Harvard SEAS, Newsletter, Spring 2004

Vo l u m e I I I • I s s u e 1 • S p r i n g 2 0 0 4

faculty members like Henry Ehrenreich may bring to mind those who inspired you to work (and play) hard at Harvard. Be sure to take a look at the quotes from our graduates that appear in the alumni section. Serving society Our people and discoveries travel the globe. Ralph Mitchell’s quest to preserve the U.S.S. Arizona and Steve Wofsy’s efforts to understand our forests, both throughout the Northeast as well as in places as far away as Bhutan, are only two examples of the many ways we extend beyond Cambridge. And how we do it … As you know, what we do is not defined or constrained by independent disciplines. We tackle tough problems, developing and using whatever knowledge and tools are needed to get the job done. That means an applied mathematician studying insects (p. 2), a computer scientist influencing economics (p. 4), and an electrical engineer changing medicine (p. 14). I was struck at our recent faculty retreat by how important this open structure and culture is to the way we work. While we’ve expanded, we can all still meet in one place to discuss everything from our academic program to the interplay of technology and society. Such interaction is not possible at many other institutions and highlights the importance of our size and spirit. Ultimately, through the work we do and by the way we do it, our goal is to use what we discover, learn, and create to make the world a better place. I hope you will let me know how your experiences here – in ways great and small – may have helped you reach out and connect with other people and society. F

What’s inside
Crosscurrents Mahadevan finds the profound in the mundane, Parkes negotiates and navigates, Suo takes a drive, and Wofsy works in the wild. Faculty News Mitchell protects our past, Morrisett lands at EECS, Grosz and Vadhan create a cluster, and Division members in the news. In Medias Res An innovator returns home, researchers give the common tap a new angle, scientists poke holes in the ozone debate, and a grid is born. Student News A Bioinnovation course offers an amusing assignment, a selection of recent awards, and ES 51 students go off-road. In Profile Khaneja looks for the right path and Ehrenreich shares four decades of wit and wisdom. Outside the Quad The GK-12 program lights up teachers, Unilever provides food for thought, a list of recent industry collaborations, and faculty-student patents. Alumni Notes Iansiti explains the physics of business, grads share memories, and a calendar of upcoming events hosted by the Division. Connections Snapshots of collaborative science in action, and how to keep in touch. 20 18 16 14 12 10 6 2

Reaching out
n this issue of the newsletter you will see how members of the Division of Engineering and Applied Sciences reach out to solve tough problems, collaborate with industry, enlighten students, and serve our society and the world. I am proud not only of this work, but how we do it – without walls, without departments, and without limits. The work we do … Fundamental research Innovative research is one of the cornerstones of the Division. As you read about Zhigang Suo’s molecular car, you might imagine how his insights could radically change manufacturing. Industrial collaboration We connect with companies and they with us to share knowledge and knowhow. An award from IBM with support from Intel will allow the Division to develop the Crimson Grid, a computer network designed to solve a variety of complex problems right from a researcher’s (and perhaps one day, your) desktop. Education and mentoring Teaching makes a profound connection, as students take what they’ve learned and apply it to everything they do. The GK-12 program, TECH-supported courses in bioinnovation, and

Dean’s Message

I

DEAS Spring 2004 I 1

L

Moreover, stopping to smell (and study) the roses might someday help with creating self-assembling nanostructures, The Division’s Lakshminarayanan Mahadevan (Maha for one of the most critical components of the emerging field of short), Gordon McKay Professor of Applied Mathematics nanotechnology. and Mechanics, has not only posed the question, but is trying to solve countless others like it. Using mathematics to understand how materials move and behave, he places particular emphasis on phenomena visible to the naked eye Maha’s dark, vibrant eyes flit behind large round glasses that and closely tied to experiments or experience. He’s explored reflect the light in his sizable, but test tube–free third floor the way honey coils (important for geologists who study the office of Pierce Hall. In his experience, you don’t necessarily flow of molten rock within the Earth), why insects can adhere need a lab or complicated devices to do meaningful experito surfaces (leading to the creation of ments and research. “Why struggle to new types of adhesives), how hair coils “Natural systems offer a rich find something worth studying when on water (helpful in understanding the you have quick access to rich events, arena to learn about the principles of self-assembly), and the like how a flag flutters, that you can way fabrics fold and wrinkle (providing interplay between geometry easily play with? Being able to radically insight about the spiky surface of a disand physics in the real world.” change parameters – a light breeze vereased red blood cell). sus a strong wind – without losing the

ike a slow-motion release of a firework, a flower’s bud bursts forth into a delicate display. Looking at a bouquet at the florist, you might never ask how the intricate petals, stems, and stigmas, each contained in a green orb the size of a gobstopper, emerged perfectly unfolded without the slightest rip.

“Natural systems offer a rich arena to learn about the interplay between geometry and physics in the real world. Folding is not just for flowers, but critical to our very existence. It happens in our tightly bound-up DNA,” Maha points out.

Crosscurrents



“I find joy in discovering the sublime in the mundane,” says Maha, who recently relocated from one Cambridge (England) to another (Massachusetts). “I try to uncover explanations for everyday events that are easily seen but not well understood. They typically turn out to be more relevant than I first imagined.” Think again about the complexity of the flower as you recall how you’ve struggled to fold a map without tearing, or at least swearing. Within the blooming process lies what Maha calls a theory for “self-assembled origami.” The bud can unpack its suitcase and iron its clothes without the help of even a finger.

effect is ideal for experimentation.” With today’s emphasis on rapid innovation, supermarket science (creating volcanoes with baking soda) and everyday experimentation (looking out the window rather than at an LCD monitor) may seem passé. Yet Maha’s hands-on experimentation, most of which could have been done by true renaissance engineers, does not imply that such research is any less difficult or fruitful. Rather, he acknowledges that good science can arise from simple observation. Not surprisingly, Maha’s “mundane” investigations cross – if not leap – over traditional boundaries in physics, math, engineering, and bi-

By paying attention to how a flower unfolds its petals, a nature lover also learns to appreciate fundamental issues in physics and applied mathematics.

2 I DEAS Spring 2004

ology. In fact, figuring out what comes naturally requires continuous collaboration with scientists from many disciplines at Harvard, MIT, and throughout the world. Maha emphasizes that his dedicated students – “the lifeblood of my enterprise” – deserve as much credit as he does for illuminating the physics of everyday life. “Ultimately, any robust event is likely to be interesting for its own sake, since it explains something essential about how the world works,” observes Maha. And he’s not alone in appreciating such rough magic. In regard to Maha and his colleagues’ celebrated theory of how wrinkles form, Nature editor Philip Ball wrote, “It is humbling to find in a high-powered journal like Physical Review Letters that we have limped along for years without an understanding of what controls the wavelength and amplitude of wrinkling in a sheet. There’s plenty still to be learned from the $20 experiment.”

Ordinary Research with Extraordinary Results
Since all the world’s his lab, Mahadevan studies a wide variety of problems using lessons from every discipline at his disposal. Françoise Brochard-Wyart and Nobel laureate Pierre-Gilles de Gennes captured the interdisciplinary spirit that pervades all his work when discussing his and E. Cerda’s groundbreaking research on the geometry and physics of wrinkling. They said, “The paper provides a beautiful and simple understanding of many natural phenomena – bridging geometry, mechanics, physics, and even biology.”


Maha explains his passion for research by referencing a classic tale about the young Krishna, from the Bhagavata Purana. The child, an incarnation of the Hindu god Vishnu but brought up incognito by foster parents, had a reputation for mischief. One day, friends accused Krishna of eating dirt. Dismayed, his mother demanded an explanation. Krishna denied the charge, saying “I have not eaten dirt. They are all lying!” To force a confession, his mother told him to open wide. But Krishna, to avoid being caught in a lie, played a trick. Instead of muddy teeth, he revealed the entire universe to her – the earth, the stars, and the elements of all creation. And then, to keep his cover, Krishna quickly cast a spell of forgetfulness over her to clear her memory. “One way to look at the story,” Maha explains, “is to understand that meaning and the answers to the deepest questions can be found in the stuff all around us. Science is about looking for connections and finding joy in discovery itself.” Lucky for us, Maha – unlike Krishna’s mother – has not forgotten where the universe lies, but continues to look deep inside simple things, like flowers or dirt, to pull out the profound. F To read more about Mahadevan’s research, visit www.deas.harvard.edu/softmat/ and see the following recent articles: “The Physics of … Wrinkles: Lines of Least Resistance,” in the November 2003 issue of Discover Magazine “Gel gains lifelike motion” in the December 31, 2003/January 4, 2004 issue of Technology Research News “Envelope physics sheds light on ice sheets” and “The physics of haute couture” in the December 2, 2003 and February 4, 2004 editions of Nature Science Update, respectively

Plumbing for pests
Some sap-loving aphids live their entire lives deep inside galls, or the abnormal swellings of plant tissue. Since what goes in (what they eat) must go out (as waste), these aphids could easily drown inside the enclosed spaces. Mahadevan and his colleagues discovered that these snug bugs secrete powdery water-repellent wax on the surface of their homes and on their waste products. The wax-on-wax formula turns the excrement droplets into “liquid marbles” that the critters can then roll clean away. Mahadevan hopes to learn from these tiny engineers how to improve attempts at efficiently manipulating minute volumes of liquid on small surfaces.

range of locomotion for radically different animals,” says Mahadevan, “and also hints that there’s an underlying, similar process for how all of them move.” This discovery might lead to new motion techniques for tiny machines, robots, or for use in manufacturing processes that involve moving substances across surfaces.

Uncovering wrinkles
Mahadevan and his colleagues from Cambridge, England, proposed a now-famous general theory about an everyday bother that keeps dry cleaners happy – the wrinkling of fabrics and other materials. When you press down on a spring you crush it, and in so doing, the spring stores elastic energy. Likewise, a sheet can either stretch or bend; the resulting deformed sheet adopts the shape that minimizes its total bending energy. Mahadevan’s laws of wrinkling predict the amplitude and wavelength of the resulting wrinkles, and work for a variety of materials – plastics, fabrics, and even human skin. Understanding how a cape falls over you or how our sheets look after a restless sleep could lead to more realistic computer animations or better-fitting clothes.

Getting gell-o to jog
One of Mahadevan’s research teams has created an “artificial animal” from a filament of cylindrical hydrogel (cut with small scales on the bottom) that, when vibrated atop a sheet of glass, can mimic – and hence, help explain – the movement of snakes, snails, and other creatures. By varying the angle, scales, and direction of the vibration, the team derived different patterns of motion. “A simple experiment can explain a wide

DEAS Spring 2004 I 3

Crosscurrents

A

lmost two decades before the iPod and a decade before the Internet, Apple quietly introduced a prototype called the Knowledge Navigator. The tablet-shaped device featured a software agent, a talking digital butler, who helped the user access an information network, manage messages and a schedule, and even take phone calls. We aren’t there yet, but David Parkes, an Assistant Professor of Computer Science on the Gordon McKay endowment, is working on it. He is developing the technology and protocols to enable software agents to help us navigate and, more important, negotiate through our daily economic and informational lives. In the near future “you will have a software agent that is helping you negotiate with other agents and managing your computational and informational needs,” says Parkes. Common examples of agents include passive software that suggests songs or movies based upon your past purchase and preference patterns and active code that automatically bids (within defined limits) on an item at an online auction site.

I want a cool collection of dance songs for my party, but I don‘t want to pay more than $20.00. I know John likes dance tracks by Outkast and Madonna since over the past few weeks he’s purchased their songs and has both artists on his favorites list. I’ll go check out what some leading online retailers suggest based on John’s profile.

John Doe

John Doe

Great selections! My friends danced till dawn! For future reference, however, I’d steer clear of Fountains of Wayne.

John‘s Agent

John‘s Agent

To create a great party mix for John, I’ll purchase 3 songs by Fountains of Wayne and 4 by Duran Duran from iTunes at $1.00 a track. I’ll buy 4 tracks by NERD and 4 by Kylie Minogue from BMG for $1.50 each. While pricey, I think they’re closer to his profile. That’s $19.00 total and in budget. Let’s deliver these and dance!

I think John might enjoy songs by Fountains of Wayne and Duran Duran. Our other members have said both are great for dance parties. I will sell you songs for $1.00 a piece.

Based on what John’s bought in the past, we’d recommend tracks by NERD and Kylie Minogue. I will sell you songs for $1.50 each.

“My overall research agenda is to design distributed systems where you need to get the economic incentives right. There is often a beautiful tension to resolve between optimal economic properties and optimal computational properties,” says Parkes. Consider peerto-peer systems (P2P) like the original Napster. “How do you get a P2P system to be well running despite the self-interested nature of participants?” he asks. Even the most well-designed e-commerce site will have little value if it allows cheating, harms the buyer (fraud), or harms the seller (like illegal file swapping). If done right, “[even] if everyone behaves in a self-interested way, you can still drive toward a goal, whether it be efficiency, fairness, or the best price,” Parkes explains. You also need robust computation to back it all up – from the interface to the behindthe-scenes algorithms to clear expressive markets – and to keep things run-

The above diagram shows how an agent might “negotiate and navigate” on behalf of John Doe to create a dance party mix. Based on John’s past purchases, preferences, and budget, his agent negotiates with various vendors such as iTunes and BMG. In turn, the sellers suggest options by using the data (what John likes and wants) the agent has provided. These e-retailers might recommend songs by comparing John’s profile with thousands of other users with similar tastes. After the agent makes the purchases, John can evaluate its performance. Over time, as the agent gains more information, it will become better and better at making the “right” decisions.

ning smoothly and quickly. Similarly, if a user must enter detailed information to set his/her preferences, constantly adjust the agent, or remains worried about privacy and security, then desirable properties could unravel. Creating better software agents is, of course, about more than just getting the best price on eBay. “We are moving away from a computational model where you buy a computer and put it on your desk. Instead, companies buy computational time, in what’s called on-demand computing,” Parkes says. In other words, a company may only need to buy a few hours’ use of a software product or only need extra server power during peak times. Amazon and eBay might compete for those limited server resources during the busy holiday

shopping season. A trusted agent – one that knows the rules and knows what a company wants – will be essential for good business. Parkes sees a great opportunity to build both the computational and economic tools necessary to drive the future of e-commerce and automated negotiation. For him, the real action – and the software agents – lie at the interface. The butler, however well spoken, is likely to remain virtual (even in the virtual world) for some time to come. F For more on Parkes’ work, visit www.eecs.harvard.edu/~parkes/ www.eecs.harvard.edu/p2pecon/

4 I DEAS Spring 2004

Driving for answers
While some strive to move mountains, Gordon McKay Professor of Mechanics and Materials Zhigang Suo wants to move a molecule. Suo and his team at DEAS have envisioned a new technology, the molecular car, designed to shuttle target molecules like passengers. Imagine if you could operate a tiny vehicle that specifies and positions individual molecules to influence a chemical or biological process. The level of control, like commanding grains of sand to march to their proper place to create the ultimate castle, would be unprecedented. What would a molecular car look like? Consider a short-chain molecule with three features (see diagram). One end adsorbs – attracts and holds molecules to the surface of its molecules – to a solid, its middle has a group with an electric dipole moment normal to the solid surface, and its other end is a passenger receptor for the molecule. To move the “car,” Suo imagines constructing an array of electrodes underneath the substrate. When the
A conceptual model of Suo’s molecular car.

Molecule (passenger) Receptor (seat) Dipole (engine)

Binder (wheel)

On-chip infrastructure
electrodes are charged sequentially, the resulting moving electric field makes the car move. Because the electric field pattern is programmable (the individual electrodes can be turned on and off at will), the car can move forward or backward, make a sharp turn, and park. For passenger pickup and drop-off, the car would attract a molecule on board in one area with favorable conditions, such as the right pH and temperature levels, move on the highway, and then release the passenger in another area using the reverse method, altering conditions to make the molecule want to hop off.

Substrate (pavement)

Suo admits that the concept of the molecular car raises as many questions as potential solutions. How could the car be controlled or account for difficult conditions such as thermal fluctuation, akin to being in a perpetual earthquake? What would happen if two cars bump into each other? At present, the molecular car only exists in computer simulations. Suo and his research colleagues are looking for ways to overcome the technical challenges required to go from the drawing board to the showroom. It may turn out that moving a molecule will be a mountainsized challenge, but the drive is likely to be exciting. F

Into the woods
Some research at the Division requires a bit of a hike. Abbott Lawrence Rotch Professor of Atmospheric and Environmental Science Steven Wofsy knows this from experience. Under Wofsy’s guidance, the Division hosts the Northeastern Regional Center (NERC) of the National Institute for Global Environmental Change (NIGEC), funded by the U.S. Department of Energy. Recently, Steven Wofsy described what it’s like to do research in the wild. What research are you most excited about? We have an integrated long-term study, which means that we looked at the changes in the ecosystem over time. The University of New Hampshire looked at how the ecosystem is responding to air pollution, and a group from Woods Hole Marine Biological Lab is investigating how the system responds to heating the soil. I am very excited about the fact that all of these things are being brought together. Do you act differently (what you drive, eat, wear) due to your research? I think I have become more respectful of the complexity of the environment than I was when I was more of an advocate. You never know everything you need to know – and that’s true of stock investing right on to the environment. What are the challenges of conducting research in the wild? Our sites get hit by lightning maybe two or three times a year. And then there are the snakes, mosquitoes, airplane schedules, customs, permission to work in a particular location, and politics. Do you need any more? But it’s exciting. F For more information about NIGEC and NERC, visit http://nigec.ucdavis.edu/

Steven Wofsy conducts a field study in Bhutan. “We met someone at Harvard who was a botanist in Bhutan who invited us to go there and help them with their forest inventory. So we spent a month learning about how they do things and showing them how we do things.”

DEAS Spring 2004 I 5

Faculty News
only the “day of infamy” at Pearl Harbor – December 7, 1941, when a surprise air attack by the Japanese crippled the U.S. Pacific Fleet and left 2,390 dead – but, through its bent metal and lost heroes, the turmoil of the entire war. During the attack, a 1,760-lb. armorpiercing bomb blew up half the ship, causing the rest to sink to the shallow bottom, forty feet down. Today, a bowed white walkway lies like a silk ribbon over the remaining hull. After taking a short boat ride to the memorial, the 1.4 million annual visitors can peer down at the ghostly shadow through a mix of blue waves churning uneasily, but often beautifully, with the Bunker C fuel oil that inexorably bleeds from the ship. “The ship still contains about half a million gallons of oil,” says Mitchell, who recently returned from a site visit. “Meanwhile, at about a quart a day, there’s a steady drip of oil that’s
Ralph Mitchell holds up a flask containing a sample taken from the site of the U.S.S. Arizona.

how microorganisms adhere to and grow on surfaces and form biofilms,” he says. Part of his work is to study the chemical transformations mediated by microorganisms found on the surfaces of both living and artificial materials. Microbes live and breed on everything, most commonly in the form of thin sheets called biofilms. A common example is the plaque that forms on your teeth. If you neglect to brush, what you leave behind will eventually eat away at the tough enamel, leaving you with pain, a disapproving dentist, and a hefty repair bill for the cavity. That process may be exactly what’s happening to the Arizona – the brew of seawater, microbes, fuel oil, and time is taking its toll. In addition, with the right combination of conditions the bacterial population can be transformed from something benign into microbes that can chew through metal. Like many of our national monuments – the Statue of Liberty or the Lincoln Memorial – the Arizona does not sit in the climate-controlled safety of a museum. Much of the ship’s impact is in its ability to give visitors a real link to the past. Not being encased in glass, however, is especially perilous for a sunken ship. The leaking fuel oil is now only a drip, but as corrosion continues, the risk of an oil spill increases. The National Park Service is not simply letting nature take its course. In the past several years, as part of their ongoing research to preserve the ship, they have sent divers down to take water and oil samples from around the vessel. In addition, ROVs (tethered remotely operated vehicles) have explored parts of the ship.

ordon McKay Professor of Applied Biology Ralph Mitchell watches his mail like a teenager expecting a bulky college acceptance envelope. “This week we are getting archival metal that came off the ship. And we are getting oil. We already have microorganisms …” he says with the trace of an Irish accent, hinting at his undergraduate days at Trinity College, Dublin, where he studied microbiology. The deliveries, looking like props for an episode of the hard-hitting, but often less than hard science, television drama “CSI,” actually add up to an act of conservation. Mitchell is collaborating with a team of marine archeologists and microbiologists, led by the National Park Service. Their task: to preserve and protect one of the United States’ most sacred national monuments, the U.S.S. Arizona.

G

“...there’s a steady drip of oil that’s rising to the surface of the harbor. The survivors call it the tears of the men who died.”
rising to the surface of the harbor. The survivors call it the tears of the men who died. You can actually smell the oil. It’s not ephemeral, but real.” While the Arizona is a tomb for 1,177 fallen sailors and marines, it also harbors life. Over the decades, a host of marine organisms have settled on the silt-covered metal hull and decks. That’s where Mitchell comes in. “The underpinning for everything I do is understanding

The submerged remains of the 608foot-long, 31,000-ton naval battleship lie off the coast of Honolulu, Hawaii. The Arizona literally embodies not

6 I DEAS Spring 2004

Conservation scientists need to take great care; the Arizona, after all, is a gravesite. “The survivors, rightly, will not allow anyone inside,” says Mitchell. Further, the ship itself and the oil that threatens the harbor inextricably mesh with the memories of the survivors and the friends and families of those buried in the ocean. “So what’s really going on in there at a biochemical level? We simply don’t know.” Mitchell is convinced that the ship is home to an increasing number of harmful microbes. He is one of the world’s leading experts on biofilms and the microbiology of surfaces. With financial backing from the Division, his laboratory has the means to play a critical part in the preservation of the memorial. The three U.S.S. Arizona samples – water, oil, and metal – sent to him by the Park Ser-

vice marine archeologists will provide him with the raw materials for learning more about what’s happening without having to disturb the site. Mitchell hopes to help determine if the interaction between the organisms and the fuel oil is accelerating the corrosion, how quickly the ship’s metal is decaying, and how long it will be until the oil starts leaking at a fast rate. His Laboratory of Applied Microbiology is using modern methods of genetic analysis to identify the destructive bacteria, as well as electrochemical techniques to determine the rate of deterioration of the metal. For Mitchell, the Arizona intersects several interests – scientific (exploring biofilms), conservationist (he runs one of the only fellowships in biological

conservation in the U.S.), and, of course, personal (as an American). “We know very little about the physical, chemical, and biological processes responsible for degradation of the historic artifacts that are essential to the cultural life of our nation,” Mitchell says. His work on the World War II memorial may lead to discoveries about how to preserve other historic sites – from sunken ships to historic buildings to battlefields. “I feel fortunate to take part in this cooperative research, to use what I know to help protect something that I care about. The Arizona is a labor of love for me as well as for other scientists committed to its preservation. It is as much about securing the future of our heritage as it is about science.” F For more on the U.S.S. Arizona Memorial, visit www.nps.gov/usar/

Over 1.4 million people visit the U.S.S. Arizona Memorial each year. Chief architect Alfred Preis said of its design, “The form, wherein the structure sags in the center but stands strong and vigorous at the ends, expresses initial defeat and ultimate victory. The overall effect is one of serenity.” (Photo courtesy of the National Park Service.)

DEAS Spring 2004 I 7

New arrivals

T

It’s a matter of trust

Faculty News

he goal of Greg Morrisett’s work is to eliminate the need to discover the intent (good or bad) of a program before a user downloads or runs it. For instance, when you receive an e-mail with an attachment, the file may contain a virus or worm that could damage your computer, steal personal information, or be used to attack third parties (e.g., someone else on your shared network). Likewise, when you click on a Web page, open a document, update a device driver, or install a game, you may be executing hidden code that introduces security holes or installs a Trojan Horse (a malicious, securitybreaking program that is disguised as something benign on your machine). Unfortunately, we do not yet have a robust technical solution to easily counter these problems. Self-mutating code (programs that change and adapt) can fool standard virus scanners. Digital signatures (ways to authenticate the identity of the sender/signer of a document and to ensure that it has not been tampered with) only establish the provenance of code, not its trustworthiness. Put another way, Morrisett is working on technology that would eliminate the need for trust. He advocates an architecture that would require programs to come with an explicit, formal security policy as well as mathematical proof that the code will respect that policy when run. To make that a reality, he has been developing security and self-checking tools for software engineers. Such safeguards make it possible to automatically check that particular code will not do something “bad” when executed. In essence, it shifts the burden of proof from the code consumer to the code producer. Imagine the benefits of opening attached documents or downloading (legally, of course) music files that have already been road-tested for reliability and safety. Given the wide implications of the problems Morrisett tackles, it is not surprising that the Division’s interdisciplinary atmosphere attracted him to Harvard. “I am able to work with a wide variety of scientists and academics here. A lot of computer science problems are not just technical, but legal and social, and ultimately determine policy decisions that affect our everyday lives.” F

Greg Morrisett provides trust and support for his colleagues working at Maxwell Dworkin.

John Gregory (“Greg”) Morrisett
Allan B. Cutting Professor of Computer Science John Gregory (“Greg”) Morrisett joins Harvard after spending seven years at Cornell University, an institution as famous for its gorges as for its ivy. While there, he was an Assistant Professor (1996–2001) and then an Associate Professor (2001– 2003) in the Department of Computer Science. During the 2002–2003 academic year he left Ithaca for industry, serving as a Visiting Researcher at Microsoft’s Cambridge (U.K.) Research Lab. Morrisett studied Mathematics and Computer Science at the University of Richmond (B.S., 1989) and received both his M.S. (1991) and Ph.D. (1995) in Computer Science from Carnegie Mellon University.

Nota Bene
Laser vision ... Cable network provider Comcast featured Federico Capasso on “Technogenesis” ... In April, he will receive the 2004 Caterina Tomassoni and Felice Pietro Chisesi prize at the University of Rome ... Light my wire ... The January 29th New York Times Circuits section and the January 27th Boston Globe Health and Science section showcased Limin Tong and Eric Mazur’s work on nanowires ... Baker’s dozen ... The Harvard University Gazette chose a profile piece on biomedical engineer Kit Parker as one of its top 12 stories of the year ... Gell-o that jogs and the latest fashions ... The January 19th issue of The Scientist and the January 7th Technology Research News featured research by L. Mahadevan and colleagues on hydrogels, and the February 4th edition of Nature Science Update highlighted Mahadevan’s work on wrinkling and folding ... Shape is destiny ... The February 13th Science magazine featured a perspective on packing spheres by David Weitz ... Tour de force ... The March 3rd “Charlie Rose Show” highlighted Bill Gates’ visit to the Division ... Nano know-how ... The March 6th Boston Globe highlighted the Center for Imaging and Mesoscale Structures (CIMS) in a story about local nanotechnology expertise. F

8 I DEAS Spring 2004

Collaborations
I
Grosz and Vadhan lead a random walk through Radcliffe
s solving a problem more difficult than verifying a solution? Can any efficient process be efficiently reversed? Can you infer a global property of an object by inspecting a tiny portion of it?

What’s a Cluster?
It’s fundamental scientific research being done at Radcliffe Institute. It’s complementing, not competing with, individuals from MIT. It’s hosting international experts who speak the same languages – computer science and mathematics. Likewise, practicing innovative science relates to crafting an environment where thought is free to flow. The cluster gives its members a fresh environment – the peaceful Putnam House and the modern Maxwell Dworkin – in which to work. In lieu of the typical academic year, the program offers researchers a different tempo – the time to tackle a single area with little interruption. While these changes seem simple, cluster leader Salil Vadhan says, “Being out of your element encourages you to think in new ways, be more open to collaboration, and take risks.”

At the Radcliffe Institute for Advanced Study this year, a sixmember research cluster on randomness and computation, affiliated with the Theory of Computation Group at DEAS, is searching for answers to such questions. “Randomness is past the point of just being a topic in computer science; it has permeated the entire subject,” says Assistant Professor of Computer Science Salil Vadhan ’95, who brought together the cluster and serves as its chair. One of the chief goals of the computer science cluster is to use randomness to design efficient algorithms for complex computational tasks. Cluster members believe they can design “randomized” algorithms that offer reasonably close approximations of solutions to problems too large and complex to solve in their entirety. “The science is terrific, they’re getting work done, and their seminars and research are involving students and other faculty from Harvard and MIT,” says Barbara Grosz, Dean of Science at the Radcliffe Institute and Higgins Professor of Natural Science. “At other levels, they’re interacting with Radcliffe fellows across many disciplines. What more could you ask from a cluster?” F For more information, visit www.radcliffe.edu/research/2003_random.html
Adapted from “Understanding Randomness,” Radcliffe Quarterly, April 2004

Members of the Randomness Cluster
The cluster members comprise an eclectic mix of scientists based in academia and industry, located coast to coast and throughout the world, including: Eli Ben-Sasson, Harvard University & MIT Irit Dinur, University of California at Berkeley Oded Goldreich, Weizmann Institute of Science Shafi Goldwasser, Weizmann Institute of Science & MIT Dana Ron, Tel Aviv University Ronitt Rubinfeld, NEC Laboratories Madhu Sudan, MIT Salil Vadhan, Harvard University

Randomization is an important tool for finding an approximate answer – especially when finding an exact answer is infeasible either because the problem is difficult or because the input is very large. A well-known example is analyzing the data from a poll. By using a random sample (where each point has the equal probability of being selected) the estimate will be close to the correct fraction of purple to red points in the entire population. If the sample is large enough, you can estimate the correct fraction with a high degree of probability.

DEAS Spring 2004 I 9

Selected articles about the Division

In Medias Res

Dean Venky welcomes Bill Gates Col ’77 back to the Division. While here, Gates toured Maxwell Dworkin (a building made possible by his and Steve Ballmer’s ’77 support, and named for their mothers), met with faculty, signed a poster of the BASIC program code he wrote as an undergraduate, and spoke with students about the challenges and opportunities in computer science and engineering. (Photo courtesy of Microsoft.)

Bill Gates comes home
Microsoft Chairman William H. Gates III Col ’77 delivered a relaxed, sometimes humorous talk to about 350 students, faculty, and administrators at Lowell Lecture Hall on Thursday, February 26, outlining a software future that features smarter, more secure machines, and encouraging students to develop computing’s next big idea. Gates, who dropped out of Harvard in 1975 to found Microsoft, told students that the computer industry needs new energy and fresh minds. He said that despite advances in the past decades and the prominence of the computer in today’s society, people are still underestimating the power of current developments. Gates identified computer security as the biggest threat to his vision of a wired future, saying that if people think their financial or personal information isn’t secure, they won’t use applications that could otherwise make their lives easier. He also pointed to artificial intelligence (AI) as an area of computer science awaiting a big breakthrough. Artificial intelligence has such potential to transform the industry, he said, that a student today making a breakthrough in AI could conceivably create the next Microsoft. Fewer people are working in artificial intelligence today, however, than 20 years ago. Gates visited Harvard as part of a fivecampus speaking tour, also making stops at the Massachusetts Institute of Technology, Cornell University, Carnegie Mellon University, and the University of Illinois. Division of Engineering and Applied Sciences Dean Venkatesh Narayanamurti introduced Gates, eliciting a laugh from the crowd when he described Gates’ famously shortened stay at Harvard, saying “He spent precisely as much time as he needed to.” “You have a unique opportunity tonight to hear from someone who not too long ago was where you are now,” Narayanamurti said. F
Reprinted with permission from the Harvard University Office of News and Public Affairs, February 26, 2004 Pieces about Gates’ visit to the Division also appeared in the Boston Globe, Boston Herald, The New York Times, and Time.

News from the nozzle
Triangular nozzles provide the tiniest droplets, say Harvard Physics Department graduate student Henry Chen and Gordon McKay Professor of Applied Mathematics and Applied Physics Michael Brenner, who used a mathematical algorithm to determine that a miniature threesided tap could produce drips some 21 percent smaller than a conventional round nozzle. The miserly taps – which could, in theory, create drops just 8 billionths of a millimeter in size – might prove a boon for technologies that employ sprays of costly materials. For instance, triangular taps could boost the resolution of ink-jet printers, which work by squirting fine droplets of ink onto surfaces. They could also cut the size of traditional silicon chips and biochips, both of which feature patterns that are sometimes produced by a tightly controlled spray of droplets. “We hope that the theoretical methods we used to answer this problem will prove broadly applicable,” says Brenner. “We are trying to develop the general mathematical methods that are needed for carrying out mathematical optimizations of structures used in engineering.” In addition to taps, Chen and Brenner see their mathematical methods applied to a number of other examples, including a new switch with a shape deemed optimal and a coffee cup whose form facilitates boiling and convection. Chen and Brenner’s work is funded by the National Science Foundation and Harvard’s Materials Research Science and Engineering Center (MRSEC). F
Adapted from FAS Communications, December 12, 2003

10 I DEAS Spring 2004

Up in the atmosphere ...
Harvard researchers have observed for the first time a key molecule that destroys atmospheric ozone, confirming researchers’ theories about the stratospheric chemistry that creates Arctic and Antarctic ozone holes each winter. Rick Stimpfle, a senior project scientist with the Division, was the lead author in a paper published this month in the Journal of Geophysical Research-Atmospheres that outlined the findings. Stimpfle conducted the research along with David Wilmouth, a postdoctoral fellow in Atmospheric Chemistry, Philip S. Weld Professor of Atmospheric Chemistry James Anderson, and Ross Salawitch, a researcher at NASA’s Jet Propulsion Laboratory.

… And down on earth
Harvard researchers are weighing in on the ozone pollution debate, asserting that federal assumptions about natural background levels are wrong and may result in national standards that permit too much ozone pollution. Arlene Fiore, a postdoctoral fellow at Princeton University who conducted the research while at Harvard, and Daniel Jacob, Gordon McKay Professor of Atmospheric Chemistry and Environmental Engineering, made their assertions in a recent article in the Journal of Geophysical Research.

“Our work suggests that the Environmental Protection Agency (EPA) is presently overestimating background concentrations,” Fiore says. An image taken from NASA’s Total Ozone Mapping “This overestimate might lead to a Spectrometer (TOMS) Web site. greater health risk, due to imposing Results of the research also showed that weaker regulations on ozone.” The two researchers will share the chemical chain reaction involving the molecule likely their conclusions with federal lawmakers at a Senate briefing runs 20% – 30% faster than scientists had expected, a finding in Washington, D.C. The testimony comes as the EPA conducts that will allow them to adjust critical computer models to beta regular re-evaluation of its air pollution standards. F ter approximate real conditions in the atmosphere. F

Both articles adapted from the Harvard University Gazette, February 26, 2004

DEAS, IBM connect to create computing ‘grid’
The Division of Engineering and Applied Sciences and computer giant IBM are teaming up through an applied research award to create a pilot computer grid that, if successful, could one day provide researchers access to greatly increased computing power. The effort, called the Crimson Grid Test Bed, is being launched with the help of several parties, including IBM, Intel Corp., the Faculty of Arts and Sciences (FAS), and University Information Systems (UIS). The initiative is just part of a larger DEAS effort to work with computer industry leaders in developing new computational tools and techniques that can benefit researchers in other parts of the University. “A grid could potentially provide the tools to solve any type of problem, from a complex literature search to mining the genome,” explains DEAS’s Chief Information Officer and Information Technology Director Jayanta Sircar, who is also the Primary Investigator for the project. “Our goal is to provide the enabling infrastructure for state-ofthe-art research computing,” Dean Narayanamurti says. “Such an infrastructure is critical to several scientific disciplines, spanning areas such as high-energy physics, materials science, computer science, astronomy, and biology.” Bruce Harreld, IBM senior vice president of strategy, said in a statement, “This grid project can open doors to new research and help both organizations to draw on complementary strengths, including IBM’s expertise in grid computing, computational biology, and advanced IT (information technology) solutions.” F
Adapted from the Harvard University Gazette, January 22, 2004

DEAS Spring 2004 I 11

Student News
To be judged successful, the roller coaster design must be free-standing and have at least one loop; the marble must undergo five changes of direction, leave and then rejoin the track, and complete the course without falling or stopping. Students in ES 143/243 use overhead lights and chairs to balance their coaster creations.

Roll with it

H

arvard students from all fields of study – from government to biology to engineering – joined together at The Children’s Museum of Boston in February to collaborate on building a working roller coaster. Instead of steel and wood, students used pipe insulation tubes and toothpicks. The fearless rider of the twists, turns, and loop was an eager, but never-at-risk, marble.

The spirited “play” was part of ES 143/243, a course developed by Gordon McKay Professor of the Practice of Biomedical Engineering David Edwards with Paul Bottino, executive director of the Technology and Entrepreneurship Center at Harvard (TECH). The class encourages creativity and team building to help inspire students to translate biomedical scientific innovations into improved healthcare solutions. Dr. Louis B. Casagrande, president and CEO of The Children’s Museum, kicked off the event by sharing his experiences as an educator and museum director. “Creative play is work,” he said. The exercise’s purpose was twofold: for students to work together as a team to solve a single problem (to learn how to work and re-work ideas), and for them to explain to everyone else what worked, what didn’t, and why. Understanding how to build a roller coaster – “a kind of perverse ‘startup’ analogy,” Edwards pointed out – is more than just an exercise. By appreciating the dynamics of what first appears as a simple problem, Harvard students who want to use science to influence society gain a critical understanding of how to respond to real-world issues.

“In the end, no matter how sophisticated the scientific solution to a problem, researchers need to explain how things work in the simplest possible terms, and listen – and have fun,” said Edwards. As a case in point, one group in the current class is focused on developing a new treatment for malaria. “How to treat malaria, an epidemic in many nations of poverty, requires a non-standard pharmaceutical level of thinking,” explained Edwards, “since such regions lack the infrastructure and basic medical knowledge that most pharma solutions take for granted.” Casagrande said his ultimate mission at The Children’s Museum is to “bring out the genius in every child.” Likewise, Edwards and Bottino are striving to bring out the healthcare inventor in every student. The class has already helped apply several ideas outside of the classroom. In the fall of 2002, students helped found MEND (MEdicine in NeeD). The nonprofit enterprise has since built an internationally renowned board and advisory board and forged important pharmaceutical, NGO, and governmental relationships in its effort to develop a new treatment of tuberculosis in the developing world. In the spring of 2003, Edwards worked with several students to create Pulmatrix, a pharmaceutical startup in the Cambridge area that is currently developing proprietary aerosol products that treat, prevent, or inhibit the spread of airborne infectious diseases. While the designs for the student-built roller coasters are not likely to be appearing at a theme park near you, what was learned in creating them may serve as an inspiration to produce real ways to counter and treat diseases. F

12 I DEAS Spring 2004

Awards
2003 SAME Award … In recognition of her hard work and
dedication to research, Harvard College senior Christine Susan Mulvey was awarded the 2003 Colonel and Mrs. S.S. Dennis III Scholarship by the New York City Post of the Society of American Military Engineers (SAME). Ms. Mulvey, a 2004 candidate for the S.B. degree in Engineering Sciences with a concentration in Bioengineering, is from Canton, Massachusetts, and graduated from Canton High School in 2000. Her primary academic interest is the study of bone mechanics, or discovering how bones move and respond to stimuli. For her senior design project, Ms. Mulvey is working on a spinal testing device in consultation with the Orthopedic Biomechanics Lab at Beth Israel Deaconess Medical Center in Boston, where she interned during the summer of 2003. After graduating, she plans to pursue an academic career.

2004 Analog Devices Awards … Graduate student Yong Liu and William Andress ’05, in Assistant Professor of Electrical Engineering Donhee Ham’s lab, have won 2004 Analog Devices Outstanding Student Designer awards. F

Dean Venky presents Christine Susan Mulvey ’04 with a certificate of accomplishment and a scholarship check for $1,000.

Student drivers

(Left) Students drive for grades in ES 51; (Above left) Rob Howe adds color commentary and applause; (Above right) a student team basks in the spoils of engineering excellence. (Photos courtesy of Jon Chase and the Harvard News Office.)

At the end of last semester, DEAS undergraduates in “Computer-Aided Machine Design” (ES 51) competed for spills, thrills, and a potential A (for their final grade). Before hitting the road, students first “reverse engineer” cordless screwdrivers and then (with the help of a few additional parts from $40 worth of gears) turn them into remote-control vehicles. In addition to grease and good sense, students use 3-D modeling software that generates “code” (i.e., instructions and measurements) that is programmed into a lathe at the machine shop to cut out custom pieces. The software helps ensure that the builders have pieces that are accurately sized and shaped. On race day, the vehicles were put to a series of tests, including power (towing a heavy trailer up a ramp) and maneuverability (driving through a “slalom” course made of beakers and funnels). The grueling circuit is designed to challenge all aspects of their knowledge of engineering, from mechanical to electrical. F
DEAS Spring 2004 I 13

D

Consider NMR, a particular interest of Khaneja’s. The technique uses a sequence of radio frequency pulses to manipulate the magnetic fields surrounding an atom. Since many atomic nuclei are magnetic (due to the quantum mechanical properties of spin), these radio frequency pulses cause them to dance in specific ways. A detector can trace and identify these patterns “Using optimal control theory to of movement. Observing how the manipulate quantum systems helps nuclei of a particular researchers design molecule dance and conduct bet- “Optimal control theory is helps researchers ter experiments,” about determining the determine the strucKhaneja explains. ture of molecules “Many scientific best path to take in order themselves. tools involve control over quantum phenomena.” Examples include existing technologies such as nuclear magnetic resonance (NMR) spectroscopy (used to understand the structures of proteins), and future innovations such as quantum computers (designed to significantly increase storage and speed).

uring the cold war, the race to the moon wasn’t just about coming in first. Physicists on both sides had to determine the safest, most direct, and most fuel-efficient flight path for the rocket and its crew, all in a strange environment: space. Assistant Professor of Electrical Engineering Navin Khaneja isn’t planning the next trip to the moon, but his research on optimal control theory has significant parallels to that famous voyage. Down on earth, he is trying to find the best “flight” paths for quantum evolution in the molecular universe.

From the moon to molecules
Faculty member Navin Khaneja’s research on optimal control theory requires a broad understanding of physics and applied mathematics; the engineering savvy to understand scientific tools; and an appreciation of experimental design.

In Profile

researchers are interested in complex, large protein molecules, reducing dampening (or “relaxation” losses) is critical. To compensate for relaxation, Khaneja is developing new methods to manipulate the dynamics of coupled nuclear spins. Using optimal control theory to design radio frequency pulses, he can minimize relaxation losses. An optimal design could improve the sensitivity of experiments and shorten the time it takes to analyze the structure of complex molecules. “Imagine it like this,” he explains: “The paths to your office are slushy from the snow. You have a critical meeting and you want to arrive looking good. My challenge is to help you find the best and fastest route, so your clothes don’t get ruined. Optimal control theory is all about determining the best path to take in order to get the desired result.” At the quantum level, such guidance will help scientists improve the sensitivity and efficiency of their experiments so they can reach their destination (i.e., data) with less mess. “It’s critical to do this work,” says Khaneja. “Even a small improvement in an NMR experiment might allow researchers to understand a new aspect of a protein. And what we learn from one tool can be used to improve other devices and be applied to related problems.” Is navigating through slush or space really comparable to manipulating atomic particles? As Khaneja explains, absolutely. “You essentially use the same methods and ideas from control theory to guide a rocket to the moon or plot your course to your office as you do to optimally control quantum phenomena. It speaks to the universality and beauty of mathematical thought.” F To read more about Khaneja’s research, visit http://hrl.harvard.edu/~navin/

to get the desired result.”

Why does the use of NMR technology need improving? “With small molecules,” Khaneja says, “it is relatively straightforward to determine their threedimensional structure. But as the size of a molecule increases, the dance patterns are dampened, making it harder to extract the desired information.” Since

An elegant rendition of the optimal paths of quantum states in the presence of relaxation losses. (Image courtesy of Steffen Glaser, Technische Universität München.) For more information, see “Boundary of Quantum Evolution under Decoherence,” N. Khaneja, B. Luy, S.J. Glaser. PNAS, USA 100, 13162–13166 (2003).

14 I DEAS Spring 2004

It must be tremendously interesting to be a schoolmaster, to watch [students] grow up and help them along. I don’t see how you could ever get old in a world that’s always young. – Goodbye, Mr. Chips

Lifelong learning

W

hile the Division keeps evolving, students who graduated years, if not decades, apart often speak fondly of the same professor – still an inspiration, still hard at work. Consider Clowes Research Professor of Science and University Ombudsperson Henry Ehrenreich. In addition to serving Harvard and the Division for four decades, he offers a quiet wit and candor to everyone he encounters. Ehrenreich, a physicist who has chiefly studied semiconductor theory, says he has always felt a sense of obligation toward the University. “Harvard has been very good to me in many ways because of all the opportunities to do things … and I really enjoyed the interaction with people outside of the Division.” After some coaxing by Harvard President Larry Summers, Ehrenreich landed his latest role as University Ombudsperson in February of 2003. The Ombuds Office, which he shares with Lydia Cummings, offers employees a safe haven to discuss conflicts or problems. Accepting the position, while impressive, is yet another example of his dedication to community outreach. During his time at Harvard, he has chaired the Science Center Executive Committee and the Committee on the Science Core Curriculum, and at one time directed the Materials Research Lab (now MRSEC). He has also served as a member of the Committee on Free Speech Guidelines and continues to work on the University’s Committee on the Environment. Teaching has always held a privileged place for Ehrenreich. “Teaching kids in Core courses who are not going to concentrate in the sciences really teaches you about how to explain science. It is a very important thing to learn.” A case in point is when he has worked on advisory boards for the departments of Defense and Energy and as a consultant to the White House Office of Technology and Policy. “When I testified before

Henry Ehrenreich, who balances a dual role as Clowes Research Professor of Science and University Ombudsperson, enjoys a rare moment of relaxation in his office.

concerned by the tough demands on today’s faculty. With research, funding, and family obligations all vying for an instructor’s time, “I feel like I lived in the age of leisure,” he jokes. “Yet what’s impressive is that, despite these deOf course, teaching has helped him mands, there are a do more than just translate science “Teaching kids in Core courses lot of people around to senators. “I rem- who are not going to concen- who do teach Core, who do serve on the ember a first-year Faculty Council, student from Boise, trate in the sciences really and who all do so Idaho, who always teaches you about how to many other things.” came to lecture Congress, I needed to look for the right expressions and the right level of discourse,” he says. “Teaching in the Core provided me with a tremendous amount of experience to do just that.” “Henry is one of early. She looked explain science.” those dynamic peolost, so I made a ple who represent what’s best about the point of being around a few minutes Division,” says Dean Venky. “He reminds before class. Perhaps not as a result us that what might first seem like an of my conversations – though maybe obligation is in fact an opportunity for they played a part – she ended up being outreach and growth. The Division, and the editor of the Harvard Crimson. Her everyone in it, must be part of the wider Harvard experience completely transworld to truly be a success.” F formed her.” Because he equally values spending time inside and outside the lab, Ehrenreich is
DEAS Spring 2004 I 15

Outside the Quad

R

elief from the cold and dark February weather came early for local Cambridge high-school teachers. A Division-backed seminar on Photovoltaics and Semiconductors provided some welcome heat, light, and electricity.

Seminars for the 2003–2004 School Year
Building Design and Energy Efficiency Photovoltaics and Solar Panels Photovoltaics and Semiconductors Fuel Cells and Electrochemistry Human Activity and the Environment

Science education advocate Eric Mazur, who holds a triple appointment as Harvard College Professor, Gordon McKay Professor of Applied Physics, and Professor of Physics at Harvard University, led the two-hour workshop. “I enjoyed giving this workshop so much!” he said. “The teachers were a great audience and the workshop was a great opportunity for me to put my own research in a broader perspective. I am excited that some of it will reach the high-school level.” Mazur’s team of faculty and graduate students shared fundamental knowledge and hands-on techniques covering the latest in semiconductors and photovoltaics (the science of converting light to electricity) with eager teachers like John Samp. “After working with high-school students all day,” Samp commented, “it’s refreshing to experience a more rigorous educational environment. The seminar is a perfect opportunity for high-school and college teaching to intersect.” The talk, one of many, is part of the National Science Foundation (NSF)– funded GK-12 program, which has paired DEAS graduate students with teachers to develop and deliver learning activities for the Cambridge Public School System. The arrival of Kathryn Hollar, the first full-time Director for Educational Programs at the Division, has energized the already top-notch outreach program. Hollar is uniquely qualified for the task: She holds a B.S. degree in Chemical Engineering and English from North Carolina State University and a Ph.D. in Chemical Engineering with a Biochemistry minor from Cornell University. Most recently she was an Assistant Professor in the Chemical Engineering

(Top) The Division’s Eric Mazur gives a thumbs up to teachers John Samp and Marion Levinstein. (Above) Maureen Haverin, the teacher liaison for the GK-12 program, shows fellow educator Manjula Subramanian how to get students charged about research.

Department at Rowan University in Glassboro, New Jersey. Jim Carey, a research assistant in the Division, shed some light on why the outreach to students is so critical. “With more and more attention being given to renewable energy sources, it is likely

that solar cells will be an important part of young students’ lives. The earlier we can give them a physical understanding of how these modern devices work, the better.” F

16 I DEAS Spring 2004

The halls of innovation
Maintaining relationships with industry is one of the main priorities of the Division. TECH, the Industrial Outreach Program (IOP), and the Harvard Industrial Partnership (HIP) all provide excellent ways for faculty and students to collaborate with research organizations. The following examples illustrate how the Division interacts with industry.

Better brushing with Unilever…
Working with a company like Unilever Research Inc., which makes and sells food, home, and personal care goods, might not seem like an obvious fit for someone doing research on microfluids, soft materials, and biophysics. But Gordon McKay Professor of Applied Physics and Professor of Physics David Weitz explains it otherwise. “Unilever is profoundly interested in exploring use of the latest technologies to make improvements in everything, from better soaps and detergents to foods that stay fresh longer,” he says. The academic-industry interface gives his lab a chance to test out ideas in the real world and to learn equally from Unilever’s research team, who might approach a problem from a completely different angle. “The results of this work are not only good science, but also have the potential for technological impact. We can then take what we discover and be proud when it results in a new or better product, while at the same time plowing our findings back into the Division,” he says. Ultimately, the collaboration helps both groups do first-rate research. “Applications can drive innovations in science just as much as research can influence the everyday products that you and I use,” concludes Weitz. F
For more information, visit www.unilever.com/brands/innovation/ www.deas.harvard.edu/projects/weitzlab/
Examples of products that rely on colloidal structures include toothpaste and food items such as milk and mayonnaise.

Research on the physics of soft condensed matter like colloids — mixtures of small particles suspended in a liquid — has led to practical applications for consumer goods companies.

Other recent industry collaborations …
The Division collaborates with a host of industries. Some of the most welldeveloped and extensive relationships include those with: Agilent Technologies Inc. Alkermes Inc. Colgate-Palmolive Infineum USA L.P. Kraft Foods Microsoft Corp. National Storage Inc. Nortel Paramitas Corp. Procter & Gamble Co. Rhodia Research Seagate Technology Weld Foundation

Select joint faculty-student inventions
Faculty and students are research and invention partners at the Division. Recent applications for patents and issued patents include: Case 2119, “Soft Output Detector for Channels with Deletions and Insertions” by Alek Kavcic (faculty) and Wei Zeng (student) Case 2254, “Improved Near Maximum Likelihood Method for Decoding LowDensity Parity-Check Codes” by Alek Kavcic (faculty), Ned Varnica (student), and Marc Fossorier (University of Hawaii faculty) Case 1135, “Process for Structural Alteration of Selected Material (‘MicroEngraving’) Within Transparent Materials” by Eric Mazur (faculty) and Eli Glezer (student) Case 1818, “A Method for Fabricating Optical Waveguides and Other Optical Devices in Three Dimensions Inside Bulk Glass Using Femtosecond Laser Oscillator” by Eric Mazur (faculty), Chris Schaffer, André Brodeur, and Rafael Gattas (students) F

DEAS Spring 2004 I 17

Alumni Notes

with Marco Iansiti
There and back again

M

arco Iansiti, ’83, Ph.D. ’88, may not have traveled far from home, but his journey from physics to business, especially radical at the time, has provided him with a fruitful intellectual adventure. He received his two degrees in Physics from Harvard before joining the faculty of the Harvard Business School (HBS) in 1989. Iansiti helped create, and initially co-chaired, the I.T. and Management program offered jointly by DEAS and HBS. What are some of your favorite memories of your time at DEAS? It was two weeks before my oral exam, and Mike Tinkham showed me his paper and said “Why don’t you work on this?” I said, “I’m not interested in research right now – I’m interested in my oral exam!” But it was fun and led to a bunch of really neat things. I [also] remember the first experiment that involved creating a new device – a new tunnel junction – and cooling it at low temperatures. I remember at 3 a.m. taking the first measurement and seeing something interesting. It felt really fascinating to see something that hadn’t been done before and get some interesting data.

Do you think that professional schools should seek out more people like you – people who have a background in physics rather than those who studied business or economics straight through? I think they should. In my own department, I’m next door to someone who studied materials science at MIT and in the next room is someone who has a degree in electrical engineering, also from MIT. I work with people who have backgrounds in organizational psychology. The management of technology is not a discipline, it’s a set of problems, and the best way to tackle that set of problems is by leveraging a variety of different disciplines. It’s too easy to get lost in the complexity of the data and I think having strong training really helps people. Are you excited about the direction in which economics is moving, in terms of behavioral economics and some of the work that has emerged? I think it is very exciting. Right now there are much more powerful tools and more powerful ways to become closer to the application world, and I

think that’s really good. It’s very hard to develop a solid understanding of a business problem if there isn’t a strong business foundation in economics or psychology or sociology or some field that can give you some structure and discipline. What are some of the research topics that you are most interested in right now? I’m very interested in networks of companies. In the old days, innovation was primarily done by individuals in isolated labs, such as Bell Labs. Now it is really a network phenomenon, where literally thousands of organizations are working on related problems. Is there a modern Bell Labs out there now? No, the concept has changed. There are a number of R&D labs that are really important in a variety of ways. You can find them in universities and you can still find them in the private sector. Looking forward, or even looking back a few years, I think you’ll find that any innovation is really a collaborative phenomenon. F

18 I DEAS Spring 2004

Quick quotes
Graduates share their thoughts and memories about their time at the Division Unexpected applications … “I ended up taking a structural engineering course that turned out to be really useful. During a summer job with a little engineering outfit, I used a lot of what I learned from the course. My claim to fame is that I wrote the first pass of a program for them – an interactive program that let you lay out the geometry of a geodesic dome – that was later used to lay out the EPCOT Center.” – David Robinson ’72 Teaming up … “I really liked the teamwork between the students in my group. We reviewed each other’s papers and gave comments and feedback. I think the relationships I developed at DEAS will last a lifetime.” – Raquel Hill Ph.D. ’02 Favorites … “My favorite class was ES 145, Bioengineering – looking at systems in the human body – because of Professor McMahon. He passed away the week before the final exam. He was a very special person; he made it a point to talk to students and get to know them outside of class.” – John Basbagill ’00 The real world … “I will always remember the story Professor Abernathy told about the John Hancock Building and the vibration problems they had been having. I probably could recite the story today, word for word, if I had to. For me, I benefited from the practice – from understanding why we were studying all these concepts and how you would apply them.” – Aldona Clottey ’95

Applying medicine … “In medical school there are a variety of topics of physiology that you can appreciate because you have studied the way fluid flows, or if you’ve studied mechanics you can appreciate the skeletal system more than if you are coming in without that background. As for orthopedics itself, it is all applied engineering. You have fractures and you have to build something that contains the fracture. That’s a lot of structural engineering.” – Jeremy Moses ’96, M.D. ’00 The past and present … “I have a lot of fond memories of being [at Harvard], coming in at night, meeting with the graduate students that were there. The new [computer science] building is 100 times better [than Aiken]. The way it lets people collaborate, the size of it. It’s wonderful to see the top work in computer science being done there.” – Bill Gates Col ’77 F

Events
In addition to almost daily seminars and colloquia – from computer science to squishy physics – the Division also sponsors major workshops. Please visit http://deas.harvard.edu/newsandevents/ for the latest details, dates, and times. Graduates are always welcome (and encouraged) to attend events. Here’s a selection of what’s to come later this year.

Privacy and Security: Technologies, Policies, and Society
Dates: April 22–23, 2004 Information & registration: www.radcliffe.edu/events/conferences/ security/

Industrial Outreach Program (IOP) 2004 Workshop
Frontiers in Materials and Nanoscience: Innovation & Collaboration Dates: May 20–21, 2004 Information & registration: www.deas.harvard.edu/industry/

Second Workshop on the Econoics of Peer-to-Peer Systems (P2P)
Dates: June 4–5, 2004 Information & registration: www.eecs.harvard.edu/p2pecon/

Harvard Industrial Partnership (HIP) 2004 Workshop
Dates: October 20–21, 2004 Planning for HIP 2004, an excellent opportunity to learn about the latest from Electrical Engineering and Computer Science, is well under way. Please stay tuned for more. F

Michael Rabin, Thomas J. Watson, Sr. Professor of Computer Science (far left, seated) and other attendees during a panel session at the HIP 2003 workshop last November.

DEAS Spring 2004 I 19

Inside information
While this issue has explored the influence of engineering and the applied sciences on society, connections to the outside world often start on the inside. As the Division has developed to occupy over 340,000 square feet of interconnected labs, classrooms, clusters, and offices, a concerted effort has been made to craft an environment (both physical and social) that encourages and inspires scientists and researchers to work together. Applied physicists work alongside chemical engineers. Fiveminute catch-up meetings, a favorite way for Dean Venky to keep current, happen in the hallways, allowing ideas to reach all interested ears. Students are free to drop by faculty offices and labs rather than make appointments. With bridges and tunnels linking old buildings with new, great ideas (and minds) have little fear of getting rained on. F
Table talk … (Below Left) Computer science faculty Margo Seltzer and Matt Welsh never table a good discussion. Turn the page … (Below Right) Camilla Lau ’04, who designed an automatic piano music page turner, goes solo at her Senior Design Project Presentation. She gave “bravos” for the support of her faculty advisors, fellow students, and friends .

Take five ... (Left) Dean Venky catches up with undergraduate Will Fithian ’06. Impromptu “hallway meetings” are just one way Division faculty, students, and staff reach out and connect. Beyond boundaries ... (Below) Without the constraints of traditional departments, Division members unite naturally to solve problems in all types of weather.

Connections

An ongoing series of photo essays dedicated to showcasing how DEAS inspires collaborative work and encourages interdisciplinary research.

Feedback loop
We welcome and appreciate your comments, suggestions, and corrections. Please send feedback to: [email protected] or call us at 617-496-3815. This newsletter is published by: The Division of Engineering and Applied Sciences Communications Office Harvard University Pierce Hall 29 Oxford Street Cambridge, MA 02138 Managing Editor/Writer: Michael Patrick Rutter Design and Production Coordinator: Eliza Grinnell This publication, including past issues, is available on the Web at www.deas.harvard.edu Copyright © 2004 by the President and Fellows of Harvard College

Face the future … Donhee Ham and Robert M. Westervelt look forward to future collaborations on novel biochip design (Above). Altered states … Howard Stone and David Weitz investigate a problem from different points of view (Above Left).

20 I DEAS Spring 2004