

tron affinities or ionization potentials. -2 m -1 Bandwidth 3x1018 60

The narrow electronic bandwidth in organic 18 semiconductors has a few consequences. First, the 2x10 40 absorption-spectrum bandwidth is narrower than 18 in conventional inorganic semiconductors. Conse- 1x10 20 quently, a single organic material can be potentially Spectral Photon Flux (s 0 0 photoactive only in a narrow optical-wavelength range 400 600 800 1000 1200 of the solar spectrum (Figure 2). Although this is a dis- Wavelength (nm) advantage in terms of harvesting solar flux, multiple absorbers in stacks of solar cells connected in series FIGURE 2 Overlap of the absorption spectrum of P3HT (dashed line) with the AM can be engineered to expand the absorption range. 1.5 solar spectrum (solid line). Integrated spectral-photon flux below each photon Because the valence band and conduction band are wavelength is also shown (dotted line). concentrated in narrower energy regions, the absorp- tion coefficient resulting from the excitation of elec- Production of Heterojunction Devices trons from the valence band to the conduction band is The simplest organic solar cells can be made by sand- very strong, typically >10–5 cm–1 at peak absorption. wiching thin films of organic semiconductors between This high absorption coefficient means that only a thin two electrodes with different work functions. The work (100–200 nm) film is required to absorb most incident function is the amount of energy necessary to pull an light, an attractive characteristic for solar cells because electron from a material. When such a diode is made, less material is required to make them. electrons from the low-work-function metal flow to Second, the charge carriers in organic semiconduc- the high-work-function metal until the Fermi levels tors do not exhibit band-like transport as they do in are equalized throughout the structure. This sets up a inorganic semiconductors. Instead, they move around built-in electric field in the semiconductor. When the by a hopping mechanism between localized states. The organic semiconductor absorbs light, electrons are charge-carrier mobilites in organic semiconductors created in the conduction band, and holes (positive- are, therefore, inherently low, with typical values of charge carriers) are created in the valence band. Thus, <10–2 cm2/Vs. This low charge-carrier mobility puts a in principle, the built-in electric field can pull the constraint on the thickness of organic materials that photogenerated electrons to the low-work-function can be used in a solar cell because recombinative loss electrode and holes to the high-work-function elec- increases with increasing thickness. Fortunately, this trode, thereby generating a current and voltage drawback is offset because only a very thin layer of (Figure 3a). In practice, however, these cells have very organic materials is necessary because organic semicon- low power-conversion efficiency (< 0.1 percent) ductors are highly absorptive. Organic solar cells may because the electric field is not strong enough to sepa- potentially perform better than conventional solar cells at rate the bound excitons (i.e., the excited-state species higher temperature, because hopping is a thermally acti- formed in organic semiconductors described above). vated process. The performance of inorganic solar cells A significant improvement in the performance of typically decreases as operating temperature increases. organic solar cells was achieved by Tang (1986). His A third key difference between organic and inorganic device consisted of a heterojunction between donor and semiconductors is that organic materials do not have acceptor semiconductors, resembling a p-n junction in The 36 BRIDGE conventional solar cells (Figure 3b). The benefit of this before the electron and hole recombine if carbon-60 device derived from the use of two organic materials (C60) derivatives are blended into the polymer with offset electron affinities (lowest unoccupied mole- (Figure 3c). Blend solar cells were made simply by cular orbital, LUMO) or ionization potentials (highest blending the C60 derivative, which acts as an electron occupied molecular orbital, HOMO). Excitons that dif- acceptor, into the polymer at concentrations in the fuse to the interface undergo efficient charge transfer, as range of 18–80 wt. percent. At these concentrations, this offset in the energy levels provides a sufficient chem- the polymer and the C60 derivatives form a connected ical potential energy to overcome the intrinsic exciton- network to each electrode. The key to making effi- binding energy. Upon charge transfer, the electrons are cient blend solar cells is to ensure that the two materi- transported in the acceptor material and the holes in the als are intermixed very closely at a length scale less donor material to their respective electrodes. than the exciton diffusion length so that every exciton

The efficiency of this type of planar heterojunction formed in the polymer can reach an interface with C60 device is limited, however, by the exciton diffusion to undergo charge transfer. length, which is the distance over which excitons At the same time, the film morphology has to enable travel before undergoing recombination, approximately charge-carrier transport in the two different phases to 5–10 nm in most organic semiconductors. Excitons minimize recombination. The film morphology (i.e., formed at a location further than 5–10 nm from the phase separation between the two materials) and, ulti- heterojunction cannot be harvested. The active area of mately, the efficiency of the device, are determined by this type of solar cell is thus limited to a very thin the concentration of materials, film-casting solvent, region close to the interface, which is not enough to annealing time, temperature, and other parameters. adsorb most of the solar radiation flux. Solar cells made by this method have continuously improved to better than 2 percent power efficiency under Blend Cells solar AM 1.5 conditions over the last few years (Padinger In the mid 1990s, Yu and colleagues (1995) showed et al., 2003; Shaheen et al., 2001), and recently, an that excitons can be rapidly split by electron transfer efficiency of 5 percent was reported (Ma et al., 2005).

FIGURE 3 A schematic energy-band diagram of (a) a single-layer semiconductor polymer solar cell, in which indium-tin oxide (ITO) serves as a transparent high-work-function electrode and calcium serves as the low-work-function electrode; (b) photoinduced electron transfer from the lowest unoccupied molecular orbital (LUMO) of an electron donor to the LUMO of an electron acceptor in a planar heterojunction cell; (c) a bulk heterojunction solar cell based on semiconducting polymer and C60 derivative. WINTER 2005 37

The work on polymer/C60-derivative blend cells has between the two semiconductors; (3) forward electron created a new paradigm in the field of organic-based solar transfer; and (4) charge transport. One must also con- cells, which is the notion of bulk heterojunction devices, sider undesirable recombination processes that can wherein two semiconductors with offset energy levels are limit the performance of the cell, such as geminate interpenetrated at a very small length scale to create a recombination of electrons and holes in the polymer high interface area for achieving high-efficiency devices. and back electron transfer from the electron acceptor to Since then, similar bulk heterojunction devices using the polymer (Figure 4b). electron acceptors other than the C60 derivative, such as CdSe nanorods (Huynh et al., 2002), a second Light Absorption semiconducting polymer (Granstrom et al., 1998), and The necessity of absorbing most of the solar spectrum titania nanocrystals (Arango et al., 1999), have been (process 1) creates two requirements. First, the band demonstrated, albeit with slightly lower efficiencies. gap must be small enough to enable the polymer to absorb most of the light in the solar spectrum. Calcula- Limits on Performance tions to determine the band gap that optimizes the To understand the limits on the performance of bulk amount of light that can be absorbed and the voltage heterojunction devices and find ways to improve them, that can be generated show that the ideal band gap is it is important to consider all of the processes that must approximately 1.5 eV, depending on the combination of occur inside the cells for electricity to be generated. semiconducting polymers and electron acceptors These processes, shown in Figure 4a, are: (1) light (Coakley and McGehee, 2004). Second, the film must absorption; (2) exciton transport to the interface be thick enough to absorb most of the light. For most organic semiconductors, this means that films must be 150–300 nm thick, depending on how much of the film a Desirable processes consists of a nonabsorbing electron acceptor. The opti- mum film thickness will absorb much incident light (2) (3) without significant recombination losses. e- (4) Exciton Transport (1) Once an exciton is created in the polymer, it must dif- fuse or travel by resonance energy transfer (process 2) to the interface with the other semiconductor and be split (4) h+ by electron transfer before it recombines (process 5). Experiments have shown that an exciton can diffuse electrode semiconducting electron acceptor electrode approximately 5–10 nm in most semiconducting poly- polymer (e.g., C60 or TiO2) mers before recombination. Therefore, no regions in the polymer can be more than 5–10 nm from an inter- b Undesirable recombination events face. Templating or nanostructuring of the donor and e- acceptor phases to fabricate ordered bulk heterojunction e- with controlled dimensions is an attractive approach to achieving full exciton harvesting (Figure 5) (Coakley (5) (6) and McGehee, 2003). Some small-molecule semicon- ductors have been shown to have larger exciton diffu- h+ h+ sion lengths (Peumans et al., 2003). Research is under way to improve exciton transport in organic semiconductors, for example by using resonance electrode semiconducting electron acceptor electrode

polymer (e.g., C60 or TiO2) energy transfer to funnel excitons directly to an absorber located at the charge-splitting interface or by incorpo- FIGURE 4 Schematic energy diagrams of the semiconductors and energy levels in rating phosphorescent semiconductors, which exhibit a bulk heterojunction solar cell showing the (a) desirable and (b) undesirable longer excited-state lifetimes (Liu et al., 2005; Shao and processes that can occur. Yang, 2005). The 38 BRIDGE

that are only 100-nm thick so that the carriers do not have to travel very far. Unfortunately, most of the light is not absorbed by films this thin. If the films are thick enough to absorb most of the light, then only a small fraction of the carriers escape the device. Many researchers are now trying to optimize the interface between the two semiconductors and improve charge transport in the films so that the charge can be extracted from 300-nm-thick films before recombination occurs.

Future Challenges The outlook for organic solar cells is very bright. FIGURE 5 Scanning electron micrograph of mesoporous titania film. This film was Efficiency greater than 5 percent has been achieved used in combination with P3HT to fabricate ordered bulk heterojunction solar cells. (Ma et al., 2005; Xue et al., 2004), and many are Source: Coakley and McGehee, 2003. Copyright 2003, American Institute of Physics. optimistic that 20 percent can be achieved by opti- mizing the processes described above. Once this goal Forward Electron Transfer is achieved, a primary research challenge will be The actual process of charge transfer (process 3) making cells that are stable in sunlight and that can requires that the offset in LUMO levels of the donor and handle wide temperature swings. The survival of many acceptor be sufficient to overcome the exciton-binding organic pigments in car paints in sunlight and the pro- energy. However, this drop in energy must not be exces- duction of organic light-emitting diodes with opera- sive, because the maximum voltage attainable from this tional lifetimes greater than 50,000 hours are type of bulk heterojunction solar cell is determined by encouraging signs that the required stability can be the gap between the HOMO of the electron donor and achieved. The final challenge will be scaling up the the LUMO of the acceptor. The gap becomes smaller as process and manufacturing the cells at a cost of approx- the LUMO of electron acceptors is moved farther away imately $30/m2. Shaheen and colleagues (2005) have from the LUMO of the polymer, which corresponds to a described several approaches to making organic cells. larger driving force for charge transfer. As can be seen from processes 1, 2, and 3, the design References of an efficient organic solar cell involves optimizing the Arango, A.C., S.A. Carter, and P.J. Brock. 1999. Charge various energy levels to achieve the optimum level of transfer in photovoltaics consisting of polymer and TiO2 extracted current with respectable voltage, as the power nanoparticles. Applied Physics Letters 74(12): 1698–1700. supplied by a solar cell is the product of current and volt- Coakley, K.M., and M.D. McGehee. 2003. Photovoltaic cells age. Fortunately, the wealth of chemical synthetic made from conjugated polymers infiltrated into mesoporous knowledge and the dependence of electronic properties titania. Applied Physics Letters 83(16): 3380–3382. of organic molecules on their molecular structures allow Coakley, K.M., and M.D. McGehee. 2004. Conjugated poly- for flexible tuning of the band gap and energy levels of mer photovoltaic cells. Chemistry of Materials 16(23): organic semiconductors by chemical synthesis. Signifi- 4533–4542. cant research on band engineering of this type should Granstrom, M., K. Petritsch, A. Arias, A. Lux, M. Andersson, yield very promising results in the near future. and R. Friend. 1998. Laminated fabrication of polymeric photovoltaic diodes. Nature 395(6699): 257–260. Charge Transport Gregg, B. 2003. Excitonic solar cells. Journal of Physical After forward electron transfer, the holes in the poly- Chemistry B 107(20): 4688–4698. mer and the electrons in the electron acceptor must Huynh, W.U., J.J. Dittmer, and A.P. Alivisatos. 2002. Hybrid reach the electrodes (process 4) before the electrons in nanorod-polymer solar cells. Science 295(5564): the acceptor undergo back electron transfer to the poly- 2425–2427. mer (process 6). Even in the best bulk heterojunction Lewis, N.S., and G. Crabtree. 2005. Basic research needs cells, this competition limits the efficiency of the cells. for solar energy utilization. Available online at: http:// The problem can usually be mitigated by making cells www.sc.doe.gov/bes/reports/files/SEU_rpt.pdf. WINTER 2005 39

Liu, Y., M.A. Summers, C. Edder, J.M.J. Fréchet, and M.D. Shaheen, S.E., D.S. Ginley, and G.E. Jabbour, editors. 2005. McGehee. 2005. Using resonance energy transfer to Organic Photovoltaics. MRS Bulletin 30(1). Special Edi- improve exciton harvesting in organic-inorganic hybrid tion. photovoltaic cells. Advanced Materials (in press). Shao, Y., and Y. Yang. 2005. Efficient organic heterojunction Ma, W., C. Yang, X. Gong, K. Lee, and A.J. Heeger. 2005. photovoltaic cells based on triplet materials. Advanced Thermally stable, efficient polymer solar cells with Materials (in press). nanoscale control of the interpenetrating network mor- Tang, C.W. 1986. Two-layer organic photovoltaic cell. phology. Advanced Functional Materials 15(10): Applied Physics Letters 48(2): 183–185. 1617–1622. Xue, J.G., S. Uchida, B.P. Rand, and S.R. Forrest. 2004. Padinger, F., R.S. Rittberger, and N.S. Sariciftci. 2003. Effects Asymmetric tandem organic photovoltaic cells with hybrid of postproduction treatment on plastic solar cells. planar-mixed molecular heterojunctions. Applied Physics Advanced Functional Materials 13(1): 85–88. Letters 85(23): 5757–5759. Peumans, P., A. Yakimov, and S.R. Forrest. 2003. Small mol- Yu, G., J. Gao, J.C. Hummelen, F. Wudl, and A.J. Heeger. ecular weight organic thin-film photodetectors and solar 1995. Polymer photovoltaic cells: enhanced efficiencies cells. Journal of Applied Physics 93(7): 3693–3723. via a network of internal donor-acceptor heterojunctions. Shaheen, S.E., C.J. Brabec, N.S. Sariciftci, F. Padinger, T. Science 270(5243): 1789–1791. Fromherz, and J.C. Hummelen. 2001. 2.5% efficient organic plastic solar cells. Applied Physics Letters 78(6): 841–843. The 40 BRIDGE

NAE News and Notes

NAE Newsmakers

Robert B. Fridley, Professor of the tree harvester for mechanical pioneering contributions to the devel- Emeritus and former department harvesting of tree fruit. He is also opment and understanding of CMOS chair of Biological and Agricultural the author of several books on the gate dielectrics, the basis of silicon Engineering at University of Cali- agricultural industry. chips. The award is given in recogni- fornia, Davis, received an Award of Edward D. Lazowska, Bill and tion of outstanding contributions to Distinction from the College of Melinda Gates Chair in Computer solid-state devices and technology. Agricultural and Environmental Science and Engineering at the George M. Whitesides, Wood- Sciences, University of California, University of Washington, received ford L. and Ann A. Flowers Univer- Davis. The Award of Distinction is the Computing Research Associa- sity Professor, Department of given to individuals whose contri- tion 2005 Distinguished Service Chemistry and Chemical Biology, butions and achievements have Award for outstanding service to Harvard University, was awarded enriched the image and reputation the computing research community. the Welch Foundation 2005 Welch of the college and improved its pub- Tso-Ping Ma, Raymond John Award in Chemistry. Dr. White- lic service capabilities. Dr. Fridley’s Wean Professor, Department of Elec- sides received the award for his con- research, which was focused on the trical Engineering, Yale University, tributions in many areas of needs of agriculture, forestry, and is the recipient of the 2005 IEEE chemistry and his leadership in the aquaculture, led to the development Andrew S. Grove Award for his scientific community.

2005 Annual Meeting

Washington, D.C., for the 2005 NAE followed President Wulf’s address. Annual Meeting. An orientation The program continued with the session for new members on Saturday, presentation of the 2005 Founders October 8, was followed by the NAE Award to C. Dan Mote Jr. and the Council dinner in the Great Hall of Arthur M. Bueche Award to Leo the National Academies Building in Young. Dr. Mote, president of the honor of the 74 new members and University of Maryland and Glenn 10 new foreign associates. L. Martin Institute Professor of Engi- NAE Chair Craig R. Barrett neering, was recognized “for the cre- opened the public session on Sun- ation of a comprehensive body of day, October 9, with an appeal to work on the dynamics of moving new members to advocate for more flexible structures and for leadership research, better K–12 education, in academia.” Dr. Young, director of and a more innovation-friendly research (retired) for the U.S. environment (p. 42). President Department of Defense, received the Stephen Intille, research scientist and technology direc- Wm. A. Wulf then addressed the Bueche Award for “sponsoring col- tor, House_n Consortium, Massachusetts Institute of group and entreated support for laborative research programs among Technology, delivered the Lillian M. Gilbreth Lecture. improving the competitiveness and academic, industrial, and govern- security of the United States in a ment engineers and scientists.” On October 8, NAE members, for- globalized world (p. 44). The induc- Following short presentations by eign associates, and guests gathered in tion of the NAE Class of 2005 Drs. Mote (p. 48) and Young (p. 50), WINTER 2005 41

electrical and computer engineer- NAE activities and issues relevant ing; Leah Jamieson, Ransburg Pro- to the engineering profession. The fessor of Electrical and Computer business session was followed by a Engineering, associate dean, and symposium, “Adapting Engineering director, EPICS; and William C. Education to the New Century,” Oakes, assistant professor of engi- moderated by G. Wayne Clough, neering and co-director of EPICS, president of Georgia Institute of all of Purdue University. They Technology. Speakers included received their award at a ceremony Charles M. Vest, President Emeri- in Washington, D.C., last February. tus of MIT; George Peterson, execu- After a break, Dr. Stephen Intille tive director of ABET Inc.; presented the Lillian M. Gilbreth Theodore C. Kennedy, founder of Lecture, which recognizes outstand- BE&K Inc.; Jacquelyn F. Sullivan, ing young engineers. Dr. Intille, codirector of the Integrated Teach- research scientist and technology ing and Learning Laboratory at the Bernard Amadei, professor of civil engineering at the director of the Massachusetts Insti- University of Colorado; and Susan University of Colorado, described the activities of Engi- tute of Technology (MIT) Home of Ambrose, associate provost for edu- neers Without Borders. the Future Project, spoke on using cation and director of the Eberly ubiquitous computing technologies Center for Teaching Excellence at Edward Coyle, a recipient of NAE to encourage aging in place. Carnegie Mellon University. In the 2005 Gordon Prize, spoke on behalf The final speaker was Bernard afternoon, members and foreign of the prizewinning program, the Amadei, professor of civil engineer- associates participated in NAE sec- Engineering Projects in Community ing at the University of Colorado, tion meetings at the Keck Center. Service (EPICS) Program at Purdue who spoke in his capacity as founder The final event that evening was University (p. 52). The Gordon of Engineers Without Borders. The the annual reception and dinner Prize is awarded in recognition of day ended with a reception for dance, held at the Andrew W. new modalities and experiments in members and their guests. Mellon Auditorium. Music was pro- education that develop effective At the Annual Business Session vided by the Radio King Orchestra. engineering leaders. The recipients on Monday, October 10, members The next annual meeting is in 2005 are Dr. Coyle, professor of had an opportunity to discuss specific scheduled for October 15–16, 2006. The 42 BRIDGE

Chairman’s Remarks

report focuses on this country, I think last two decades. But go a step fur- it has equal application to just about ther and consider the achievement every other established economy in of the average 12th grade student in the world today. The report talks math and science, which is of special about competition—economic com- interest to us here. Twelfth graders petition—going forward. And if you in the United States rank in the bot- translate that, very simply, economic tom 10 percent among their inter- progress and economic strength national peers. translate to standard of living, which I urge you to support the recom- is something we are all interested in. mendations in the report on K–12 Craig R. Barrett The report lists three general areas of education, which, I think, is first, significance for the strength of any second, and third on the list of pri- These remarks were delivered on country or any economy: (1) educa- orities. The ability of any country October 9, 2005, at the NAE tion, (2) the creation of ideas, and to educate its young people, espe- Annual Meeting. (3) an environment where educated cially in math and science, is Some of you may have heard about people can take ideas and create absolutely key to its success. All of the DARPA grand challenge— new products, new services, and us must raise our voices, collective- autonomous, computing vehicles new businesses that drive eco- ly, to local officials, state officials, attempting to navigate a 130-mile nomic growth and a better standard and national officials. I don’t see course through the desert near Las of living. how our economy can continue to Vegas. These specially equipped The report discusses in detail each thrive with a 30-percent dropout vehicles are totally computer driven, of those areas and focuses on actions rate and an inability to educate with no onboard human interven- we must take for the United States young people to be competent in tion. Last year, the Carnegie Mellon to remain competitive. I will talk math and science. vehicle, which led the race, went a very briefly about some of those The second issue is the genera- total of 7.5 miles but eventually ran actions. I hope this will encourage tion of ideas. Ideas are typically off the road. No vehicles finished the new members of the Academy to generated through research and course last year, but even 7.5 miles join with existing members to make development (R&D) efforts and was a great accomplishment. This some of these things happen. expenditures. The good news is that year, five vehicles finished the entire The first issue is education. In the United States still has the pre- 130-mile course, which is quite an recent rankings of universities, the mier basic research operation in the engineering achievement. It also United States has 18 of the top 20 world—our research universities, shows the tremendous improvement and 35 of the top 50 universities in which are funded primarily by the and accomplishment in just one year. the world in a semi-quantitative federal government, but also by Of course, I am pleased to report that ranking system. The free competi- local governments and private the vehicle from Stanford (my alma tion of our university system contin- industry. But funding in physical mater) finished first. ues to serve us well, even though sciences comes primarily from the I noted with interest that, coinc- there is increased competition from National Science Foundation, and dent with the induction of the NAE some prestigious foreign universities. some from the U.S. Department class of 2005, the National Aca- The real issue in the United of Defense and U.S. Department of demies has released a new report States, by any objective measure, is Energy. For about the last two (Rising Above the Gathering Storm: K–12 education. Consider the decades, funding for basic research Energizing and Employing America for dropout rate of young children, in the physical sciences in the a Brighter Economic Future) on com- roughly 30 percent, who do not even United States has been flat in petitive challenges the United States graduate from high school. This per- absolute dollars, roughly $5 billion faces going forward. Although the centage has remained steady for the spent in our major universities. WINTER 2005 43

To put that into perspective, Intel broadband conductivity in the 1800s and was writing poetry about Corporation alone spends $5.5 bil- United States. The United States the time the National Academies lion, not only on basic research, but now ranks about 15th per capita in was created. The title of the poem, on total research. And Intel is not terms of broadband penetration. If appropriate for the National Aca- alone. Companies like Microsoft you assume that the Internet is the demy of Engineering, is “The Bridge and IBM spend equal amounts. The vehicle for information access, Builder.” fact that one company can spend as communication, and decision mak- An old man, going a lone highway, much on R&D as the entire federal ing going forward, then you realize Came, at the evening, cold and gray, government of the United States is we are at a disadvantage. It is To a chasm, vast, and deep, and wide, troubling. The new Academies unfortunate that every small busi- Through which was flowing a sullen tide. report recommends a substantial ness, every home, and every large increase in funding for basic business in the United States is not The old man crossed in the twilight dim; research to keep our universities at yet connected with broadband The sullen stream had no fears for him; the top. I encourage you all to sup- capability. I encourage you to sup- But he turned, when safe on the other side, port that recommendation. port the recommendations in the And built a bridge to span the tide. Ironically, the United States report related to increasing our “Old man,” said a fellow pilgrim, near, spends roughly $25 billion a year on broadband capability. “You are wasting strength with building here; agricultural subsidies, five times the I want to leave you with this Your journey will end with the ending day; amount we spend on basic R&D in thought: The Irish poet, William You never again must pass this way; physical sciences. Is it better to fund Butler Yeats, talked about education You have crossed the chasm, deep and wide. the industries of the nineteenth in a simple, yet profound, way. He Why build you a bridge at the eventide?” century than the industries of the said education is not like filling a The builder lifted his old gray head: twenty-first century? If you talk to a pail but like lighting a fire. The “Good friend, in the path I have come,” he said, congressman, ask him why we spend challenge for our society is to light “There followeth after me today, $25 billion to support traditional that fire in every young child in the A youth, whose feet must pass this way. industries like agriculture at the United States and give every child expense of basic R&D. an opportunity to grow to his or her This chasm, that has been naught to me, The third area is creating and full potential. To that fair-haired youth may a pitfall be. fostering an environment for inno- As an engineer, I’ve never before He, too, must cross in the twilight dim; vation. A number of topics are quoted two poets in the same Good friend, I am building the bridge for him.” included in this area, from patent speech, but this poem is appropriate I think our task at the National systems to communication laws, here. The poet is Will Allen Drom- Academies is to follow the advice in rules, and regulations that have goole. Actually, Will was a she—a that poem, to build a bridge for the inhibited the proliferation of Tennessean who lived in the late next generation. The 44 BRIDGE

President’s Address

full list in A Century of Innovation: prosper. I expect you have either Twenty Engineering Achievements read or heard about Tom Friedman’s That Transformed Our Lives. But book, The World Is Flat: A Brief consider this. History of the Twenty-first Century In 1900, almost no home had (Farrar, Straus and Giroux, 2005). electricity; very few people had auto- Friedman’s premise is that the (eco- mobiles, and there were just a few nomic) playing field has become tens of miles of paved road in the more level—flatter in his par- entire country; the first airplane had lance—and off-shoring, outsourc- not been flown; the average life ing, and Lou Dobbs’ “Exporting of Wm. A. Wulf expectancy was 46 (it’s now 77, and America” are all manifestations of 20 of the additional 31 years are this. Flattening, Friedman argues, These remarks were delivered on attributable to clean water and san- has happened because it is now October 9, 2005, at the NAE itation); almost no one had a tele- technologically possible to locate call Annual Meeting. phone; there was no refrigeration, centers in the Philippines, coordi- It is an immense honor to wel- radio, TV, or antibiotics; and, of nate the complex supply chains and come once again our new members course, there were no computers, work flows that enable manufactur- and foreign associates. The knowl- Internet, or satellites. In 1900, ing in China, and perform “back edge and experience you bring to 50 percent of the U.S. population office” work in India, including the academy enables us to continue lived on farms, and it took that having Indian radiologists read to play a unique role in service to many to feed the other 50 percent. x-rays and CAT scans taken in our democracy. The academy is Today, 2 percent live on farms and U.S. hospitals. renewed and enriched by each new feed not only the other 98 percent, Friedman is not the first one to class. I also want to acknowledge but also a good number of folks over- say these things, nor is his analysis the families and friends of the new seas. And the list goes on . . . impeccable. But he has captured members, whose support has been The phenomenal transformation the attention of the country. He essential to every inductee. of our quality of life has been fueled lists ten “flatteners” that have led to My topic this year is what I by innovations created by engineers, the leveling of the economic playing believe the nation must do to pros- and the pace of innovation, if any- field—nine of which are technolo- per in the 21st century. By any thing, is accelerating. Many of us gies we (engineers) created. Engi- objective measure, the United have believed that the United neers made possible, and now real, States is in great shape! We are the States has been a particularly fertile what many believe is a serious prob- only superpower, our economy is the place for innovation. Our great lem for the United States—compe- largest in the world and growing research universities have created tition on a rough-and-tumble, flat nicely, both unemployment and streams of new knowledge and pro- playing field. inflation are low, and we are domi- vided educated engineers to exploit Friedman argues that, despite the nant in most areas of science, tech- that knowledge. We have had a dangers, the trend toward a flat nology, and high-tech industry. ready supply of capital, and our cul- world is a good thing, both eco- Engineers have been responsible ture encourages risk taking. nomically and geopolitically. Lower for much of that dominance and the costs benefit consumers and share- resulting prosperity. There is no bet- The Flat, Global Playing Field holders in developed and develop- ter way to convey the impact of The world is not static, however, ing countries alike, and a rising engineers on our quality of life than and recent books and reports have middle class in India and China will by reading a list of the 20 greatest begun to raise warning flags that become consumers of their own engineering achievements of the the status quo in the United States products, and ours. That same ris- twentieth century. You can find the will not suffice for us to continue to ing middle class has a growing stake WINTER 2005 45

in frictionless international com- leadership position are no longer to 2004, more than one-third of merce—and hence in stability, sufficient—and perhaps not even Nobel Prizes in the United States peace, and the rule of law. But, he appropriate—in a globalized world. were awarded to foreign-born scien- says, there will be problems during Most Americans instinctively tists. To me those facts suggest that the transition, and whether global know that the way to prosper is to we have been skimming the best and flatness will be good for any particu- innovate. But innovation means brightest from around the world and lar country will depend on whether change, and change can be difficult, that much of our prosperity is the that country is prepared to compete especially when you are on top. As result of our access to that incredible on the new global playing field. Charles Darwin said, “It is not the talent pool. A few lines at the very end of strongest of the species that survives, At a congressional hearing, only Chapter 6 inspired me to choose this nor the most intelligent, but rather the congressmen get to ask ques- topic. “But have we [the United the one most adaptable to change.” tions, of course, so I don’t know their States] really been investing in our Clayton Christenson, the author opinions. But reading between the future and preparing our children of The Innovator’s Dilemma (Collins, lines, it seemed to me that many of the way we need to for the race 2003), noted that the best run com- the subcommittee members had a ahead? . . .The answer is no.” When panies are often the most resistant to different take on these facts— I combine those lines with the fact new, “disruptive” technologies. It is namely that every foreign student is that engineers are the ones who cre- no accident that only one of the 100 a potential spy trying to steal our ated the enablers of a flat world, I largest U.S. firms in 1900 was still technology and that the United can’t help but think that we have a on the list in 2000. Ironically, the States would be better off if there responsibility to prepare the United most dangerous place to be seems to were no foreign students at Ameri- States to play on a flat field. be at the top—for both companies can universities. One congressman NAE recently released Engineering and countries. It is hard for the said explicitly that if there were no Research and America’s Future, a leaders to change because what they foreign students, there would be study chaired by Jim Duderstadt, have been doing is what got them to room for all of the U.S. students former president of the University of the top and because a large vocal who want to be scientists and engi- Michigan. The report documents cadre believes that deviating from neers but can’t get into college. the decline of (1) federal support the current course will lead to disas- I was stunned. Like the corporate for research in the physical sciences ter. Therefore, the tendency is to types who cling to the notion that and engineering; (2) the number of circle the wagons—to protect the they are doing the right thing, these U.S. students in physical sciences current advantage! representatives obviously believe and engineering (the United States A recent experience will illustrate that the United States is the sole actually produces only 7 percent of this kind of thinking. A few weeks possessor of leading technology and new engineers in the world); (3) the ago, I testified before the House Sub- the sole source of talent that can U.S. share of papers and patents; and committee of the Judiciary Commit- produce the next important ideas. (4) the U.S. capacity for innovation. tee responsible for immigration. The So, let’s circle the wagons and make Another report, Rising Above the subject was foreign-born students, visas to study in the United States Gathering Storm, issued by the especially in the physical sciences even harder to get! National Research Council on and engineering. Here are a few What would have happened if October 12, also documents how the undisputed facts: between 1980 and that had been the prevailing attitude global innovation environment is 2000, the percentage of Ph.D. scien- in the past? Fifty years ago many of changing. Chaired by NAE member tists and engineers employed in the our scientific leaders came from Norm Augustine, the report was United States who were born abroad Europe—Einstein, Fermi, and Teller produced by a committee of CEOs, increased from 24 percent to 37 per- (without whom we might not have university presidents, and Nobel lau- cent; the percentage of foreign-born been the first to build the atomic reates. Backed by copious data, the Ph.D. engineering students today is bomb), von Braun (without whom report provides a reasoned discussion close to 60 percent; one-fourth of the we might not be ascendant in rock- of why the strategies by which the engineering faculty at U.S. universi- ets and space), and von Neumann United States achieved its current ties was born abroad; and from 1990 (without whom we might not be The 46 BRIDGE leaders in computing and informa- collectively encourages, or discour- calling for reform in engineering tion technology). Today, Europeans ages, innovation. A few of the com- education since I became NAE pres- aren’t the only ones contributing to ponents of this environment are a ident, and I think we have moved our prosperity and our security— vibrant research base; an educated the ball down the field a ways. But think of Praveen Chaudhary (now workforce; a culture that permits, not nearly far enough. We will not director of Brookhaven National even encourages, risk-taking; a be able to compete with Chinese Laboratory); C.N. Yang (Nobel lau- social climate that attracts the best and Indian engineers on price, so we reate physicist, from the Institute for and brightest from anywhere in the must make sure our engineers are Advanced Study in Princeton); and world to practice engineering; worth five times as much. We can Elias Zerhouni (who was born in “patient capital” available to the only do that by reforming engineer- Algeria and is now the director of entrepreneur; tax laws that reward ing education. the National Institutes of Health). investment; appropriate protections Third, we must provide a nurtur- There are also an enormous number for intellectual property; and laws ing social climate for U.S.-born stu- of journeyman scientists and engi- and regulations that protect the dents to pursue careers in physical neers whose individual contributions public while encouraging experi- sciences and engineering. The pro- will never be as celebrated, but with- mentation. We must do better in portion of U.S. undergraduate stu- out whom the United States would every one of these areas. dents studying engineering is the be neither as prosperous nor as lowest in the developed world—4 to secure as it is today. Priority Areas for Improvement 5 percent, as opposed to 12 to 13 per- Some of you saw my article in the First, the erosion of our physical cent in most European countries and fall issue of The Bridge in which I sciences and engineering research more than 40 percent in China. The characterized the issue of visas for base and the increasing focus on U.S. currently produces only 7 per- foreign students as just one tile in a short-term results will lead to a long- cent of new engineers in the world mosaic that depicts short-term term decline in the quality of U.S. each year. Although engineering thinking, attempts to preserve the research capability. It takes 15 years has historically been considered a status quo, and a lack of long-term for ideas to make their way from pathway to upward economic mobil- investment—in short, exactly the research laboratories to products, so ity and, for decades, classes were kind of thinking that Christensen the consequences of this neglect will overwhelmingly populated with argues dooms industry-leading, good not be apparent for a long time. immigrants and their offspring, companies that fail to adopt disrup- When they do become apparent, minorities are largely absent from tive technologies. even assuming we can muster the our engineering classes today. I don’t have time to discuss all of political will to reverse the decline, Clearly, young people do not the tiles in this mosaic, but I’ll just it will take a long time to undo the consider engineering an inviting, mention a few: proposed new poli- damage. Alas, it is clear that physi- interesting, and rewarding occupa- cies for handling “deemed exports”; cal sciences and engineering tion. That is partly perception and the creation of an undefined class research is not a current public pri- partly reality. Yes, K–12 teachers, of sensitive, but unclassified, infor- ority, and hence not a priority of our counselors, and many parents are mation; and continuing reductions government. Only the federal gov- not well informed about engineer- in federal support for research in ernment can reverse this decline— ing as a career and do not urge their physical sciences and engineering by increasing funding for university charges to consider engineering as in favor of support for more short- research and providing incentives a career. And yes, there is an incor- term research. for industry-funded research. rect stereotype of engineers as geeks Second, we must ensure the high and nerds. And yes again, we have Strategies for the Future quality of the workforce. This is our been incorrectly blamed for causing There is a widespread consensus problem, and we, the engineering some environmental problems. that innovation is critical to our community, have the ability to fix But, we have also contributed to future prosperity. In my view, there it! A little money would probably the uninviting image with our is no simple formula for innovation. help, but even without money, we boot-camp style of curriculum and A multicomponent “environment” can change a great deal. I have been our nineteenth-century pedagogy! WINTER 2005 47

Fourth, we must provide a wel- The U.S. government must change National Academies to review all coming social climate for inter- this. Fortunately, it will cost noth- aspects of the environment that will national students and scholars. As ing, but we have to make the case. support innovation with a view to most of you know, after 9/11, the Finally, I alluded earlier to a multi- suggesting reforms and renewals in United States imposed stringent component environment that sup- light of current technology. requirements on students—even ports innovation. Some aspects of senior scholars—for getting visas to this environment were created in Conclusion enter the United States. Thanks to the context of technologies of the My message today can be summed the efforts of the National Acade- past, and today they are being up simply. The United States is mies and others, the average time strained to the breaking point to enormously prosperous, in no small for processing a student visa is now cope with emerging technologies. measure because of the innovative less than two weeks. However, that Here are some examples. First, the contributions of engineers. In the is not the whole story. double-blind clinical trial, consid- process of developing the very tech- Even though the average time is ered the gold standard of the FDA nologies that have made us prosper- less than two weeks, there is a long approval process, is not well suited ous, however, we have also enabled tail on the distribution—and many to ensuring the efficacy and safety of others to compete with us on a more visas still take a year. Moreover, the emerging “designer drugs,” that is, level playing field. This is generally process, both at the embassy and at drugs created to treat a specific dis- a good thing, because a rising tide the border, can be demeaning. Inter- ease in a specific patient. Second, lifts all ships, and a more prosperous national press reports tend to focus the intellectual property system, world will surely be a safer world. not on the average time to process a which was designed for macro- But the strategies that helped us get visa request, but rather on the scopic, physical machines, is being to the top are not the ones that will extreme cases. More worrisome than strained, to say the least, when lead us to greater security, prosperity, the visa situation itself is that, in just applied to sequences of microscopic and health in the future. As difficult a few years, the image of the United DNA. Third, antitrust laws that and uncomfortable as it is, we must States abroad has changed from were designed to break up railroad change—and we must do it before it an inviting “land of opportunity” to and steel monopolies are being is too late. Some of what needs to a hostile, xenophobic country. The applied to software companies. be done is under our control or can best and brightest have other In my remaining two years as NAE be influenced by us. Therefore, it’s options—and they are taking them! president, I hope to find a way for the time we got started. Like NOW! The 48 BRIDGE

2005 NAE Founders Award Acceptance Remarks

Berkeley to become an engineer. She responded, “Hey, that’s great. Those driv’n jobs pay real good.” At the time I thought she was a little out of touch. I was so young I didn’t get the underlying message. Ever since then, I have been con- tinually amazed by questions like “what do engineers do?” and “what is engineering?” In the 1970s, I ran the undergraduate mechanical engineer- ing program at Berkeley and heard these questions all the time from anxious parents and bewildered stu- dents. I used to say, “Look at every- Craig Barrett, Dan Mote Jr., Wm. A. Wulf, and Paul Peercy. thing around you. If you didn’t dig it out of the ground or grow it, it’s engi- The 2005 National Academy of Moore received the Founders Award neering.” Now even growing it is Engineering Founders Award was in recognition of his remarkable probably engineering, too. presented to C. Dan Mote Jr., accomplishments. I recall thinking It is impossible to exaggerate the president of the University of that it was indeed a fitting tribute to importance of engineering in our soci- Maryland and Glenn L. Martin Moore, but that I would sooner be ety. How could something so obvious, Institute Professor of Engineer- elected president of the United so omnipresent, and so increasingly ing. Dr. Mote was honored “for States than qualify for it myself. critical to the future of the world, let the creation of a comprehensive Actually, I feel the same way at this alone to each of us individually, body of work on the dynamics of moment. This is a truly humbling remain so obscure to so many smart moving flexible structures and for experience. people—even the majority of people? leadership in academia.” These As is customary, the Founders Compounding the confusion is remarks were delivered on Octo- awardee offers some comments to that essentially every engineering ber 9, 2005, at the Annual Meet- the captive audience that has some professional society has been pro- ing of the National Academy of obligation to listen, if the remarks moting the public understanding of Engineering. are short enough. That, I can engineering for decades. It is a plank On the day of my induction into promise. This may be an opportu- in every professional society mission the Academy, I sat out in the audi- nity to offer you a nontraditional statement and an item on every torium, inspired by the entire thought. As my friends lament, I annual agenda. These societies have Academy enterprise, especially by seem to have a few of them, but one put time and money into improving its members, its responsibilities, its in particular has gnawed at me for public understanding of engineering opportunities, and indeed its man- decades and this may be an opportu- through books, photographs, TV date to leadership in the most nity to put it before you. programs, our own A Century of advanced technological society in While in high school I drove a Innovation coffee-table book, and history. I haven’t gotten over all delivery truck for a fuel and ice other truly marvelous stuff. We that yet. And over the years since, company in my hometown. On my deserve an A+ for effort for sure. the Academy has gotten stronger, last day on the job, after dropping But still, most of us agree, the more confident, and more active in off a 50 pound sack of ice cubes at a patently obvious remains remarkably fulfilling its responsibilities. local bar, I told the proprietress I obscure. Highly educated people, On my induction day, Gordon was quitting my job and going to and many on our university campuses WINTER 2005 49

too, still ask, “What is engineering?” them over on the other side about decided to be a part of the public, felt They simply don’t get it. the values of engineering. “Public a part of the public, and were seen So we have to face the reality that understanding of engineering” is as part of the public, the task of so far the “public understanding of framed as a “them and us” concept. improving the public understanding engineering” theme has not sold Furthermore, it seems that both of engineering might be more suc- well. If we were a business selling engineers and the public have cessful. After all, in that case us that product, we would have gone accepted the “them and us” idea. would be them. bankrupt long ago. Still, we persist, Our acceptance of it shows up in It seems that a necessary first step possibly because it seems like the various ways, like engineering curric- in solving this problem might be a right thing to do, or possibly because ula at universities that favor tech- mission plank titled “Engineering nobody has a better idea, or possibly nical/scientific topics and place less Understanding of the Public.” The because we don’t really care. emphasis on humanistic and social better we understand the public, My wife once went shopping sciences. It shows up in Dilbert car- the more likely we will come up with a friend who bought scarves for toon characters, whose dress, wit, with an efficacious method of her daughters. When asked if her and cynicism signal them as engi- improving its understanding of all daughters wore scarves, her friend neers. It shows up in the “nerd char- sorts of things . . . including engi- said no . . . but they should! Does acterization” that many engineers, neering. Our understanding of that remind you of anything? Our and the public, use to describe the the public would lead us to new public understanding theme seems bright, eccentric, antisocial charac- problems and new values for engi- to fit Einstein’s definition of insan- ters who can do things but are a little neering, possibly even new com- ity fairly well. He noted that, strange. We don’t mind it. Actually, mitments for the Academy. “Insanity is doing the same thing we like it. An engineer as highly On Earth Day 1970, Pogo over and over again and expecting organized but lightly humanized is remarked, “We have met the enemy, different results.” widely accepted. Just recently at a and he is us.” I wonder if Pogo was So why hasn’t it worked? Why charity event in Washington, a CEO an engineer. But alas, we are who does the public still not get it? That expressed surprise to me that an engi- we are, although we can also shape might be a good place to start. I can neer could serve as a president of a what we will become. It’s the latter only speculate, because this is obvi- public university. I mean, it’s so pub- thought that offers me hope. ously a difficult question. But I will lic, I heard! And that was not the Thank you for your patient atten- tell you what I think. first time, either. tion and the truly esteemed honor I believe the phrase itself, “public Many of us lament that the engi- conferred by the 2005 Found- understanding of engineering,” pro- neering profession is not well repre- ers Award. jects the problem. The statement sented in Congress. But we are The Founders Award was established characterizes a divide between more likely to fault Congress for this in 1965 by the National Academy of engineering and the public that has deficiency than to fault ourselves. Engineering to honor an outstanding to be bridged. Ironically, the state- Congress is more them than us. I NAE member or foreign associate who ment is an engineering concept in don’t recall us ever deciding to get has upheld the ideals and principles itself (as is the Academy logo) that after this problem. of NAE through professional, educa- fosters the idea that “engineering” I have long suspected that this tional, and personal achievement and is represented by engineers on one them and us positioning has kept the accomplishment. For further informa- side of the bridge and the “public” public distant from engineering. The tion, contact the NAE Awards Office is represented by everybody else society’s deepening technological at (202) 334-1237. on the other side. It says that we base has excavated the divide rather engineers on this side need to teach than bridged it. If we engineers The 50 BRIDGE

2005 Bueche Award Acceptance Remarks: Some Issues Affecting Science and Technology Policy

research, why should the federal government pay most of the cost?” The answer lies in the nature of basic research—its outputs are knowledge and understanding that are retained mainly in the mind of the investigator, who is free to leave Company A, which might have paid for the work, and join Company B, which competes with Company A and never intended to support its competitor. Naturally, Company A prefers to let the government pay for its basic research. The federal gov- ernment in turn is motivated to pur- sue its basic research goals wherever Craig Barrett, Leo Young, Wm. A. Wulf, and Paul Peercy. it can find the most competent sci- entists and engineers, regardless of The 2005 Arthur M. Bueche ation of many dedicated co-workers affiliation. Award was presented to Leo and the opportunities provided by 2. My second example concerns Young, director of research my employers. In the interests of full secrecy versus openness in science (retired), Office of the Secretary disclosure, the five employers with and technology. This issue has been of Defense, U.S. Department of whom I have been associated in the addressed at length by committees I Defense, for “sponsoring collabo- last 52 years are Westinghouse Elec- have served on at both the National rative research programs among tric Corporation, Stanford Research Academies and the American Asso- academic, industrial, and govern- Institute, the Naval Research Labo- ciation for the Advancement of Sci- ment engineers and scientists.” ratory, the U.S. Department of ence. You have probably admired These remarks were delivered on Defense (DOD), and, since retire- the childlike statue of Einstein in October 9, 2005, during the ment from full-time employment, as front of this Academy building; you Annual Meeting of the National consultant to Filtronic. may even have noticed the three Academy of Engineering. Thus, my career in research and quotations from Einstein’s writings development has spanned all three I want to express my sincere that are inscribed on the stone wall sectors—government, academia, and thanks to the Academy for honoring behind the statue. One of them industry. Much of my work has been me with the Arthur M. Bueche reads as follows: “The right to search in basic research, where the govern- Award for contributions to science for truth implies also a duty; one ment is usually the largest bill payer, and technology policy. I met Dr. must not conceal any part of what academia is the major performer, and Bueche more than 25 years ago one has recognized to be true.” This industry is the chief beneficiary. I when he gave a talk at an annual is in direct contrast to the Code would like to give you three examples IEEE technology policy conference of Ethics of the National Society of of science and technology policy here in Washington. I had no Professional Engineers, which states issues that I have encountered in my inkling then that one day you would in part: “Engineers shall not reveal career in basic research. bestow on me an award in his name. facts, data, or information without I could not have received this 1. People sometimes ask: “If indus- the prior consent of the client or award without the help and cooper- try is the chief beneficiary of basic employer . . . ” WINTER 2005 51

Both statements make sense in may escalate to a major policy issue see the science and technology pic- context, so I don’t think there is a buried in a relatively minor budget ture as a whole, to look for unex- major problem in the broad outlines detail that can have unintended con- pected or unintended consequences of policy. However, the devil is in sequences. DOD budgeted $1.5 bil- of policy changes, to ensure open the details. Industry has proprietary lion for basic research for fiscal communications in fundamental interests in competitive products 2005. However, for fiscal 2006, research, to promote multidiscipli- or process developments. Similarly, DOD submitted to Congress a bud- nary research, and to encourage academic scientists competing for get request of just $1.3 billion for contact and cooperation among publication priorities or Nobel prizes basic research, that’s $200 million government, academia, and indus- have been known to be even less less (13 percent less, not counting try, which often boils down to some- forthcoming than engineers, in spite inflation). Since DOD concentrates thing as simple as showing respect of Einstein’s admonition. For exam- much more heavily on engineering for, and understanding of, another ples, just read what two Nobel lau- than other agencies that sponsor person’s point of view. reates have written, biologist James research in science and technology, Thank you for your attention. Watson in his book, The Double cutting the DOD basic research The NAE Council established the Helix: A Personal Account of the Dis- program will disproportionately Arthur M. Bueche Award in 1983 to covery of the Structure of DNA cut into university research in engi- honor statesmanship in science and (Simon & Schuster, 1998) and neering, although that was surely not technology. Arthur M. Bueche was physicist Charles Townes in his the intention of the budget cutters. senior vice president for corporate tech- book, How the Laser Happened: This is a policy issue that concerns nology at General Electric and a mem- Adventures of a Scientist (Oxford this Academy and that Congress ber of the NAE Council who spoke out University Press, 2002). needs to address. for the advancement of technology. 3. Now let’s return to the issue of A final word. Policy gurus and funding for basic research, which research directors must endeavor to The 52 BRIDGE

Gordon Prize Lecture

difficult in a traditional engineering curriculum. Therefore, we needed a new curricular structure that would not only continue to provide tech- nical depth, but would also provide experiences that would build these additional capabilities. At about the same time, commu- nity service organizations were faced with having to take advantage of technology to improve, coordinate, account for, and deliver services to the people who depend on them. Gordon Prize Team: William Oakes, Leah Jamieson, and Edward Coyle. Their challenge was to find long- term, low-cost, customized technical On February 20, 2005, the coming decades. With our students assistance. National Academy of Engineer- in mind, we will present an EPICS was created out of the real- ing recognized the achievements of overview of the EPICS Program, ization that these mutual needs pro- the Engineering Projects in Com- describe the needs it was designed to vided a unique opportunity for munity Service (EPICS) Pro- address and its success in meeting long-term partnerships between the university and the community. Ide- gram with the Bernard M. those needs, and identify the chal- ally, these partnerships would provide Gordon Prize for Innovation in lenges facing EPICS and other inno- two benefits: (1) academic credit for Engineering and Technology Edu- vative programs. significant learning opportunities for cation. The lecture was delivered EPICS was created as a result of engineering students via long-term, by Edward Coyle on behalf of the discussions about the state of under- large-scale, real-world design projects EPICS team. graduate engineering education in that would benefit the community; the early 1990s. At the time, the My colleagues, Leah Jamieson and (2) access for community part- engineering education community and Bill Oakes, are sitting with you ners to the low-cost technical exper- was drawing fire from industry and in the audience . . . but are with me tise they need to improve their in spirit here on the stage. As elsewhere for graduating students capabilities to serve the community. founders and directors of the EPICS with strong technical backgrounds The combination of challenging (Engineering Projects in Communi- but few of the other skills they engineering design projects and long- ty Service) Program, it is a great needed for successful careers. The term service to the community has pleasure for us to be here to present consensus was that students needed proven to be extremely successful. the Gordon Prize recipient lecture. professional skills, such as the ability EPICS students are learning both the The three of us once again thank to work in a team environment, technical and professional skills they the National Academy of Engineer- communicate effectively, work with need . . . and the products they devel- ing, Mr. and Mrs. Gordon, and the customers, and manage projects; op and deliver are being used every Gordon Foundation for choosing awareness of the many issues that day by their partners in the commu- the EPICS Program for this honor. affect engineering projects, including nity. Perhaps the most compelling We also thank you for creating this ethical, legal, and environmental measure of success has been the dis- superb prize, which fosters innova- issues; and the ability to work with semination of the EPICS model. In tions in engineering education that people from many different back- just 10 years, 15 other universities are critical to preparing engineering grounds and in many social settings. have adopted it. In addition, many students at Purdue and elsewhere for The challenge to educators was that EPICS alumni have remained the challenges they will face in the teaching these skills is notoriously involved with the program. WINTER 2005 53

Examples of EPICS Projects down the length of the track. With the EPICS Program—the EPICS EPICS projects fall into four practice, they can make the car Entrepreneurship Initiative, which broad categories: education and really zip down the track! is intended to spread the benefits of outreach, human services, access EPICS products by commercializing The OL JMG Joint Services Project at and abilities, and the environment. the ones that address the most sig- Bedford-North Lawrence High School We have chosen three of the more nificant unmet needs. The poten- than 200 projects around the world The second example is an access tial of products is determined via an to profile. Taken together, these and abilities project in the EPICS annual product-feasibility competi- three projects illustrate the techni- program at Bedford-North Lawrence tion called the EPICS Idea-to-Prod- cal depth, multidisciplinary breadth, (BNL) High School in Bedford, uct Competition. and community impact of EPICS Indiana. The community partner, The BNL EPICS team’s system and demonstrate how it has been OLJMG Joint Services, is a five- won second place in the competi- adapted for different environments. county agency that coordinates tion last April. They used the funds special education in the North they won to file a provisional patent The Imagination Station Project at Lawrence Community School Sys- on their product. Not bad for kids Purdue University tem. This is the first EPICS program still in high school! Imagine what This education and outreach pro- not associated with a university. they might do once they have engi- ject is based at Purdue University. We are proud to say that this pro- neering degrees! The community partner, the Imagi- ject was started by EPICS alums who nation Station, is a science and had taken jobs near Bedford. Work- The Waiheke Island Waste space museum in Lafayette, Indiana, ing with their employers, the Crane Resource Trust Project at the that provides hands-on science, Division of the Naval Surface War- University of Auckland space, and technology experiences fare Center and Visteon, and with The dissemination of the EPICS to stimulate young minds. The teachers at the local secondary Program made a great leap forward museum offers a variety of inter- school, they created the first high with the addition this past year of our active displays, many of which were school EPICS program. The com- first international site at the Univer- developed by Purdue EPICS teams. mitment of these alums to continued sity of Auckland in New Zealand. The projects developed with and service to the community demon- Our third example, an environmen- delivered to Imagination Station strates the long-term impact of tal project, is from this new site. cover a very wide range of disciplines, EPICS on the lives of our students. Two teams from the Auckland including electromagnetism, aero- In this project, high school stu- EPICS program are working with dynamics, and hydrology. For exam- dents developed a system consisting the Waiheke Island Resource Trust ple, an interactive wind tunnel was of several devices to enable a fellow to improve the environment and designed and created by an EPICS student with cerebral palsy to sense economy of the island. One team is team to provide elementary-school when she needs to swallow in order developing a portable glass-crushing children with an opportunity to learn to avoid drooling. One device, plant to process waste glass collected about aerodynamics. The wind tun- which measures the time between on the island into clean sand for use nel has a removable test section with swallows, is integrated into an in construction materials. This pro- different interactive modules illus- inconspicuous necklace. If the time ject is turning a waste product that trating the principles of lift and drag. between swallows does become too would otherwise have to be shipped Another project, called the Mag long, another device worn on the off the island into an economic Racer, teaches children about electro- wrist or the waist can either vibrate resource. A second team is develop- magnetism. It consists of a magnetic or make a noise to remind her to ing a pilot facility for processing car inside a tube-shaped track that swallow. This is clearly an innova- waste cooking oils, primarily from runs through a series of electromag- tive device—the students found restaurants on the island, into nets. The kids try different strategies nothing else like it, despite diligent biodiesel. A potential pollutant is for activating the electromagnets to patent and product searches. thus being turned into an alterna- figure out how to make the car race This brings us to a new aspect of tive fuel for municipal vehicles. The 54 BRIDGE

The Curricular Structure the students and disciplines they this end, the three of us have do- of EPICS need for their projects. nated our personal portions of the The unique curricular structure of From an educational point of view, Gordon Prize to create an endow- EPICS enables our teams to design the long-term continuity of EPICS ment for EPICS. We also continue and deliver many different products teams enables students to experience to work with schools that want to to their community partners. Some the whole design cycle, from problem start new programs, particularly EPICS teams have been in operation definition through support of fielded through an annual conference at for as long as 10 years and have deliv- projects. The EPICS Entrepreneur- which representatives of current pro- ered many projects to their partners. ship Initiative takes this cycle one grams share their experiences. When The EPICS curriculum is imple- step further by providing opportuni- we talk with other schools about mented as a “track” of courses, and ties for our students to learn about EPICS, we emphasize major themes an EPICS team corresponds to a and pursue the commercialization of woven into a student’s EPICS experi- division or lab section of a course. the products they create. ence. Two of the most important Each team has 10 to 20 students The long-term continuity of pro- themes are context and time. and is vertically integrated—that jects also enables each student to The Context Theme is, composed of freshmen, sopho- play different roles on the team. As mores, juniors, and seniors. A stu- new members, they are trained in Long-term community partner- dent may be a member for up to their team’s technologies and ships have created a context for four years, registering for one or two processes by returning members. As compelling projects that can engage credits each semester. When seniors seniors, they are often leaders of sub- students and hold their interest and graduate, returning students move teams or the entire team. commitment over the course of sev- up a year and new students are eral semesters and years. Because A Brief History of the added to the team. Many teams projects are defined by needs identi- EPICS Program have even developed training fied by the community, the students processes for new members. The EPICS was launched at Purdue in know that, if the project is done large team size, vertical integration, the fall of 1995 and now has 30 teams well, it will be put to use. This adds and credit structure enable a team and a total enrollment of more than the very important dimensions of to continue with some returning 300 students per year. More than responsibility, accountability, and students each semester and each 2,000 students have participated at commitment, which help students year. In effect, each team functions Purdue, and in a typical semester mature as individuals, as engineers, as a small engineering design firm more than 20 disciplines are repre- and as citizens. with the community partner as its sented. The teams have delivered Students must also address key customer. This enables our teams more than 200 projects to the questions. Will an exhibit for a to tackle and complete projects of Lafayette, Indiana, community. children’s museum be able to with- significant size, complexity, and EPICS has been named an exem- stand the use and abuse of hundreds impact in the community. plary program by the National Sci- of children over a period of many EPICS began in electrical and ence Foundation (NSF) Corporate months? Will an upgraded software computer engineering, but good and Foundation Alliance. Its dis- interface enable social-service solutions to real problems almost semination has been supported by workers to work more efficiently or always require contributions from NSF, the Corporation for National will it slow them down—or, even other disciplines. Thus, EPICS and Community Service, and sev- worse, annoy them? Will a system spread rapidly to other areas of engi- eral corporate partners. With this meant to detect and reward good neering—first mechanical engi- support, EPICS programs have posture in young children with neering, then civil engineering, and been created at 15 additional uni- cerebral palsy measurably improve so on. It also spread to disciplines versities and one high school. their posture? These kinds of ques- outside of engineering—to com- But this is just the beginning. We tions rarely arise—and certainly puter science, sociology, and then hope that the recognition brought to cannot be answered—in project- to many other disciplines. EPICS EPICS by the Gordon Prize will based courses that do not have real teams advertise each semester for accelerate its dissemination. Toward customers. Yet, only by answering WINTER 2005 55

these questions do students learn if must identify new challenges for passionate about what they can they are really making a difference engineering education, determine achieve and, therefore, passion- in the world. how the EPICS Program can pro- ate about what they are learning. vide insights into those challenges, The Time Theme and be aware of what else lies ahead. Changing the Face of Engineering Very few real-world projects fit A second challenge facing engi- neatly into 15-week semesters or Integrating Diverse Skills into neering is the question of who will 30-week academic years. One of the Curriculum become an engineer. We are hope- the most innovative aspects of the As demands on engineering edu- ful that EPICS partnerships between EPICS curriculum is that it enables cation increase, we must ask our- engineering and the community will project teams to bridge these artifi- selves fundamental questions about not only transform our students, but cial boundaries. Our teams have how to design efficient, affordable will also help transform the face of time to deliver well designed, well curricula that will prepare our stu- engineering. tested projects, to gather feedback dents for future careers. As more is Since the early 1990s, the United and improve their designs, and to expected of our graduates, we face States, as a nation, has made almost work with their community partners what could be called a grand chal- no progress in diversifying engineer- to identify new opportunities. lenge for engineering education— ing in some key dimensions. There From the students’ perspective, how to integrate an increasing are fewer first-year female engineer- extended participation on an EPICS number of necessary skills—for ing students today than there were team gives them time to learn both example, analytical, problem solv- in 1990 and significantly fewer disciplinary depth and multidiscipli- ing, design, and professional skills— female students in computer sci- nary breadth. It also gives them into the curriculum. We believe ence. EPICS, however, is resonating time to gain a sense of the role of that programs such as EPICS, which with young women. engineering in society, develop self- can integrate sometimes disparate There is a growing awareness and team-awareness, and master a components of the curriculum, can that, for women and for students wide range of professional skills. prepare students to meet these from some underrepresented groups, The long-term participation of advi- expectations. a major factor in career choice has sors, as well as students, provides We also believe that context and to do with making a difference. many opportunities for mentoring. time will be key enablers of this inte- However, engineering has not tradi- One of the greatest rewards of advis- gration for many reasons: tionally been thought of as one of ing an EPICS team is getting to the “caring professions.” But this is • It motivates learning of fundamen- know students well and watching an image we can build, both for our tals via compelling applications. them develop throughout their aca- students and for the community as a demic careers. • It provides a setting where stu- whole, through the results of our dents can bring both analysis and students’ work. Thoughts for the Future of design to bear. Engineering Education Conclusion • It encourages students to learn on Long-term, large-scale, community- A fitting way to conclude this their own to meet the needs of based projects have been a very overview is with an anonymous their projects. effective way of providing students quote from an end-of-semester course with many of the educational expe- • It realizes efficiencies in the cur- evaluation by an EPICS student at riences they need to prepare them riculum by continually building Purdue: “No longer is engineering for their careers. EPICS has thus on past experience. just a bunch of equations. Now I see successfully addressed many of the it as a means to help mankind.” • It provides time that bridges challenges for which it was designed. As William Butler Yeats said: semesters and years to weave con- Of course, as we all know, the “Education is not the filling of a pail nections between content knowl- world changes and the challenges but the lighting of a fire.” EPICS is edge and design skills. facing engineers continue to evolve clearly one of the matches that can . . . and we must evolve with them. • It takes advantage of compel- light this fire. To keep up with these changes, we ling contexts to make students The 56 BRIDGE

Christine Mirzayan Science and Technology Policy Graduate Fellows

engineering. Her long-term goal is to contribute to the establishment of sound national and international science policy. Amit S. Mistry is currently pursu- ing a Ph.D. in bioengineering at Rice University, where he also earned a B.S. in chemical engineer- ing. His dissertation will address the degradation and biocompatibility of Tammy Bosler Amit Mistry a polymer/ceramic nanocomposite bioengineered for the treatment of Tammy Bosler was awarded a educational resources for educators severe bone injuries. Amit has also Ph.D. in physics from the Univer- and to stimulate interest in math taught chemistry, physical science, sity of California-Irvine (UCI) in and science. Tammy also taught and algebra to high school stu- October 2004. Her research was physics and astronomy classes at dents in an underserved community focused on observational astro- UCI and was awarded a departmen- in New Orleans as a participant in physics in the context of stellar tal award for teaching. She taught a Teach for America. Currently, he is spectroscopy, stellar evolution, and course on thermodynamics at the a volunteer with Engineers Without galaxy formation. Dr. Bosler also University of Regensburg, Ger- Borders, where he works with other earned an M.S. from UCI and a many, and recently developed a students to design and implement B.A. in physics from Temple Uni- two-day astronomy workshop for engineering technologies to versity in Philadelphia. the general public, which was first improve the quality of life in devel- Tammy has a great deal of experi- presented in July 2005 at The oping countries. Amit also volun- ence as an educator. She was coor- Crossings: A Progressive Learning teers with Asha for Education, an dinator of an astronomy and Center, Meeting Place, and Well- organization that supports educa- astrophysics outreach program at ness Spa in Austin, Texas. tion projects in India. UCI, for which she created basic At NAE, Tammy is working at At the National Academies, astronomy curricula and demon- the Center for the Advancement of Amid hopes to work on broadening strations for students from 3rd grade Scholarship on Engineering Educa- the study of science. He plans to through high school. The goal of tion (CASEE) to develop the combine his scientific knowledge the program was to help schools “Guide to Proposing and Managing with his policy experience gained at that had low scores on standardized NSF Engineering Education Pro- the Academies to pursue a career in tests in math and science develop jects” to promote gender equity in science policy. WINTER 2005 57

ExxonMobil Scholar-in-Residence

Center for the Advancement of Department of Chemical and Bio- Scholarship on Engineering Educa- chemical Engineering at UMBC. tion (CASEE). She received her Her long-term goal is to work at the B.S. from the University of Mary- interface of technology and society land Baltimore County (UMBC) to promote public knowledge of and her Ph.D. from the University health issues. of Delaware, both in chemical At CASEE, Camelia’s first prior- engineering. Camelia’s thesis was ity is to explore existing K–12 engi- focused on the development of an neering activities to identify automated insulin delivery device common objectives, dissemination Camelia Owens for the management of blood glu- routes, and standards. She will then cose levels in people with Type 1 develop a systems model of engi- Camelia Owens is an Exxon- diabetes mellitus. During the past neering education that highlights Mobil Scholar-in-Residence at the year (2004–2005), she was a visit- the critical steps in the education of National Academy of Engineering ing assistant professor in the an engineer.

CASEE Scholar-in-Residence

Christina Vogt, currently a Dr. Vogt was a senior technical CASEE Scholar-in-Residence, is manager and product manager at a working toward increasing the subsidiary of Lockheed, where she number of women in engineering ran an international technical education programs. Before relo- group working on computer graph- cating to the National Academies ics systems. After leaving the com- in Washington, D.C., Dr. Vogt puter industry, she taught in India, taught courses in the Urban Educa- eastern Asia, and the Middle East. tion Program at the University of When she returned to the United Southern California (USC), where States, she spent seven years at the Christina Vogt she had extensive experience with USC Rossier School of Education, Title I school teachers and learned where she held a number of teach- the ins and outs of the challenges ing and research positions. facing inner-city school systems. The 58 BRIDGE

U.S. Frontiers of Engineering Holds 2005 Meeting at GE Global Research Center

third talk addressed advances in RFID and activity recognition. The engineering for developing communities session included pre- sentations on challenges and oppor- tunities for engineering to alleviate poverty and achieve sustainability. Talks focused on the challenges of implementing appropriate tech- nologies, exemplified by the experi- ence of the DISACARE wheelchair project in Zambia; the contributions of engineering to the safe water system program of the Centers for Disease Control and Prevention; sustainable development through Eduardo Misawa (NSF) and Silvia Ferrari (Duke University) discuss an issue during one of the U.S. FOE meeting breaks. green engineering; and the impor- tance of making sustainability sci- The eleventh annual U.S. FOE to monitor, identify, and track ence a guiding scientific principle, Symposium was held at GE Global objects and people. Two talks as illustrated by the development of Research Center in Niskayuna, New focused on face recognition—first, solar electricity markets in develop- York, on September 22–24. NAE an overview on the difficulties of ing nations. member William F. Banholzer, for- ensuring reliable identification, and The third session was on engi- mer vice president of GE Plastics, second, a talk on challenge prob- neering complex systems, from arranged for GE Global Research to lems and independent evaluations metabolic pathways and ecosystems house the meeting, which not only in automatic face recognition; a to the Internet and the propagation gave participants an opportunity to visit a corporate research facility, but also defrayed a substantial portion of the cost of the symposium. The 103 engineers who attended the 2005 symposium heard talks on four topics: ID and verification tech- nologies, engineering for developing communities, engineering complex systems, and energy resources for the future. Papers based on six of the pre- sentations are published in this issue of The Bridge. The papers in the session on ID and verification technologies were based on the premise that the prolif- eration of cheap and novel sensors, Daniel Kammen (University of California, Berkeley) responds to a question about his presentation in the U.S. FOE faster computers, and intelligent session on Engineering for Developing Communities. Speaker Julie Beth Zimmerman (University of Virginia and algorithms has made it much easier Environmental Protection Agency) is standing on the left. WINTER 2005 59

of HIV infection. Overall, the pre- labs. This was followed by a won- Agency, U.S. Department of Defense sentations provided an overview of derful dinner at the Saratoga Auto- (DDR&E-Research), U.S. Depart- theoretical and experimental tools mobile Museum. One of the high ment of Homeland Security, for meeting the challenges posed by points of the meeting was an inspir- National Science Foundation, complex systems in a systematic ing talk on Thursday evening, Microsoft Corporation, Cummins way. Talks covered network theory “Engineering for a New World,” by Inc., and Dr. John A. Armstrong as a tool for describing, analyzing, Shirley Ann Jackson, president of and other individual donors. and understanding complex systems; Rensselaer Polytechnic Institute. NAE has hosted annual U.S. FOE the engineering of biological sys- She described the links between symposia since 1995. FOE also has tems; and agent-based modeling, the critical energy issues facing the bilateral programs with Germany, which is being used to study diverse world and the importance of Japan, and India. FOE meetings systems, from ant colonies to trading encouraging innovation in future bring together outstanding young in financial systems, traffic patterns, engineers and scientists. engineers from industry, academia, and the spread of epidemics. Pablo P. Debenedetti, Class of 1950 and government and provides them The session on energy resources Professor, Department of Chem- an opportunity to learn about for the future included a presenta- ical Engineering, Princeton Uni- cutting-edge developments, tech- tion on organic-based solar cells; an versity, chaired the organizing com- niques, and approaches in many overview of research by the U.S. mittee and the symposium for the fields of engineering, which is Department of Energy on hydrogen third (and final) year. NAE presi- becoming increasingly important as production and storage; and a talk dent Wm. A. Wulf expressed his engineering becomes more interdis- on advances in fuel cells. appreciation for Dr. Debenedetti’s ciplinary. The meeting also facili- A new feature at the meeting contribution to the FOE Program tates contacts and collaborations this year was “get-acquainted ses- in his opening remarks. The 2006 among the next generation of engi- sions” on the first afternoon, which organizing committee, chaired by neering leaders. enabled attendees to get to know Julia M. Phillips, director of the All of the presentations of the 2005 more about each other relatively Physical, Chemical, and Nano Sci- symposium (including the papers in early in the program. Each partici- ences Center at Sandia National this issue of The Bridge) will be pub- pant was asked to bring one slide Laboratories, is already planning for lished in the annual FOE volume, that captured the essence of his or the next U.S. FOE meeting, which which will be available in February her research or technical work. will be held September 21–23, 2006, 2006. For more information about After breaking into small groups, at Ford Motor Company in Dear- the symposium series or to nominate participants presented their slides born, Michigan. an outstanding engineer to par- and answered questions. On the Funding for the September 2005 ticipate in future meetings, contact second afternoon, participants were symposium was provided by GE, Air Janet Hunziker at the NAE Pro- given tours of GE’s nanotech- Force Office of Scientific Research, gram Office at (202) 334-1571 or by nology, energy, and biotechnology Defense Advanced Research Projects e-mail at [email protected]. The 60 BRIDGE

NAE Hosts Japan-America Frontiers of Engineering Symposium

computers, and DNA-directed for- and Kazuhiro Sakurada, head, mation of nanoscale wires for use in Nihon Schering Research Center, a DNA identification system. Nihon Schering K.K., co-chaired The Thursday evening dinner the organizing committee and the speech was given by Thomas Baer, symposium. Dr. Fujimoto, who consulting professor in applied completed his third year as U.S. co- physics at Stanford University and chair of JAFOE, will be succeeded founder of Arcturus Bioscience Inc. by Glenn H. Fredrickson, professor In a talk titled “Engineering Entre- of chemical engineering and materi- preneurs,” Dr. Baer described the als and director of the Mitsubishi process of starting and sustaining an Chemical Center for Advanced entrepreneurial enterprise, particu- Materials at the University of Cali- larly in terms of financing cycles and fornia, Santa Barbara. The 2006 human resources. He noted that the JAFOE meeting will be held unique environment of Silicon Val- November 9–11, 2006, in Japan and JAFOE meeting participants watch Sony robot QRIO ley, which is an “incubator” for new will cover topics in systems biology, during the poster session on humanoid robots. companies, alleviates some risk solid-state lighting and displays, because of the proximity of poten- cybersecurity, and biomechatronics. On November 3–5, the 2005 tial employers if a venture fails. Funding for the 2005 JAFOE Japan-America Frontiers of Engi- Other highlights included poster meeting was provided by Hitachi neering (JAFOE) Symposium was sessions on the first afternoon, when Global Storage Technologies, held at Hitachi Global Storage all participants had a chance to National Science Foundation, Japan Technologies (HGST) in San Jose, describe their technical work or Science and Technology Agency, California. NAE member Frederick research; presentations by HGST and the NAE Fund. For more infor- (Rick) Dill, Distinguished Engineer staff on their work; and a tour and mation about the symposium series at HGST, was instrumental in facili- dinner at the Computer History or to nominate an outstanding tating the hosting of this event. Museum in Mountain View. engineer to participate in a future Approximately 60 engineers— James G. Fujimoto, professor, JAFOE meeting, contact Janet 30 from each country—attended, Department of Electrical Engineer- Hunziker at the NAE Program with additional representation from ing and Computer Science, Massa- Office at (202) 334-1571 or by HGST and NAE’s partners in this chusetts Institute of Technology, e-mail at [email protected]. program—the Japan Science and Technology Agency and the Engi- neering Academy of Japan. The four sessions at the meeting were focused on humanoid robots, pure water technologies, research and development on semiconduc- tors, and the detection and destruc- tion of pathogens. Presentations— by two Japanese and two Amer- icans on each topic—focused on advances in autonomous and inter- active behaviors of humanoid robots, the removal of arsenic from The Homestead, meeting site of the JAFOE symposium, hosted by Hitachi Global Storage Technologies in San drinking water, silicon quantum Jose, California. WINTER 2005 61

Leading Philanthropists Inducted into Einstein Society

on the front lawn of the National Academies Building on Constitu- tion Avenue in Washington, D.C. Each replica includes the signature of sculptor Robert Berks and is engraved with the donor’s name and the following quotation from Albert Einstein: “The right to search for truth also implies a duty: One must not conceal any part of what one has recognized to be true.” During the presentation ceremony, Dr. Wulf told the honorees, “You should take great pride in knowing that your philanthropic contribu- tions are making a significant differ- Dr. and Mrs. Leon Kirchmayer’s son and daughter receive an Einstein statuette on their parents’ behalf at the an- ence in NAE’s service as adviser to nual meeting. From left to right: Craig Barrett, Karen Demuth, Kenneth Kirchmayer, and Wm. A. Wulf. our nation.” Members of the Einstein Society are recognized at annual meetings and meetings of the Presidents’ Circle of the National Academies. They will also be recognized in the headquarters building in Washing- ton, D.C., and in annual reports and donor-recognition issues of Academies publications. For infor- mation about becoming a member of the Einstein Society, please con- tact the Office of Development at (202) 334-2431.

NAE Members of the Einstein Society From left to right: Sheila Widnall, Wm. A. Wulf, and Craig Barrett display the Einstein statuette bestowed upon John A. Armstrong for contributions to members of the Einstein Society. the endowment for the Young Engi- Three leading philanthropists to $100,000 or more. Thirty of the 86 neers Program NAE and the National Academies, members of the Einstein Society are Holt Ashley for contributions to NAE Harold K. Forsen, William W. Lang, NAE members. Independent Funds and Olga Kirchmayer (in memory of In appreciation of their exem- Norman R. Augustine for contribu- her husband, Leon K. Kirchmayer), plary commitment, NAE president tions to the Senior Scholars Program were recently inducted into the Ein- Wm. A. Wulf presented Forsen, William F. Ballhaus Sr. for contribu- stein Society, which was created in Lang, and Kirchmayer’s son and tions to NAE Independent Funds 2004 to honor members, private daughter (attending on her behalf) Jordan J. Baruch for contributions to donors, and others whose cumu- with replicas of the original ma- NAE Independent Funds and Capi- lative lifetime donations total quette of the Einstein monument tal Preservation The 62 BRIDGE

Stephen D. Bechtel Jr. for contribu- Phil Smith Fund How People Learn, Presidents’ Circle, tions to NAE Independent Funds, Dane A. Miller for contributions to the Woods Hole Renovation, Frank Press Public Understanding of Engineer- Center for the Advancement of Fund, and NAE Independent Funds ing, K–12 Engineering Education, Scholarship on Engineering Educa- Simon Ramo for contributions to NAE and Greatest Engineering Achieve- tion and NAE Independent Funds Independent Funds, Bruce Alberts ments book Richard M. Morrow for contributions Fund for Science Education, NAS Bell Family Foundation for contribu- to the Center for the Advancement Annual Fund, Frontiers of Science, tions to NAE Independent Funds of Scholarship on Engineering Edu- Young Engineers Program, Guy George M.C. Fisher for contributions cation, NAE Independent Funds, Stever Fund, Woods Hole Renova- to NAE Independent Funds and and Capital Preservation tion, and NAS Endowment Capital Preservation Kenneth H. Olsen for contributions to Walter L. Robb for contributions to the William L. Friend for contributions to NAE Independent Funds Center for the Advancement of NAE Independent Funds Doris Pankow (in memory of Charles J. Scholarship on Engineering Educa- Grainger Foundation for contributions Pankow), for contributions to Engi- tion, Capital Preservation, and Fron- to the establishment of the Grainger neering Ethics, Capital Preservation, tiers of Engineering Challenge Prize Public Understanding of Engineer- Anne P. Rowe (in memory of Joseph E. Bernard M. Gordon for contributions ing, Frontiers of Engineering, and Rowe) for contributions to Engi- to the establishment of the Bernard NAE Independent Funds neering Education and Research, M. Gordon Prize for Innovation in Jack S. Parker for contributions to the Information Technology, and NAE Engineering and Technology Educa- Presidents’ Circle Communication Independent Funds tion and for NAE Independent Funds Initiative, How People Learn, and Dolores H. Russ (in memory of Fritz J. Anita K. Jones for contributions to NAE Independent Funds Russ), for contributions to the estab- NAE Independent Funds Robert A. Pritzker for contributions to lishment of the Russ Prize Thomas V. Jones for contributions to the Bruce Alberts Fund for Science Alan M. Voorhees for contributions to NAE Independent Funds, Capital Education, NRC Presidents’ Fund, Urban Infrastructure for Sustainabil- Preservation, and Industry Scholars and NAE Independent Funds ity in China and NAE Independent Ruben F. Mettler for contributions to Allen E. Puckett for contributions to Funds support media relations, Public the NRC Presidents’ Fund, NAS Wm. A. Wulf for contributions to Understanding of Engineering, and Annual Fund, Capital Preservation, NAE Independent Funds

Calendar of Meetings and Events

2006 February 15 NRC Executive Committee April 6 NAE Regional Meeting January 17 NRC Executive Committee Meeting University of Michigan Meeting February 21 NAE Awards Forum/Awards April 11 NRC Executive Committee February 6–7 NRC Governing Board Meeting Dinner Meeting March 2–4 Indo-U.S. Frontiers of February 8–9 NAE Council Meeting All meetings are held in the Academies Building, Engineering Symposium Irvine, California Washington, D.C., unless otherwise noted. Agra, India February 9 NAE National Meeting For information about regional meetings, please Irvine, California March 14 NRC Executive Committee contact Sonja Atkinson at [email protected] or Meeting (202) 334-3677. WINTER 2005 63

In Memoriam

SPENCER H. BUSH, 85, presi- FREDERICK J. ELLERT, 76, ALAN S. MANNE, 80, Professor dent, Review and Synthesis Associ- retired general manager, Systems Emeritus of Operations Research, ates, died on October 2, 2005. Dr. Development and Engineering Stanford University, died on Sep- Bush was elected to NAE in 1970 Department, General Electric tember 27, 2005. Dr. Manne was for work on the physical and Company, died on July 13, 2005. elected to NAE in 1990 for out- mechanical metallurgy of materials Dr. Ellert was elected to NAE in standing contributions to operations used in nuclear reactors. 1987 for his leadership in the devel- research methodology and applica- opment and application of high- tions for production scheduling, MARVIN CHODOROW, 92, Pro- voltage direct-current technology plant capacity decisions, and energy fessor of Applied Physics and Elec- for large-scale electric utility power planning. trical Engineering, Emeritus, networks. Edward L. Ginzton Laboratory, RONALD F. SCOTT, 76, Dotty Stanford University, died on Octo- LEOPOLD B. FELSEN, 81, pro- and Dick Hayman Professor of ber 7, 2005. Dr. Chodorow was fessor of aerospace and mechanical Engineering, Emeritus, California elected to NAE in 1967 for his out- engineering and professor of electri- Institute of Technology, died on standing work on microwave tubes. cal and computer engineering, August 16, 2005. Dr. Scott was Boston University, died on Septem- elected to NAE in 1974 for contri- LELAND C. CLARK JR., 86, vice ber 24, 2005. Dr. Felsen was elected butions to the theory and applica- president of research, Synthetic to NAE in 1977 for his contribu- tion of soil mechanics. Blood International Inc., died on tions to the theory and application September 25, 2005. Dr. Clark was of microwave propagation in com- CHEN-TO TAI, 88, Professor elected to NAE in 1995 for inven- plex media and for his leadership in Emeritus, Department of Electrical tions and contributions to biosen- engineering education. Engineering and Computer Sci- sors and artificial organs and blood ences, University of Michigan, and for medical applications of these DONALD R.F. HARLEMAN, 82, died on July 30, 2004. Dr. Tai was devices worldwide. Ford Professor of Environmental elected to NAE in 1987 for basic Engineering, Emeritus, Ralph M. contributions to the advancement W. KENNETH DAVIS, 87, Parsons Laboratory for Water of electromagnetic theory and retired vice president, Bechtel Resources and Environment Engi- applications to antenna design. Corporation, former deputy secre- neering, Massachusetts Institute of tary of energy, and consultant, died Technology, died on September 28, HARVEY A. WAGNER, 100, on July 29, 2005. Mr. Davis was 2005. Dr. Harleman was elected to retired executive vice president, elected to NAE in 1970 for his con- NAE in 1974 for his leadership in Detroit Edison Company, died on tributions to the development and the development of theoretical and June 30, 2005. Mr. Wagner was industrial application of nuclear experimental techniques in the field elected to NAE in 1970 for leader- power technology. of fluid mechanics. ship in the development of high- temperature conventionally fueled and nuclear power plants. The 64 BRIDGE

Publications of Interest

The following reports have been dent, automotive, Aluminum Com- director, Chemical and Physical published recently by the National pany of America, chaired the study Sciences Laboratory, Ford Motor Academy of Engineering or the committee. Other NAE members Company; John B. Heywood, Sun National Research Council. Unless on the committee were Uma Jae Professor of Mechanical Engi- otherwise noted, all publications are Chowdry, vice president, Central neering and director, Center for for sale (prepaid) from the National Research and Development, 21st Century Energy and Sloan Academies Press (NAP), 500 Fifth DuPont Company Experimental Automotive Laboratory, Massachu- Street, N.W., Lockbox 285, Wash- Station, and Jennie S. Hwang, pres- setts Institute of Technology ington, DC 20055. For more infor- ident, H-Technologies Group Inc., (MIT); John G. Kassakian, profes- mation or to place an order, contact and chief executive officer, Asahi sor of electrical engineering and NAP online at http://www.nap.edu or America Inc. Paper, $35.00. director, Laboratory for Electromag- by phone at (888) 624-8373. (Note: netic and Electronic Systems, MIT; Prices quoted are subject to change Review of the Research Program of the Christopher L. Magee, professor, without notice. Online orders receive a FreedomCAR and Fuel Partnership: First Engineering Systems Division, 20 percent discount. Please add $4.50 Report. The FreedomCAR and Fuel MIT; Michael P. Ramage, retired for shipping and handling for the first Partnership is a collaborative effort executive vice president, Exxon book and $0.95 for each additional by the U.S. Department of Energy Mobil Research and Engineering book. Add applicable sales tax or GST (DOE), the U.S. Council for Auto- Company; Bernard I. Robertson, if you live in CA, DC, FL, MD, MO, motive Research (USCAR), and five senior vice president, Engineering TX, or Canada.) major energy companies to oversee Technologies and Regulatory research on a “clean and sustainable Affairs, and general manager, Truck Globalization of Materials R&D: Time transportation energy future.” The Operations, DaimlerChrysler Cor- for a National Strategy. The global goal of the project is to enable a tran- poration (retired); and Kathleen C. spread of materials science and engi- sition first to more efficient internal Taylor, retired director, Materials neering (MSE) R&D is accelerat- combustion engines (ICEs), then to and Processes Laboratory, General ing. As a result, the U.S. position in advanced ICE hybrid electric vehi- Motors Corporation. Paper, $18.00. a number of MSE subfields is in a cles, and ultimately to a private- state of flux. To analyze the impli- sector decision, by 2015, on the Earth Science and Applications from cations of these trends for the U.S. development of hydrogen-fueled Space: Urgent Needs and Opportunities economy and national security, the vehicles. This report, which builds to Serve the Nation. The Earth is a U.S. Department of Defense asked on an earlier NRC study, The Hydro- dynamic planet with changes and the National Research Council to gen Economy: Opportunities, Costs, variations that affect communica- assess the status and impacts of the Barriers, and R&D Needs, includes an tions, energy, health, food, housing, global spread of MSE R&D. This evaluation of research on hydrogen- and transportation infrastructure. report includes a discussion of the fueled transportation systems and Understanding these changes factors that influence U.S. compa- findings and recommendations for requires a range of observations from nies’ decisions about where to locate technical directions, strategies, fund- a variety of land-, sea-, air-, and MSE R&D facilities, impacts on the ing, and management. space-based platforms. To help the U.S. economy and national security, NAE member Craig Marks, National Aeronautics and Space and recommendations for ensuring retired vice president, Technology Administration, National Oceanic continued U.S. access to critical and Productivity, AlliedSignal Inc., and Atmospheric Administration, MSE R&D. chaired the study committee. Other and U.S. Geological Survey develop NAE member Peter R. Briden- NAE members on the committee these tools, the National Research baugh, retired executive vice presi- were Peter Beardmore, retired Council was asked to develop a WINTER 2005 65

decadal strategy for surveying Earth A. Armstrong, retired vice presi- underground facilities to conceal from space. The study committee dent for science and technology, and protect strategic military func- was asked (1) to identify key scien- IBM Corporation; Alice P. Gast, tions and weapons stockpiles. tific questions to be answered by vice president for research and asso- These strategic, hardened, deeply Earth and environmental observa- ciate provost, Massachusetts Insti- buried targets are often only reach- tions from 2005 to 2015 and (2) to tute of Technology; Joel Moses, able by conventional or nuclear provide a prioritized list of space Institute Professor, professor of com- earth-penetrating weapons (EPWs). programs, missions, and supporting puter science and engineering, and Recently, an engineering feasibility activities to address these questions. professor of engineering systems, study, the Robust Nuclear Earth This interim report outlines the ratio- Massachusetts Institute of Technol- Penetrator Program, was initiated nale for linking Earth observations to ogy; and Elsa Reichmanis, director, by U.S. Department of Energy and societal needs and a discussion of the Materials Research Department, U.S. Department of Defense to most urgent near-term actions. A Lucent Technologies. Paper $43.00. determine if a weapon based on the final report, due in late 2006, will rec- major components of existing ommend space missions, programs, Technology Pathways: Assessing the nuclear weapons would be an effec- and related activities. Integrated Plan for a Next Generation tive EPW. The program has gener- NAE member Warren M. Wash- Air Transportation System. In 2003, ated a great deal of controversy ington, senior scientist and section Congress directed the secretary of about the collateral damage such a head, Climate Change Research transportation to establish the Next weapon would cause. To clarify the Section, Climate and Global Generation Air Transportation Sys- issues, Congress (P.L. 107-314) Dynamics Division, National Cen- tem (NGATS) Joint Planning and directed the secretary of defense to ter for Atmospheric Research, was a Development Office (JDPO) to plan request that the National Research member of the study committee. the development of an air trans- Council conduct a study to assess Paper, $12.00. portation system capable of meeting the anticipated health and environ- potential air traffic demands for mental effects of nuclear EPWs and Policy Implications of International 2025. To break down interagency other weapons and the comparative Graduate Students and Postdoctoral barriers and promote cooperation effectiveness of using conventional Scholars in the United States. This and collaboration, all federal agen- and nuclear weapons to destroy bio- report provides an in-depth discus- cies involved in aviation are partic- logical and chemical weapons stor- sion of the impact of foreign-born ipants in the JDPO. The National age facilities. The study committee students and scholars on U.S. edu- Research Council was asked to undertook detailed numerical cal- cational institutions and the U.S. examine the first NGATS Inte- culations to compare the effective- economy. The United States has grated Plan prepared by JPDO and ness and expected collateral damage depended increasingly on human submitted to Congress in 2004. of nuclear EPWs and surface resources from abroad for its science This assessment includes a review of nuclear weapons under a variety of and engineering workforce. How- the vision and goals, operational conditions. ever, competition for talent has concepts, and R&D road map; an NAE member John F. Ahearne, been increasing as other countries analysis of the JDPO integrated director, Ethics Program of the Sig- expand their research infrastructures product teams created to plan the ma Xi Center, chaired the study and provide more opportunities for initiative; and an assessment of committee. Other NAE members workers educated in science and the implementation process. on the committee were Richard L. engineering. The report analyzes NAE member David C. Wisler, Garwin, Emeritus Fellow, IBM T.J. trends in international student manager, University Programs and Watson Research Center, Eugene enrollments and retention rates and Aero Technology Labs, GE Aircraft Sevin, independent consultant, the impact of visa policies. Engines, was a member of the study Lyndhurst, Ohio, and Robert H. NAE members on the study com- committee. Paper, $18.00. Wertheim, Rear Admiral, U.S. mittee were William G. Agnew, Navy (retired), and consultant, Sci- retired director, Programs and Plans, Effects of Nuclear Earth-Penetrator and ence Applications International General Motors Corporation; John Other Weapons. Many nations use Corporation. Paper, $33.00. The 66 BRIDGE

Review of the GAPP Science and Imple- advise CMA about the status of ana- research that might accelerate the mentation Plan. Water managers rely lytical instrumentation technology development of cures for spinal on predictions of seasonal, annual, and systems suitable for monitoring cord injuries. Although many of and interannual changes in hydro- airborne chemical warfare agents at the recommendations are presented logic cycles. Predictions of hydrologic chemical weapons disposal and stor- in the context of specific needs cycles are based on local and remote age facilities. The report includes an articulated by the New York Spinal influences, including land processes assessment of current monitoring Cord Injury Research Board, the and ocean processes, such as the systems used to detect airborne panel of experts approached El Niño Southern Oscillation. A agents at CMA facilities and of the research priorities for the field in better understanding of land-surface applicability and availability of inno- general. The panel also recom- processes can potentially improve cli- vative new technologies. The report mends ways to improve and coor- mate predictions, but using this infor- also includes a discussion of how new dinate the existing infrastructure. mation to make water-management regulatory requirements would affect Funders at federal and state agen- decisions is still risky because current CMA’s current agent monitoring cies, academic organizations, phar- models provide only limited informa- procedures and whether new mea- maceutical and medical device tion on seasonal and longer term surement technologies could be companies, and nonprofit organiza- changes. The Global Energy and effectively incorporated into CMA’s tions will find this report essential Water Cycle Experiment (GEWEX) overall chemical agent monitoring reading. of the Americas Prediction Project strategies. NAE member P. Hunter Peckham, (GAPP) was established in 2001 to NAE member Elisabeth M. professor of biomedical engineering, improve predictions of changes in Drake, retired associate director for Case Western Reserve University, water resources on intraseasonal- new energy technology, Energy Lab- was a member of the study commit- to-interannual time scales for the oratory, Massachusetts Institute of tee. Hardback, $49.95. continental United States. GAPP Technology, was a member of the developed a Science and Imple- study committee. Paper, $25.25. Going to Extremes: Meeting the Emerg- mentation Plan, which describes ing Demand for Durable Polymer Matrix strategies for improving predictions Spinal Cord Injury: Progress, Promise, Composites. Advanced polymer and decision support in the hydro- and Priorities. An estimated 11,000 matrix composites (PMCs) have logic sciences. This report by the spinal cord injuries occur each year many advantages, such as light National Research Council reviews in the United States, and more weight and high specific strength, the GAPP Science and Implemen- than 200,000 Americans suffer that make them useful for many tation Plan and suggests improve- from maladies associated with aerospace applications. Their use ments for the plan and the GAPP spinal cord injuries. These include has been limited, however, by overall program. paralysis, bowel and bladder dys- uncertainties about the long-term NAE member Soroosh Sorooshi- function, sexual dysfunction, respi- changes in properties of PMCs an, UCI Distinguished Professor and ratory impairment, temperature under extreme conditions. The director, Center for Hydrometeorol- regulation problems, and chronic U.S. Department of Defense ogy and Remote Sensing, Univer- pain. In the last two decades, long- requested a study focused on sity of California, Irvine, was a standing beliefs about the inability methodologies for predicting long- member of the study committee. of the adult central nervous system term performance. This report pro- Paper, $31.75. to heal itself have been eroded by a vides a review of the challenges to flood of information based on the use of PMCs in extreme envi- Monitoring at Chemical Agent Disposal research in the neurosciences and ronments, the current understand- Facilities. Under the direction of the related fields. However, there are ing of PMC properties and behavior, U.S. Army’s Chemical Materials still no cures, and restoring func- a discussion of the data necessary Agency (CMA), and as mandated tion after spinal cord injuries for the development of effective by Congress, the nation is destroying remains extremely complex. This models, and recommendations for its chemical weapons stockpile. This report by the Institute of Medicine improving long-term predictive study was requested by the Army to recommends directions for future methodologies. WINTER 2005 67

NAE member Kenneth L. Reifs- Public Water Supply Distribution Sys- of their structural materials, are used nider, Pratt & Whitney Chair of tems: Assessing and Reducing Risks. as research tools in a variety of scien- Design and Reliability, University of First Report. The Water Science and tific disciplines. The National Sci- Connecticut, was a member of the Technology Board of the National ence Foundation asked the National study committee. Paper, $18.00. Research Council is studying water Research Council to assess the cur- quality issues associated with public rent state of and future prospects for Tank Wastes Planned for On-Site Dis- water supply distribution systems high-field science and technology in posal at Three Department of Energy and their potential risks to con- the United States. This assessment Sites: The Savannah River Site: Interim sumers. The distribution system—a focuses on scientific and techno- Report. In response to a request from critical component of every drinking logical challenges and opportunities, Congress, the U.S. Department of water utility—poses significant rather than on specific program Energy (DOE) asked the National challenges from the operational and activities, and offers findings and Academies to evaluate its plans for public health standpoints. This recommendations for directions for managing radioactive wastes from report was requested by the Environ- research, compares U.S. efforts to spent nuclear fuel at sites in Idaho, mental Protection Agency (EPA), those of other countries, and recom- South Carolina, and Washington. which is considering revisions to the mends improvements. This interim report evaluates stor- Total Coliform Rule and potential NAE members on the study com- age facilities at the Savannah River new requirements for ensuring the mittee were David C. Larbalestier, site in South Carolina, with a par- integrity of distribution systems. David Grainger Professor and L.V. ticular focus on plans to seal the The report identifies trends relevant Shubnikov Professor of Materials tanks with grouting. The report to the deterioration of drinking Science Engineering, University finds that tanks do not necessarily water quality in distribution systems Wisconsin-Madison, and John M. have to be sealed as soon as the bulk and prioritizes the issues of greatest Rowell, Distinguished Visiting Pro- of the waste has been removed. concern. The committee reviewed fessor, Arizona State University. Postponing permanent closure nine EPA white papers and con- Paper, $39.00. allows time for the development and cluded that cross connections and application of emerging technolo- backflow, new or repaired water Prospective Evaluation of Applied En- gies for removing and immobilizing mains, and finished water storage ergy Research and Development at DOE residual waste, without increasing facilities presented the highest (Phase One): A First Look Forward. risks to the environment or delaying potential health risks. In addition, In 2001, the National Research final closure of the “tank farms.” the committee identified premise Council (NRC) completed Energy The report also recommends alter- plumbing and operator training as Research at DOE: Was It Worth It?, natives to address the lack of tank high-priority issues. This report will a congressionally mandated assess- space at the site, as well as the need be followed in about 18 months by a ment of the benefits and costs of for focused research to reduce the more comprehensive final report DOE’s fossil energy and energy effi- amount and improve the immobi- evaluating approaches to character- ciency research and development lization of residual waste and to test izing risk and identifying strategies (R&D) programs. Congress then the assumptions used in evaluations for reducing risks. directed DOE to request that the of long-term risks. NAE member Vernon L. NRC develop a methodology for NAE member Frank L. Parker Snoeyink, Ivan Racheff Professor of assessing prospective benefits. The chaired the study committee. Other Environmental Engineering, Univer- first phase of the project to develop NAE members on the committee sity of Illinois at Urbana-Champaign, a methodology began in December were Hadi A. Abu-Akeel, senior was a member of the study commit- 2003. Phase one focuses on (1) the vice president and chief engineer tee. Print-on-demand, $18.00. adaptation of the retrospective (retired), and technical advisor, methodology to a prospective con- FANUC Robotics America Inc., and Opportunities in High Magnetic Field text; (2) develops transparent Milton Levenson, retired vice pres- Science. High-field magnets, which methodology that does not require ident, Bechtel International. Print operate at the limits of the mechani- extensive resources for implementa- on Demand, $18.00. cal and/or electromagnetic properties tion and produces easily understood The 68 BRIDGE results; and (3) describes a practical be used to review the prospective Draper Professor of Aeronautics and and consistent process for using it. benefits of DOE fossil energy and Astronautics, Massachusetts Insti- In phase two, the methodology will energy efficiency R&D programs. tute of Technology, and Maxine L. be made more robust and related NAE members on the study com- Savitz, retired general manager, issues will be explored. In subse- mittee were Wesley L. Harris, Technology/Partnerships, Honey- quent phases, the methodology will department head and Charles Stark well Inc. Paper, $38.00.

The Periodicals BRIDGE Postage Paid (USPS 551-240)

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