Aegean Associates, Inc.

From high above the Florida Panhandle, the Apalachicola Bluffs — a winding system of steep ravines — look like the branching veins of a leaf.

But this is a leaf in motion: Through the millennia, the ravines have worked their way east from the Apalachicola River, rippling through the land like furrows in a plowed field. This branching network is built on a shallow system of groundwater, with hundreds of valleys carved out over the years by streams and springs.

The Apalachicola Bluffs have been described as a living landscape, their sands constantly evolving and branching through a combination of shifting soil and flowing streams. Now researchers in MIT’s Department of Earth, Atmospheric and Planetary Sciences (EAPS) have developed a model that gauges how fast the area’s hills and valleys are spreading.

The researchers studied slopes at the very tips of the branching channels, which they refer to as “channel tips.” They found that the steeper a hill, the faster the channel tips are advancing across the landscape. The model, they say, can be used to predict the evolution of similar landscapes, such as retreating shorelines and the banks of large, meandering rivers.   

“This is a speedometer for the growth of a channel network,” says Taylor Perron, the Cecil and Ida Green Assistant Professor of Geology in EAPS. “If you measure a hill slope profile, you can figure out how fast the tips of the channels are growing.”

Perron and co-author Jennifer Hamon ’10 will publish their results in an upcoming issue of the Journal of Geophysical Research.

Perron visited the Bluffs several years ago to participate in fieldwork led by Daniel Rothman, a professor of geophysics at MIT. The group hiked through the hilly terrain, observing that at the channel tips, steep hills gave way to valleys with springs at their bases. These springs form naturally when groundwater flows at shallow depths through the sandy hills: In the deepest valleys, the groundwater emerges as a trickle or a stream.

Previous researchers have found that these springs eat away at hills, transporting sand away from a hill’s base. In response, the sand on the hillside slides down to replace the sand that’s been displaced. Over the long term, the interaction between sand and water advances the valley tips outward across the landscape.

Perron and Hamon conjectured that the relationship between hills and springs may record how fast certain channel tips grow. The team focused specifically on the shapes of hillsides, studying the behavior of sand on slopes of varying steepness, and drawing up equations representing soil flux along slopes.

Through their equations, the team found that slopes that move faster across the landscape become steeper and more sharply curved. Perron inferred from the data that steep hills are created by strong, fast-moving springs.

The group tested their theory against topographic data obtained from Rothman’s research. In previous studies that related the growth of the valley network to groundwater flow, Rothman and his group examined data from airborne laser altimetry that mapped the Apalachicola Bluffs through the valleys’ dense vegetation. Perron and Hamon applied their equations to the various hill slopes along the map, assigning a relative speed to each advancing channel tip.

In the resulting map, the group identified some interesting patterns. For instance, Perron found that most of the fastest channel tips, consistent with Rothman’s groundwater studies, belonged to the same valley network, suggesting that this network grew more recently than the others. The older valleys, he says, appear to have slowed down as they grew closer to one another and the springs at their tips began to “compete” for water — a result consistent with Rothman’s groundwater studies. The more water available, the stronger a spring flows, cutting more sand from a hill.

The researchers also observed a second pattern superimposed on the map of channel-tip speeds: Southward-growing channel tips are bounded by steeper slopes, whereas gentler slopes surround channels that grow north. They say this discrepancy may be due to the sun: Southward-moving slopes — which face north — generally stay shadier and wetter, which may spur plant growth in the hot Florida climate, stabilizing the sandy soil.

“This is exciting, because we have no general theory that predicts how sunlight should shape topography,” Perron says. “But it’s clearly important. Plants lean towards the sun, but these slopes lean away.”

The group’s analysis may be used to characterize other regions with similar landscapes. Jon Pelletier, a professor of geosciences at the University of Arizona, has used similar mathematical modeling to study the valley networks of southern Arizona.

“The approach is applicable to a great many valley heads around the world where the valleys are ‘young,’” Pelletier says. “[The MIT researchers] have obtained an elegant mathematical relationship that characterized the shape of an important class of landforms.”

Perron adds, “This would be interesting in the context of floodplains … you might be able to use the bank curvature to estimate long-term rates of the migration of large rivers.”

Believe it or not, there was a time when MIT’s newest Rhodes Scholar, Stephanie Lin, didn’t excel academically.

“In fourth grade, I was a pretty bad student,” she says. “Sort of distractible, not very motivated.

“I guess I got over that,” Lin laughs. The senior biology major, who was among 32 American students to win the Rhodes in November, is poised to pursue a career in medicine after four years at MIT packed with classes, research and extracurricular activities.

Growing up, Lin says, she leaned more toward the humanities than the sciences, considering myriad professions — librarian, lawyer, poet — throughout middle and high school. It wasn’t until a summer research experience at Michigan State University before her senior year of high school that Lin found herself drawn to the puzzles of the natural world, and the dynamic environment of the research lab.

Arriving at MIT as a freshman, Lin thought she wanted to study chemistry, but quickly switched paths again, to biology. “I took 7.012 [MIT’s introductory biology course] with Eric Lander and Rob Weinberg,” she says. “I haven’t really looked back since.”

Lin recently completed her senior thesis in the lab of Jeroen Saeij, an assistant professor of biology, where she investigated Toxoplasma gondii, an infectious parasite that serves as a model for malaria because of its similarities to the parasite behind the mosquito-borne disease. Lin’s project looked at a specific protein found within Toxoplasma, and how that protein affects which genes are turned on and off in host cells invaded by the parasite.

With the help of a team in Saeij’s lab, Lin created two versions of the parasite — one with the protein in question, the other without — and infected host cells with them, comparing how the cells’ behavior differed in the two versions.

“We found that [the presence of this protein] was potentially changing the way that host cells sense nutrients,” Lin says. “Now the hypothesis, or at least the suspicion, is that it allows a parasite to usurp more nutrients from the host cell,” which in turn allows the parasite to grow and replicate faster.

In addition to her lab-based research, Lin has a passion for putting biology into practice: She has undertaken two service trips through MIT’s Global Poverty Initiative, both to a small Mexican village where farmers were struggling to increase their crop yields. While lab work is often concerned with pioneering high-tech methods, Lin says her experience in Mexico showed her that in real-world settings, great progress can be made simply by connecting people with tried-and-true resources. In the villagers’ case, the solution involved greenhouses, fertilizers and crop rotation — relatively simple strategies, but ones previously unfamiliar to these farmers.

“We have so much more access to information just because we’re students living in a more developed place,” Lin says. “We can Google anything that we want to find; we can look for people in our area who are experts in almost any field. When you’re in a rural community you just don’t have that. There’s no Internet, and there’s very little exchange of information between the city and the countryside.”

Lin’s experiences in Mexico gave her an appreciation for the challenges facing those in developing countries — but in fact, she has seen similar issues closer to home. Since her freshman year at the Institute, she has volunteered with Health Leads Boston, an organization that helps low-income patients access the social services they need, such as food stamps, subsidized housing and utility assistance.

Working at Health Leads “made me realize how many resources there are in Boston, but also how difficult they can be to access,” Lin says, adding that her most challenging case involved a pregnant patient with rheumatoid arthritis who lived two hours from Boston Medical Center and lacked personal transportation. Lin says she would often “run out of lecture and give [the patient] a call,” to make sure that the prearranged taxi had arrived on time.

“Sometimes — often, actually — it’s just about helping someone get the simple things they need,” Lin says.

Looking forward, Lin is drawn to the study of infectious diseases, which she sees as an ideal combination of biology, research and social issues. She is considering applying her Rhodes scholarship toward a master’s degree in medical anthropology at Oxford University in the U.K. before attending medical school, focusing on what she calls the “humanistic aspects of medicine,” such as the best way to get patients to adhere to a treatment plan and factors that lead to the spread of disease.

Whichever direction Lin’s career takes her, one thing’s for sure: No longer can anyone accuse her of being unmotivated.

A team of MIT researchers has developed a way of making a high-temperature version of a kind of materials called photonic crystals, using metals such as tungsten or tantalum. The new materials — which can operate at temperatures up to 1200 degrees Celsius — could find a wide variety of applications powering portable electronic devices, spacecraft to probe deep space, and new infrared light emitters that could be used as chemical detectors and sensors.

Compared to earlier attempts to make high-temperature photonic crystals, the new approach is “higher performance, simpler, robust and amenable to inexpensive large-scale production,” says Ivan Celanovic ScD ’06, senior author of a paper describing the work in the Proceedings of the National Academy of Sciences. The paper was co-authored by MIT professors John Joannopoulos and Marin Soljačić, graduate students Yi Xiang Yeng and Walker Chen, affiliate Michael Ghebrebrhan and former postdoc Peter Bermel.

These new high-temperature, two-dimensional photonic crystals can be fabricated almost entirely using standard microfabrication techniques and existing equipment for manufacturing computer chips, says Celanovic, a research engineer at MIT’s Institute for Soldier Nanotechnologies.

While there are natural photonic crystals — such as opals, whose iridescent colors result from a layered structure with a scale comparable to wavelengths of visible light — the current work involved a nanoengineered material tailored for the infrared range. All photonic crystals have a lattice of one kind of material interspersed with open spaces or a complementary material, so that they selectively allow certain wavelengths of light to pass through while others are absorbed. When used as emitters, they can selectively radiate certain wavelengths while strongly suppressing others.

Photonic crystals that can operate at very high temperatures could open up a suite of potential applications, including devices for solar-thermal conversion or solar-chemical conversion, radioisotope-powered devices, hydrocarbon-powered generators or components to wring energy from waste heat at powerplants or industrial facilities. But there have been many obstacles to creating such materials: The high temperatures can lead to evaporation, diffusion, corrosion, cracking, melting or rapid chemical reactions of the crystals’ nanostructures. To overcome these challenges, the MIT team used computationally guided design to create a structure from high-purity tungsten, using a geometry specifically designed to avoid damage when the material is heated.

NASA has taken an interest in the research because of its potential to provide long-term power for deep-space missions that cannot rely on solar power. These missions typically use radioisotope thermal generators (RTGs), which harness the power of a small amount of radioactive material. For example, the new Curiosity rover scheduled to arrive at Mars this summer uses an RTG system; it will be able to operate continuously for many years, unlike solar-powered rovers that have to hunker down for the winter when solar power is insufficient.

Other potential applications include more efficient ways of powering portable electronic devices. Instead of batteries, these devices could run on thermophotovoltaic generators that produce electricity from heat that is chemically generated by microreactors,  from a fuel such as butane. For a given weight and size, such systems could allow these devices to run up to 10 times longer than they do with existing batteries, Celanovic says.

Shawn Lin, a professor of physics at Rensselaer Polytechnic Institute who specializes in future chip-making technology, says that research on thermal radiation at high temperatures “continues to challenge our scientific understanding of the various emission processes at sub-wavelength scales, and our technological capability.” Lin, who was not involved in this work, adds, “This particular 2-D tungsten photonic crystal is quite unique, as it is easier to fabricate and also very robust against high-temperature operation.  This photonic-crystal design should find important application in solar-thermal energy-conversion systems.”

While it’s always hard to predict how long it will take for advances in basic science to lead to commercial products, Celanovic says he and his colleagues are already working on system integration and testing applications. There could be products based on this technology in as little as two years, he says, and most likely within five years.

In addition to producing power, the same photonic crystal can be used to produce precisely tuned wavelengths of infrared light. This could enable highly accurate spectroscopic analysis of materials and lead to sensitive chemical detectors, he says.

The research was partly supported by the Army Research Office through the Institute for Soldier Nanotechnologies, NASA and an MIT Energy Initiative seed grant, as well as by TeraGrid resources and the MIT S3TEC Energy Research Frontier Center of the U.S. Department of Energy.

Within a few years, people in remote villages in the developing world may be able to make their own solar panels, at low cost, using otherwise worthless agricultural waste as their raw material.

That’s the vision of MIT researcher Andreas Mershin, whose work appears this week in the open-access journal Scientific Reports. The work is an extension of a project begun eight years ago by Shuguang Zhang, a principal research scientist and associate director at MIT’s Center for Biomedical Engineering. Zhang was senior author of the new paper along with Michael Graetzel of Switzerland’s École Polytechnique Fédérale de Lausanne.


Video: Melanie Gonick
In his original work, Zhang was able to enlist a complex of molecules known as photosystem-I (PS-I), the tiny structures within plant cells that carry out photosynthesis. Zhang and colleagues derived the PS-I from plants, stabilized it chemically and formed a layer on a glass substrate that could — like a conventional photovoltaic cell — produce an electric current when exposed to light. 

But that early system had some drawbacks. Assembling and stabilizing it required expensive chemicals and sophisticated lab equipment. What’s more, the resulting solar cell was weak: Its efficiency was several orders of magnitude too low to be of any use, meaning it had to be blasted with a high-power laser to produce any current at all.

Now Mershin says the process has been simplified to the point that virtually any lab could replicate it — including college or even high school science labs — allowing researchers around the world to start exploring the process and making further improvements. The new system’s efficiency is 10,000 times greater than in the previous version — although in converting just 0.1 percent of sunlight’s energy to electricity, it still needs to improve another tenfold or so to become useful, he says.

The key to achieving this huge improvement in efficiency, Mershin explains, was finding a way to expose much more of the PS-I complex per surface area of the device to the sun. Zhang’s earlier work simply produced a thin flat layer of the material; Mershin’s inspiration for the new advance was pine trees in a forest.

Mershin, a research scientist in the MIT Center for Bits and Atoms, noticed that while most of the pines had bare trunks and a canopy of branches only at the very top, a few had small branches all the way down the length of the trunk, capturing any sunlight that trickled down from above. He decided to create a microscopic forest on a chip, with PS-I coating his “trees” from top to bottom.

Turning that insight into a practical device took years of work, but in the end Mershin was able to create a tiny forest of zinc oxide (ZnO) nanowires as well as a sponge-like titanium dioxide (TiO2) nanostructure coated with the light-collecting material derived from bacteria. The nanowires not only served as a supporting structure for the material, but also as wires to carry the flow of electrons generated by the molecules down to the supporting layer of material, from which it could be connected to a circuit. “It’s like an electric nanoforest,” he says.

As an bonus, both zinc oxide and titanium dioxide — the main ingredient in many sunscreens — are very good at absorbing ultraviolet light. That’s helpful in this case because ultraviolet tends to damage PS-I, but in these structures that damaging light gets absorbed by the support structure.


Andreas Mershin
Photo: M. Scott Brauer
Mershin thinks that because he and his colleagues have now lowered the barrier to entry for further work on these materials, progress toward improving their efficiency should be rapid. Ultimately, once the efficiency reaches 1 or 2 percent, he says, that will be good enough to be useful, because the ingredients are so cheap and the processing so simple.

“You can use anything green, even grass clippings” as the raw material, he says — in some cases, waste that people would otherwise pay to have hauled away. While centrifuges were used to concentrate the PS-I molecules, the team has proposed a way to achieve this concentration by using inexpensive membranes for filtration. No special laboratory conditions are needed, Mershin says: “It can be very dirty and it still works, because of the way nature has designed it. Nature works in dirty environments — it’s the result of billions of experiments over billions of years.”

Because the system is so cheap and simple, he hopes this will become a “way of getting low-tech electricity to people who have never been thought of as consumers or producers of solar-power technology.” He hopes the instructions for making a solar cell will be simple enough to be reduced to “one sheet of cartoon instructions, with no words.” The only ingredient to be purchased would be chemicals to stabilize the PS-I molecules, which could be packaged inexpensively in a plastic bag.

Essentially, Mershin says, within a few years a villager in a remote, off-grid location could “take that bag, mix it with anything green and paint it on the roof” to start producing power, which could then charge cellphones or lanterns. Today, the most widely used source of lighting in such locations is kerosene lanterns — “the most expensive, most unhealthy” form of lighting there is, he says. “Nighttime illumination is the number one way to get out of poverty,” he adds, because it enables people who work in the fields all day to read at night and get an education.

Babak Parviz, an associate professor of electrical engineering at the University of Washington who specializes in bionanotechnology, says this is “a very exciting paper and a very nice step toward integrating biomolecules for building solar cells. This shows a very promising and creative first step toward building organic photovoltaic cells that can use biologically (naturally) produced cores.” He adds that while the present system still needs further development, “further work in the field can perhaps improve the stability and performance of these devices.”

The research was funded in part by an unrestricted grant from Intel Corp., and also included researchers at the University of Tennessee.

The silk that spiders use to build their webs, trap their prey and dangle from your ceiling is one of the strongest materials known. But it turns out it’s not simply the material’s exceptional strength that makes spider webs so resilient; it’s the material’s unusual combination of strength and stretchiness — silk’s characteristic way of first softening and then stiffening when pulled. These properties, scientists have found, vary depending on the forces applied, as well as on the overall design of the web.

Markus Buehler, an associate professor of civil and environmental engineering (CEE) at MIT, has previously analyzed the complex, hierarchical structure of spider silk and its amazing strength — on a pound-for-pound basis, it’s stronger than steel. Now, Buehler and his colleagues have applied their analysis to the structure of the webs themselves, finding evidence of the key properties that make webs so resilient and relating those properties back to the molecular structure of silk fibers.

The lessons learned from this work, Buehler says, could not only help develop more damage-resistant synthetic materials, but could also provide design principles that might apply to networked systems such as the Internet or the electric grid.

A paper describing the new findings is published this week in Nature. In addition to Buehler, the study was carried out by CEE graduate students Steven Cranford and Anna Tarakanova, and Nicola Pugno of the Politecnico di Torino in Italy.

It turns out that a key property of spider silk that helps make webs robust is something previously considered a weakness: the way it can stretch and soften at first when pulled, and then stiffen again as the force of the pulling increases.

This stiffening response is crucial to the way spider silk resists damage. Buehler and his team analyzed how materials with different properties, arranged in the same web pattern, respond to localized stresses. They found that materials with other responses — those that either behave as a simple linear spring as they’re pulled, or start out stretchy and then become more “plastic” — perform much less effectively.

Spider webs, it turns out, can take quite a beating without failing. Damage tends to be localized, affecting just a few threads — the place where a bug got caught in the web and flailed around, for example. This localized damage can simply be repaired, rather than replaced, or even left alone if the web continues to function as before. “Even if it has a lot of defects, the web actually still functions mechanically virtually the same way,” Buehler says. “It’s a very flaw-tolerant system.”

Buehler’s research is mostly theoretical, based on computer modeling of material properties and how they respond to stresses. But in this case, to test the findings, he and his team literally went into the field: They tested actual spider webs by poking and pulling at them. In all cases, damage was limited to the immediate area they disturbed.

The effect was somewhat surprising, Buehler says: The initial response was a deformation of the entire web, since the strands are initially relatively easy to deform. But then, because of the fibers’ nonlinear response, only the threads where the force was applied carried the load — by stretching out and then becoming stiff. As the force increased, they eventually broke.

“No matter where you pull, the web always fails exactly at that location,” Buehler says. Anyone can try this simple experiment, he adds: Simply pluck a single silk thread from a spider web, and it should break only where it’s pulled. In a web made of material with a more uniform stretching response, by contrast, local stresses cause much more widespread damage.

In a strong wind, on the other hand, it’s the initial stiffness of the silk that helps a web survive. Webs in Buehler’s simulation were able to tolerate winds up to almost hurricane strength before tearing apart.

Engineers tend to focus on materials with uniform, linear responses, Buehler says, because their properties are so much easier to calculate. But this research suggests that there could be important advantages to materials whose responses are more complex. In the unusual response of spider silk, for example — initially stiff, then stretchy, then stiff again — “each little piece of that funny behavior has a fundamental role to play” in making the whole web so robust, he says. Materials with the same ultimate strength, as measured by their breaking point, often perform very differently in real-world applications. “The actual strength is not so important, it’s how you get there,” he says.

The basic principle of permitting localized damage so that an overall structure can survive, Buehler says, could end up guiding structural engineers. For example, earthquake-resistant buildings are generally designed to protect the whole building by dissipating energy, reducing the load on the structure. When they fail, they tend to do so in their entirety.

A new design might allow the building to flex up to a point, but then certain specific structural elements could break first, allowing the rest of the structure to survive; this might ultimately allow the building to be repaired rather than demolished. Similar principles might apply to the design of airplanes or armored vehicles that could resist localized damage and keep functioning.

Such “sacrificial elements” might be used not just for physical objects but also in the design of networked systems: For example, a computer experiencing a virus attack could be designed to shut down instantly, before its problems propagate. Someday, then, the World Wide Web might actually be strengthened thanks to lessons learned from the backyard version that inspired its name.

“It’s a real opportunity,” Buehler says. “It opens a new design variable for engineering.”

David Kaplan, a professor of engineering at Tufts University and director of its Center for Biological Engineering, calls these findings “quite exciting.” He says, “The combination of modeling and experiment makes this particularly attractive as a platform for study and inquiry into materials designs and failure modes in general, with structural hierarchy in mind.”

“These principles, I believe, will have an impact in a wide range of fields such as medicine, future materials and architecture,” adds Philip LeDuc, a professor of mechanical engineering at Carnegie Mellon University.

This work was supported by the Office of Naval Research, the National Science Foundation, the Army Research Office and the MIT-Italy Program.

MIT President Susan Hockfield and about a dozen members of the MIT faculty traveled to Davos, Switzerland, last week to participate in the annual meeting of the World Economic Forum (WEF).

The five-day meeting, which brings together leaders in business, politics, academia and other areas, hosted a variety of discussions revolving around this year’s theme: “The Great Transformation: Shaping New Models.”

A troubled global economy with high unemployment dominated many of the meeting’s conversations. In a panel on the future of economics, economists including Peter Diamond, Institute Professor Emeritus and a winner of last year’s Nobel Prize in Economics, and MIT graduates Bob Shiller and Joe Stiglitz agreed that many economic models failed to recognize the nature of the current economic crisis because they largely ignored the effects of contagion and connectedness — unable to factor in the financial connections between institutions and the global risk of bankruptcy cascades. Panelists said more attention to models recognizing the role of collateral and new models, such as those incorporating behavioral economics, may offer a more accurate outlook.

MIT researchers also said that basic research in neuroscience will play a significant role in shaping societies, behavior and economic progress. MIT President Susan Hockfield spoke at Davos of the importance of basic brain research in fields other than neuroscience.

“New findings from neuroscience will have profound implications in fields far beyond the brain and mind, and well beyond psychology or medicine,” Hockfield said.

An analysis of how human nature could improve society was the topic of discussion among faculty members H. Robert Horvitz, the David H. Koch Professor of Biology and a Nobel laureate; Nancy Kanwisher, the Walter A. Rosenblith Professor of Cognitive Neuroscience; Alex “Sandy” Pentland, the Toshiba Professor of Media Arts and Sciences; and Tomaso Poggio, the Eugene McDermott Professor in the Brain Sciences and Human Behavior. One example is what the group termed “trust networks”: communities, such as open-source software users, that combine technical know-how with a certain amount of trust. Code creators, sharers and users need to trust each other in order to build on and propagate a software program. The researchers argued that understanding how trust plays into such large-scale networks may foster more successful innovations for society.

The researchers also discussed new advances in brain imaging, exploring the ever-closing gap between artificial and human intelligence. In the near future, they said, neuroscientists may be able to construct a precise model of the human brain, which could serve as a testing ground for potential therapies as well as a blueprint for artificially intelligent machines.

Yossi Sheffi, the Elisha Gray II Professor of Engineering Systems, moderated a panel on vulnerabilities in the global supply chain. Piracy, the effects of climate change, and weaknesses in cybersecurity are significant risks to the global trade of goods and services; panelists suggested that governments work together to reduce the impact of such risks.

Other MIT researchers who participated in Davos this year included Ed Boyden, the Benesse Career Development Associate Professor of Research in Education; Neil Gershenfeld, Director of the MIT Center for Bits and Atoms; Carlo Ratti, Associate Professor of the Practice in the Department of Urban Studies and Planning and Director of the SENSEable City Laboratory; Adèle Naudé Santos, Dean of the MIT School of Architecture and Planning; and Tim Berners-Lee, 3Com Founders Professor of Engineering and director of the World Wide Web Consortium (W3C).

Ed Boyden participated in a panel on leadership, in which panelists looked at the complex pressures that leaders face, including rapidly changing environments, technology, and social crises. Carlo Ratti presented a project for a new, interactive urban design in Mexico, called Ciudad Creativa Digital, with Mexican president Felipe Calderon. Adèle Naudé Santos and Neil Gershenfeld contributed to a broad discussion on innovations in social and technological models. Gershenfeld noted that smart materials and the digital revolution have made it possible for ordinary people to create new technologies, while Santos argued that the physical environment — particularly modern cities — is essential to making connections between people.

Susan Hockfield and MIT hosted three private events: a breakfast discussion about neuroscience moderated by Nature Editor in Chief Philip Campbell and featuring Bob Horvitz, Nancy Kanwisher and Tomaso Poggio; a dinner discussion about the fate of the Eurozone featuring Peter Diamond and Nouriel Roubini, a Professor of Economics and International Business at New York University’s Stern School made famous for having predicted the housing crisis; and a reception for friends of MIT.

Hockfield, who serves as a Director of the WEF, said, “The large number of MIT faculty in attendance covered a lot of academic ground, but one message they all sent to Forum attendees was this: solving the toughest problems in the world requires science, mathematics and engineering. It’s an important message in any era and particularly important in times of fiscal austerity.”

A company looking to purchase an electric-powered delivery truck today will likely experience some sticker shock: Such a vehicle costs nearly $150,000, compared to about $50,000 for the same kind of truck with a standard internal-combustion engine.

However, using electric vehicles can markedly lower the costs of a fleet of delivery trucks. That’s the conclusion of a new MIT study showing that electric vehicles are not just environmentally friendly, but also have a potential economic upside for many kinds of businesses.

The study, conducted by researchers at MIT’s Center for Transportation and Logistics (CTL), finds that electric vehicles can cost 9 to 12 percent less to operate than trucks powered by diesel engines, when used to make deliveries on an everyday basis in big cities.

“There has to be a good business case if there is going to be more adoption of electric vehicles,” says Jarrod Goentzel, director of the Renewable Energy Delivery Project at CTL and one of four co-authors of the new study. “We think it’s already a viable economic model, and as battery costs continue to drop, the case will only get better.”

Another of the paper’s co-authors, Clayton Siegert, a 2009 graduate of the CTL’s master’s of engineering in logistics program and a member of the Renewable Energy Delivery Project, presented the results in January at the IEEE Power and Energy Society Innovative Smart Grid Technologies Conference, in Washington. The paper will be published in a volume of the conference’s proceedings. It originated in a thesis project by two researchers who received the master’s of engineering in logistics from CTL in 2011, Andre De Los Rios and Kristen Nordstrom.

Electric vehicles: A staple of the truck fleet?

The CTL study was conducted using data collected by the international office supplier Staples, as well as ISO New England, the nonprofit firm that runs New England’s electric power grid. Using that data, the researchers modeled the costs for a fleet of 250 delivery trucks, and examined alternate scenarios in which the whole fleet used one of three kinds of motors: purely electric engines, hybrid gas-electric engines and conventional diesel engines.

Based on the Staples data, the researchers modeled what would happen if diesel gasoline cost $4 per gallon. Trucks with internal-combustion engines averaged 10.14 miles per gallon, compared to 11.56 miles per gallon for hybrid trucks, while the electric-only trucks averaged 0.8 kilowatt-hours per mile. Staples currently has 53 all-electric trucks, manufactured by Missouri-based Smith Electric Vehicles, in use in several American cities.

The study added one new component to the projections often made by industry fleet managers: The researchers looked at what would happen if the fleets of trucks were part of a vehicle-to-grid (V2G) system in which their batteries could be plugged into the electricity grid for 12 hours overnight, as an additional resource for providing reliable electricity to consumers. In such a setup, truck owners would be paid by utility firms for the power services they provide. V2G systems are currently being tested by multiple utility companies.

After running the numbers for various scenarios in which trucks are parked at slightly different times overnight, the MIT team found that businesses could earn roughly $900 to $1,400 per truck per year in V2G revenues in current energy markets, representing a reduction of 7 to 11 percent in vehicle operating costs. Firms would also save money on fuel, and on maintenance, because electric trucks induce less wear and tear on brakes.

All told, the operational cost per mile — the basic metric all fleet managers use — would drop from 75 cents per mile to 68 cents per mile when V2G-enabled electric trucks are substituted for internal-combustion trucks. Moreover, as Goentzel notes, “almost all these costs scale down to the individual vehicle.” Firms do not need fleets as big as 250 trucks to realize savings.

Michael Payette, director of fleet equipment at Staples, suggests that the MIT analysis corresponds with his firm’s results so far — although “it is still early in our post-deployment analysis,” he notes. In reviewing the performance of electric trucks, Payette adds, there have been “no real surprises from a reliability perspective, but I was surprised by the drivers’ acceptance, to the point where they do not ever want to drive a diesel [truck] again.”

In cities, ‘almost any truck you see is a candidate’

As Goentzel acknowledges, one limitation of the concept is that it only applies to urban truck fleets; electric vehicles do not have the range to make many kinds of rural or interstate deliveries. On the other hand, he notes, opportunities abound to use midsize electric trucks in cities.

“If you’re in an urban environment, almost any truck you see is a candidate,” Goentzel says. “If there’s a commercial truck in a city, it’s likely to be part of a fleet, whether it’s a service vehicle for a cable company, an electric utility truck, a mail package-delivery truck or part of a government fleet.”

And if the V2G concept is brought to market, commercial fleets would likely be among the first vehicles to be used, partly for logistical reasons: They would provide power resources that could be connected to the grid at regular times in the same locations.  

“The initial opportunities for V2G are likely to be for fleets, because they can be managed and controlled,” Goentzel says. Knowing that, say, Staples would have 250 trucks plugged into the grid at certain overnight hours would help utilities smooth out the flow of electricity to consumers. That delivery would be harder to manage, he notes, if it depended on individual consumers plugging their autos into the grid at more random times. “There is some work to be done before the average person is able to plug in their car and get paid by the grid,” Goentzel acknowledges.

Hepatitis C, an infectious disease that can cause inflammation and organ failure, has different effects on different people. But no one is sure why some people are very susceptible to the infection, while others are resistant.

Scientists believe that if they could study liver cells from different people in the lab, they could determine how genetic differences produce these varying responses. However, liver cells are difficult to obtain and notoriously difficult to grow in a lab dish because they tend to lose their normal structure and function when removed from the body.

Now, researchers from MIT, Rockefeller University and the Medical College of Wisconsin have come up with a way to produce liver-like cells from induced pluripotent stem cells, or iPSCs, which are made from body tissues rather than embryos; the liver-like cells can then be infected with hepatitis C. Such cells could enable scientists to study why people respond differently to the infection.

This is the first time that scientists have been able to establish an infection in cells derived from iPSCs — a feat many research teams have been trying to achieve. The new technique, described this week in the Proceedings of the National Academy of Sciences, could also eventually enable “personalized medicine”: Doctors could test the effectiveness of different drugs on tissues derived from the patient being treated, and thereby customize therapy for that patient.

The new study is a collaboration between Sangeeta Bhatia, the John and Dorothy Wilson Professor of Health Sciences and Technology and Electrical Engineering and Computer Science at MIT; Charles Rice, a professor of virology at Rockefeller; and Stephen Duncan, a professor of human and molecular genetics at the Medical College of Wisconsin.

Stem cells to liver cells

Last year, Bhatia and Rice reported that they could induce liver cells to grow outside the body by growing them on special micropatterned plates that direct their organization. These liver cells can be infected with hepatitis C, but they cannot be used to proactively study the role of genetic variation in viral responses because they come from organs that have been donated for transplantation and represent only a small population.

To make cells with more genetic variation, Bhatia and Rice decided to team up with Duncan, who had shown that he could transform iPSCs into liver-like cells.

Such iPSCs are derived from normal body cells, often skin cells. By turning on certain genes in those cells, scientists can revert them to an immature state that is identical to embryonic stem cells, which can differentiate into any cell type. Once the cells become pluripotent, they can be directed to become liver-like cells by turning on genes that control liver development.

In the current paper, MIT postdoc Robert Schwartz and graduate student Kartik Trehan took those liver-like cells and infected them with hepatitis C. To confirm that infection had occurred, the researchers engineered the viruses to secrete a light-producing protein every time they went through their life cycle.

“This is a very valuable paper because it has never been shown that viral infection is possible” in cells derived from iPSCs, says Karl-Dimiter Bissig, an assistant professor of molecular and cellular biology at Baylor College of Medicine. Bissig, who was not involved in this study, adds that the next step is to show that the cells can become infected with hepatitis C strains other than the one used in this study, which is a rare strain found in Japan. Bhatia’s team is now working toward that goal.

Genetic differences

The researchers’ ultimate goal is to take cells from patients who had unusual reactions to hepatitis C infection, transform those cells into liver cells and study their genetics to see why they responded the way they did. “Hepatitis C virus causes an unusually robust infection in some people, while others are very good at clearing it. It’s not yet known why those differences exist,” Bhatia says.

One potential explanation is genetic differences in the expression of immune molecules such as interleukin-28, a protein that has been shown to play an important role in the response to hepatitis infection. Other possible factors include cells’ expression of surface proteins that enable the virus to enter the cells, and cells’ susceptibility to having viruses take over their replication machinery and other cellular structures.

The liver-like cells produced in this study are comparable to “late fetal” liver cells, Bhatia says; the researchers are now working on generating more mature liver cells.

As a long-term goal, the researchers are aiming for personalized treatments for hepatitis patients. Bhatia says one could imagine taking cells from a patient, making iPSCs, reprogramming them into liver cells and infecting them with the same strain of hepatitis that the patient has. Doctors could then test different drugs on the cells to see which ones are best able to clear the infection.

In 2010, faculty from all five MIT schools released an assessment of diversity in the Institute’s faculty. The report of the Initiative on Faculty Race and Diversity, issued through the Office of the Provost, found that while recruitment and retention of underrepresented minority faculty had vastly improved in recent years, the “climate” around race and inclusion — the day-to-day experience of minority groups within MIT’s culture — is “distinctly and sometimes painfully different from that of their majority peers.” In particular, the report revealed a tension at MIT between diversity and excellence: a belief, whether right or wrong, that by seeking to make its faculty more diverse, MIT sacrifices its core value of academic excellence.  

Last week, faculty, students and staff gathered to examine this underlying tension during the 2012 Institute Diversity Summit. During the daylong event, participants and audience members gathered in MIT’s 10-250 lecture hall for panel discussions and breakout sessions on race, gender and cultural awareness.

President Susan Hockfield launched the summit with a video address, affirming how far MIT has come since its homogenous roots.

“For a century or more, the Institute was chiefly composed of white men from New England,” Hockfield noted. “That is not, of course, the MIT that courses through the Infinite Corridor today.”

Hockfield observed that as MIT grew in scope, it also grew in its diversity and its achievements: “I’m convinced that those trend lines have risen together for a reason. … In the complex society that we serve, our diversity is essential to our excellence.”

MIT’s makeup has changed significantly in recent decades, both in gender and race. Today, 45 percent of undergraduates are women, compared with 15 percent in 1975. Among faculty, 21 percent are women, compared with 8 percent in 1980. Underrepresented minorities (URMs) make up one-quarter of the undergraduate body, compared with 10 percent in 1987. And 7 percent of today’s faculty are URMs — up from 3 percent in 1987.

Chancellor Eric Grimson sees today’s campus as a very different environment than the one he entered in 1975. As the campus grew more diverse, Grimson said, he observed a “change in atmosphere” as more women and minority groups added new perspectives to discussions in MIT’s labs, classrooms and hallways.

“We need to appreciate how a different mode of thought brings value to a problem,” Grimson said. “Excellence is enhanced precisely because no one viewpoint is best.”

Beyond numbers

While the numbers clearly show an increasingly diverse faculty and student population, Lotte Bailyn, the T. Wilson (1953) Professor of Management Emerita, cited a missed opportunity to make the MIT environment equally supportive of all groups. Bailyn helped draft the 2010 report on diversity, and said the Initiative on Faculty Race and Diversity found that while 39 percent of MIT faculty agreed that ethnic diversity was important to excellence, this percentage was mainly made up of minority faculty and women. These same groups said they did not enjoy the same climate or opportunities as white men at the Institute.

During the summit, Wesley Harris, a professor of aeronautics and astronautics and MIT’s associate provost for faculty equity, spoke of his personal challenges growing up in the segregated South. Harris attended elementary school in Virginia, never receiving brand-new textbooks. He and his classmates used books handed down from white schools, he said, with “the word, the big word, the ‘N’ word” written in the margins by white students who knew where the textbooks were headed.

Harris was raised with the mentality that “good” was never enough. “In the segregated South, you always had to be so strong — you could always stand alone, and be better and stronger than the opposition,” Harris said.

Harris recommended a scientific approach to evaluating diversity’s role in academic excellence. For example, he suggested tallying the number of patents, papers, startups and leadership roles held by a given group, and comparing the output against other groups as an “honest assessment.”

Outside the comfort zone

In addition to faculty viewpoints, the diversity summit showcased student voices. Graduate students Alex Evans and Sophie Ni spoke on a panel with postdoc Laura Lopez and undergraduate Eric Boyer ’13 on diversity within the student body.

“I think MIT is a very diverse place with a lot of different groups,” Boyer said. “But I don’t think the groups actually interact.”

Boyer said the lack of interaction partly stems from the heavy course load for freshmen. Because incoming students are stressed about labs and exams, they may not want to break out of their “comfort zone” to get to know other groups and cultures — opting instead to spend time with people of a similar background.

Boyer said MIT should consider requiring a freshman seminar entailing interaction with students from different cultures, “before they’re cemented into their major and only hang out with their one group of friends. … [Otherwise] that’s what happens at MIT because of the stress level.”

In the summit’s keynote address, Evelynn Hammonds SM ’80, dean of Harvard College and a professor of the history of science and of African and African American studies at Harvard University, spoke of the history of racial diversity in science and engineering. She noted that while there have been “fits and starts” in encouraging diversity in the sciences, there are no coordinated, well-directed programs in many ethnic communities to cultivate aspirations in science and engineering. Hammonds pressed for a long-term commitment to the issue of diversity in science, and identified a need for more mentoring programs to motivate students who might otherwise fall through the cracks.

“I’m not talking about diversity for diversity’s sake,” Hammonds said. “I’m talking about diversity for the search for the best possible talent throughout this community and ultimately throughout the country. It’s about the search for talent, and about not leaving talented people behind for social reasons.”

The Institute Diversity Summit was co-sponsored by the Committee on Race and Diversity and the Council on Staff Diversity and Inclusion.

Many critical cell functions depend on a class of molecules called purines, which form half of the building blocks of DNA and RNA, and are a major component of the chemicals that store a cell’s energy. Cells keep tight control over their purine supply, and any disruption of that pool can have serious consequences.

In a new study, MIT biological engineers have precisely measured the effects of errors in systems for purine production and breakdown. They found that defects in enzymes that control these processes can severely alter a cell’s DNA sequences, which may explain why people who carry certain genetic variants of purine metabolic enzymes have a higher risk for some types of cancer.

DNA usually consists of a sequence of four building blocks, or nucleotides: adenine, guanine, cytosine and thymine (the A, G, C and T “letters” that make up the genetic code). Guanine and adenine are purines, and each has a close structural relative that can take its place in DNA or RNA. When these nucleotides, known as xanthine and hypoxanthine, are mistakenly inserted into DNA, they cause mutations. They can also interfere with the function of messenger RNA (mRNA), which carries DNA’s instructions to the rest of the cell, and the RNA molecules that translate mRNA into proteins.

“A cell needs to control the concentrations very carefully so that it has just the right amount of building blocks when it’s synthesizing DNA. If the cell has an imbalance in the concentrations of those nucleotides, it’s going to make a mistake,” says Peter Dedon, a professor of biological engineering at MIT and senior author of the study, which is appearing in the Proceedings of the National Academy of Sciences the week of Jan. 30.

In addition to forming the backbone of DNA and RNA, purines are also a major component of ATP, the cell’s energy currency; other molecules that manage a cell’s energy flow; and small chemical cofactors required for the activity of thousands of cell enzymes.

Abnormal metabolism

Dozens of enzymes are involved in purine metabolism, and it has long been known that malfunction of those enzymes can have adverse effects. For example, losing a purine salvage enzyme, which recovers purine nucleotides from degraded DNA and RNA, leads to high blood levels of uric acid, causing gout and kidney stones — and in extreme cases, a neurological disorder called Lesch-Nyhan syndrome. Losing another salvage enzyme produces a disease called severe combined immunodeficiency.

Abnormal purine metabolism can also lead to side effects for people taking a class of drugs called thiopurines. In some people, these drugs, often used to treat leukemia, lymphoma, Crohn’s disease, rheumatoid arthritis and organ-transplant rejection, can be metabolized into toxic compounds. Genetic testing can reveal which patients should avoid thiopurine drugs.

In the new study, Dedon and his colleagues disrupted about half a dozen purine metabolism enzymes in E. coli and yeast. After altering the enzymes, the researchers measured how much xanthine and hypoxanthine was integrated into the cells’ DNA and RNA, using a highly sensitive mass spectrometry technique they had previously developed to study DNA and RNA damage caused by inflammation.

They found that the malfunctioning enzymes could produce dramatic increases — up to 1,000-fold — in the amounts of hypoxanthine incorporated into DNA and RNA in place of adenine. However, they saw very little change in the amount of xanthine inserted in place of guanine. 

Chris Mathews, a professor emeritus of biochemistry and biophysics at Oregon State University, says the finding could help researchers better understand how defects in purine metabolism produce disease. “This paper opens the door to numerous studies — for example, looking into the biological effects resulting from the accumulation of abnormal bases in DNA and RNA,” says Mathews, who was not involved in this study.

Scientists have found quite a bit of genetic variation in purine metabolic enzymes in humans, so the research team plans to investigate the impact of those human variants on xanthine and hypoxanthine insertion into DNA. They are also interested in studying the metabolism of the other two nucleotides found in DNA, cytosine and thymine, which are pyrimidines.

As the United States seeks to reinvigorate its job market and move past economic recession, MIT News examines manufacturing’s role in the country’s economic future through this series on work at the Institute around manufacturing.

Nearly a third of companies now say that the adoption of sustainable practices has added to their profitability, according to a new MIT study — and manufacturing firms are in the vanguard.

Two-thirds of more than 2,800 companies surveyed by MIT Sloan Management Review say they have made sustainability a permanent agenda topic within their companies, up from 55 percent a year ago. And most respondents — based in 113 countries, and spanning a wide variety of sizes and industries — now see sustainability as “necessary to be competitive” in today’s economy. The study was conducted in collaboration with of the Boston Consulting Group.

“The purpose of the report was to get a high-level view of how organizations are thinking about sustainability, and what they are doing about it,” says David Kiron, executive editor of MIT Sloan Management Review and a co-author of the report. “The attention and investment we see indicate the here-to-stay nature of sustainability for organizations everywhere.”

Manufacturing companies seem to be leading the way in this new approach: The survey found a particularly strong commitment to sustainability among “resource-intensive” producers of consumer products, commodities, chemicals and automobiles, as well as in energy-related companies. Respondents said product development was enhanced by a focus on sustainability, with 25 percent of companies citing “improved innovation in products and services” as among the top benefits they derived through sustainability.

Sustainable practices help cut energy and commodity prices by reducing waste, and in some cases have transformed companies from pariahs to paragons in the eyes of environmentally aware groups.

For example, paper-products manufacturer Kimberly-Clark has moved from criticism over its cutting of old-growth forests to a top Dow Jones Sustainability World Index ranking among makers of personal products, thanks to a concerted company-wide effort to make sustainability a priority. In addition to curbing former unsustainable practices, Kimberly-Clark now aims to reach 25 percent of sales by 2015 from “environmentally innovative products” — such as a new kind of toilet paper without a cardboard tube at the center.

In this study, Kiron says, sustainability encompassed not just reductions in energy use and emissions, but also more efficient use of water and natural resources; recycling of materials and careful attention to the full life-cycle impact of products, including their ultimate disposal; and attention to human rights in the treatment of employees and suppliers.

Kiron cites Starbucks’ focus on improving the sustainability of its coffee cups — of which the chain uses billions every year. To find innovative ways of reducing the waste associated with disposable cups, the company has convened conferences at MIT in an effort to come up with more environmentally friendly approaches. “There isn’t a hard line for them between environmental and social issues,” Kiron says. “It’s all part of being socially responsible.”

The study makes clear that the more deeply ingrained sustainability is within a company’s organizational structure, the more likely it is that these practices add to the company’s bottom line. For example, companies that say they have profited from their sustainability initiatives are 50 percent more likely to have a CEO strongly committed to the programs, are twice as likely to have a separate reporting process for sustainability, and are more than twice as likely to have a chief sustainability officer.

Sometimes, internal efforts to improve sustainability can lead to new product offerings. For example, UPS improved efficiency by designing the shipping industry’s most comprehensive tracking system. That made it possible to determine the exact carbon-emissions impact of each of the millions of parcels transported every day; the company now offers customers the option of paying a bit extra for a “carbon neutral” delivery, providing carbon offsets based on the actual path and types of vehicles by which a parcel travels to its destination. The service adds about a nickel to the shipping cost per parcel.

Another example of product innovation comes from automaker BMW, which set up a sustainability team that quickly attracted some of the company’s leading engineers. They ended up designing what they say is the world’s first electric car designed for mass production from the ground up. While it may be years before the new division — dubbed “Project i” — actually contributes to the company’s profits, BMW sees it as laying the groundwork for a leadership role in new automotive technology.

Sustainability turns out to have benefits for a company’s relations with its own employees as well, the study found. “In terms of retention and recruitment, having sustainability present on your agenda really has some cachet,” Kiron says, making it easier to attract and keep some of the most talented people.

Robert Eccles, a professor of management practice at Harvard Business School who was not involved in this study, says the MIT study “is one that executives in companies in every industry all over the world should read, and it identifies many of the key issues that need to be addressed.”

Eccles also says the report’s findings are congruent with those of research he has conducted with colleagues at Harvard and at London Business School.

“The research I am doing with a number of collaborators strongly supports the findings of this survey,” he says. “We get similar results in contrasting 90 ‘high sustainability’ companies with a matched set of 90 ‘low sustainability’ ones. The high sustainability companies also have distinctly better financial performance over an 18-year period of time.”

But Eccles adds that in order to benefit corporate performance, sustainability must be paired with innovation. “Without innovation, simply committing to improve sustainability performance will likely detract from financial performance,” he notes.

Eccles cautions that challenges remain in interpreting these findings: While savings in energy, water and resource use provide obvious benefits, he says, “these are easier to demonstrate and quantify than reputational and brand benefits.” He adds, “While the article rightly notes that mainstream investors are becoming interested in sustainability, this is not yet a trend and the markets still have their traditional short-term view.”

New evidence from an ancient lunar rock suggests that the moon once harbored a long-lived dynamo — a molten, convecting core of liquid metal that generated a strong magnetic field 3.7 billion years ago. The findings, published today in Science, point to a dynamo that lasted much longer than scientists previously thought, and suggest that an alternative energy source may have powered the dynamo.

“The moon has this protracted history that’s surprising,” says co-author Benjamin Weiss, an associate professor of planetary science at MIT. “This provides evidence of a fundamentally new way of making a magnetic field in a planet a new power source.”

The new paper is the latest piece in a puzzle that planetary scientists have been working out for decades. In 1969, the Apollo 11 mission brought the first lunar rocks back to Earth — souvenirs from Neil Armstrong and Buzz Aldrin’s historic moonwalk. Since then, scientists have probed the rocky remnants for clues to the moon’s history. They soon discovered that many rocks were magnetized, which suggested that the moon was more than a cold, undifferentiated pile of space rubble. Instead, it may have harbored a convecting metallic core that produced a large magnetic field, recorded in the moon’s rocks.

Exactly what powered the dynamo remains a mystery. One possibility is that the lunar dynamo was self-sustaining, like Earth’s: As the planet has cooled, its liquid core has moved in response, sustaining the dynamo and the magnetic field it produces. In the absence of a long-lived heat supply, most planetary bodies will cool within hundreds of millions of years of formation.

A dynamo still exists within Earth because heat, produced by the radioactive decay of elements within the planet, maintains the core’s convection. Models have shown that if a lunar dynamo were powered solely by cooling of the moon’s interior, it would have been able to sustain itself only for a few hundred million years after the moon formed — dissipating by 4.2 billion years ago, at the very latest.

Heavy metal rock

However, Weiss and his colleagues found some surprising evidence in a bit of lunar basalt dubbed 10020. The Apollo 11 astronauts collected the rock at the southwestern edge of the Sea of Tranquility; scientists believe it was likely ejected from deep within the moon 100 million years ago, after a meteor impact. The group confirmed previous work dating the rock at 3.7 billion years old, and found that it was magnetized — a finding that clashes with current dynamo models.

Weiss collaborated with researchers at the University of California at Berkeley and the Berkeley Geochronology Center, who determined the rock’s age using radiometric dating. After a rock forms, a radioactive potassium isotope decays to a stable argon isotope at a known rate. The group measured the ratio of potassium to argon in a small piece of the rock, using this information to ascertain that the rock cooled from magma 3.7 billion years ago.

Weiss and graduate student Erin Shea then measured the rock’s magnetization, and found that the rock was magnetized. However, this didn’t necessarily mean that the rock, and the moon, had a dynamo-generated magnetic field 3.7 billion years ago: Subsequent impacts may have heated the rock and reset its magnetization.

To discard this possibility, the team examined whether the rock experienced any significant heating since its ejection onto the moon’s surface. Again, they looked to isotopes of potassium and argon, finding that the only heating the rock had experienced since it was ejected onto the lunar surface came from simple exposure to the sun’s rays.

“It’s basically been in cold storage for 3.7 billion years, essentially undisturbed,” Weiss says. “It retains a beautiful magnetization record.”

Stirring things up

Weiss says the rock’s evidence supports a new mechanism of dynamo generation that was proposed last year by scientists at University of California at Santa Cruz (UCSC). This hypothesis posits that the moon’s dynamo may have been powered by Earth’s gravitational pull. Billions of years ago, the moon was much closer to Earth than it is today; terrestrial gravity may have had a stirring effect within the moon’s core, keeping the liquid metal moving even after the lunar body had cooled.

Francis Nimmo, a professor of earth and planetary sciences at UCSC and one of the researchers who originally put forth the new dynamo theory, says Weiss’ evidence provides scientists with a new picture of the moon’s evolution.

“We generally assume that cooling is the main mechanism for driving a dynamo anywhere,” says Nimmo, who was not involved in the current study. “This lunar data is telling us that other mechanisms may also play a role, not just at the moon, but elsewhere, too.”

MIT researchers have discovered that certain photosynthetic ocean bacteria should beware of viruses bearing gifts: These viruses are carrying genetic material taken from their previous bacterial hosts that tricks the new host into using its own machinery to activate the genes, a process never before documented in any virus-bacteria relationship.

The con occurs when a virus injects its DNA into a bacterium living in a phosphorus-starved region of the ocean. Such bacteria, stressed by the lack of phosphorus — which they use as a nutrient — have their phosphorus-gathering machinery in high gear. The virus senses the host’s stress and offers what seems like a helping hand: bacterial genes nearly identical to the host’s own that enable the host to gather more phosphorus. The host uses those genes — but the additional phosphorus goes primarily toward supporting the virus’s replication of its own DNA.

Once that process is complete, about 10 hours after infection, the virus explodes its host, releasing progeny viruses back into the ocean where they can invade other bacteria and repeat this process. The additional phosphorus-gathering genes provided by the virus keep its reproduction cycle on schedule.

In essence, the virus, or phage, is co-opting a very sophisticated component of the host’s regulatory machinery to enhance its own reproduction — something never before documented in a virus-bacteria relationship.

“This is the first demonstration of a virus of any kind — even those heavily studied in biomedical research — exploiting this kind of regulatory machinery in a host cell, and it has evolved in response to the extreme selection pressures of phosphorus limitation in many parts of the global oceans,” says Sallie “Penny” W. Chisholm, a professor of civil and environmental engineering (CEE) and biology at MIT, who is principal investigator of the research and co-author of a paper published in the Jan. 24 issue of Current Biology. “The phages have evolved the capability to sense the degree of phosphorus stress in the host they’re infecting and have captured, over evolutionary time, some components of the bacteria’s machinery to overcome the limitation.”

Chisholm and co-author Qinglu Zeng, a CEE postdoc, performed this research using the bacterium Prochlorococcus and its close relative, Synechococcus, which together produce about one-sixth of the oxygen in Earth’s atmosphere. Prochlorococcus is about one micron in diameter and can reach densities of up to 100 million per liter of seawater; Synechococcus is only slightly larger and a bit less abundant. The viruses that attack both bacteria, called cyanophages, are even more populous.

The bacterial mechanism in play is called a two-component regulatory system, which refers to the microbe’s ability to sense and respond to external environmental conditions. This system prompts the bacteria to produce extra proteins that bind to phosphorus and bring it into the cell. The gene carried by the virus encodes this same protein.

“Both the phage and bacterial host have the genes that produce the phosphorus-binding proteins, and we found they can both be up-regulated by the host’s two-component regulatory system,” Zeng says. “The positive side of infection for bacteria is that they will obtain more phosphorus binders from the phage and maybe more phosphorus, although the bacteria are dying and the phage is actually using the phosphorus for its own ends.”

In 2010, Chisholm and Maureen Coleman, now an assistant professor at the University of Chicago, demonstrated that the populations of Prochlorococcus living in the Atlantic Ocean had adapted to the phosphorus limitations of that environment by developing more genes specifically related to the scavenging of phosphorus. This proved to be the sole difference between those populations and their counterparts living in the Pacific Ocean, which is richer in phosphorus, indicating that the variation is the result of evolutionary adaptation to the environment.

The new research indicates that the phage that infect these bacteria have evolved right along with their hosts.

“These viruses ... have acquired genes for a metabolic pathway from their host cells,” says David Shub, a professor of biological sciences at the State University of New York at Albany who was not involved in this research. “Now Zeng and Chisholm have shown that these particular viral genes are regulated by the amount of phosphate in their environment, and also that they use the regulatory proteins already present in their host cells at the time of infection. The significance of this paper is the revelation of a very close evolutionary interrelationship between this particular bacterium and the viruses that seek to destroy it.”

“We’ve come to think of this whole system as another bit of evidence for the incredible intimacy of the relationship of phage and host,” says Chisholm, whose next steps are to explore the functions of all the genes these marine phages have acquired from host cells to learn more about the selective pressures affecting the phage-host interactions in the open oceans. “Most of what we understand about phage and bacteria has come from model microorganisms used in biomedical research. The environment of the human body is dramatically different from that of the open oceans, and these oceanic phage have much to teach us about fundamental biological processes.” 

This research was supported in part by the Gordon and Betty Moore Foundation, the National Science Foundation’s CMORE program and Biological Oceanography program, and the U.S. Department of Energy.

Jeffrey Grossman says Cambridge has a better climate than California — for carrying out materials science research, that is. That’s why Grossman decided, two years ago, to make the move from the University of California at Berkeley to a position at MIT.

“I really don’t like sunshine that much,” he says, in a wry tone that clearly suggests otherwise: It was the innovative climate inside the halls of MIT, not the frigid weather outside, that drew him here.


Jeffrey Grossman
Photo: M. Scott Brauer
Grossman hadn’t been to Cambridge prior to the move, but there was some family history here: His father, Neal, now an associate professor of philosophy emeritus at the University of Illinois at Chicago, is an MIT alumnus (SB ’63). “I have his slide rule,” Grossman says, “that was given to him by his uncle, who was also an MIT alum.”

Growing up in suburban Evanston, Ill., Grossman says he decided very early on that he wanted to study physics when he got to college. “It was those mysteries that physics could answer that attracted me,” he recalls. “This was very exciting to me.”

He pursued that ambition, majoring in physics at Johns Hopkins University, and then continuing his studies at the University of Illinois at Urbana-Champaign, where he earned his doctorate in computational physics. He accepted a postdoc position in physics at Berkeley, “and I knew I wanted to move more toward applied physics,” he says. “I love physics, but what excites me the most are those intersections between engineering and science.” He soon decided that materials science was exactly the right field in which to pursue his research vision.

Grossman also decided that energy, in particular, presented “the perfect combination of a global-scale problem that we face, that happens to need — at its core — new materials,” he says. “And we’re also at a point where we can predict and make new materials as never before. I don’t know if we’ve ever had a challenge this big that faces us, where the answer really is in the design of new materials.”

Putting a move on

Although he and his wife, Katherine Moschandreas, graduated from Evanston Township High School together, Grossman says, they never knew each other at the school. It was only at their 10th reunion that they finally met: She was living in Cambridge (having earned two master’s degrees from Harvard University) but was about to move to Berkeley, where he was then a postdoc. They began dating soon after she moved, were married soon after that, and now have three children (ages 8, 7 and 3).

Grossman says that as a graduate student and postdoc, he found it important to have outside interests that could provide a different kind of focus. For him, that came in the form of competitive ballroom dancing, which he pursued seriously throughout those years.

“When you go dancing and look around, everyone is smiling. No one dances and is not having fun,” he says. It’s a good outlet, he adds, because “it’s something that’s not work, but it requires quite a bit of concentration. That kind of balance between my work and other interests actually allows me to be more productive and more creative. But a balance doesn’t come easy — you have to work to maintain it.”

Shaping molecules, and solar panels


Photo: M. Scott Brauer In 2009, Berkeley offered him a faculty appointment — and so did MIT. Much as he enjoyed his work in California, he says, “Professionally, I had never seen a place like MIT, where people are so problem-driven. It’s not just materials scientists — in all the departments, everyone I meet, everyone is cross-disciplinary, working on both the practice and the theory. This is where people, and I mean everyone — from the undergraduate students to graduate students, postdocs, staff scientists and faculty — are the most driven I’ve ever seen to solve hard problems together.”

So it became an easy choice for him, and he joined MIT in fall 2009. “It’s a thrill to be here, because of that culture and that environment. … If you want to try to change the world and make it better with science and engineering, this is just a really unique place,” he says.

Grossman, who was awarded tenure this year and is now the Carl Richard Soderberg Associate Professor of Power Engineering, has continued his research on developing and applying computational techniques to understand and design new materials at the scale of atoms and electrons, mostly for energy applications. For example, he and one of his postdocs recently showed how a particular material can collect solar energy in a form that can be stored indefinitely by changing the molecule’s configuration, and then release that energy on demand by snapping back to its original form in response to a small nudge such as an increase in temperature.

Grossman has also created a balance between his group’s computational predictions and practical demonstrations of those ideas. For example, he leads a new team at MIT working to make efficient, quantum-dot-based solar cells. He and his students are also arranging solar panels in three-dimensional shapes to increase the power produced over the course of a day and smooth out its variability, and are in the process of building full-scale prototypes of such systems on a rooftop at MIT.

Grossman’s work extends beyond new energy technologies into other areas, such as designing environmentally sustainable concrete and producing clean water. “One of the advantages of the kinds of simulations we do is that the same methods can be used to impact diverse technological challenges,” Grossman explains. “For example, since we study how electrons move through materials, and how this controls key properties that limit the material’s performance, we can tackle old problems … in new ways.”

Whether simulating new materials solutions or fostering collaborations among his MIT students and colleagues, it seems that moving from California to Cambridge has sharpened Grossman’s resolve to make the most out of every precious hour of sunlight.


Video: Melanie Gonick

On Monday, high school students from across the country assembled in a lecture hall at MIT, patiently awaiting a call from NASA.

For four months, these students worked in teams as part of MIT’s Zero Robotics Challenge, a competition in which high school students program small robots to fly aboard the International Space Station (ISS). The robots, named SPHERES, were originally conceived and built by students in MIT’s Space Systems Laboratory.

These robots — roughly the size and shape of a basketball — run on compressed gas, and can be programmed to spin, revolve, hover and navigate through the air. In 2006, astronauts brought several of them aboard the ISS; a few years later, astronaut Greg Chamitoff PhD ’92 helped launch the Zero Robotics Challenge, making the robots accessible to high school students.

Chamitoff was on hand Monday, along with several colleagues who served as mentors during the challenge: former astronauts Leland Melvin, John Grunsfeld ’80 and Jeff Hoffman, now a professor of aeronautics and astronautics at MIT.

For this year’s challenge, students were given a “mission” to program robots to look for, mine and return alternative energy from fictitious asteroids in deep space. They were given coordinates for virtual asteroids located within the ISS and then had to develop computer codes to make a robot perform using various strategies, each of which earned a certain number of points.

For the past few months, student teams have been testing their codes in computer simulations, maneuvering virtual robots and competing against other teams in online games. The finalists — 38 teams with the best simulation scores — assembled at MIT on Monday for a chance to watch their codes play out in real robots on the ISS, 250 miles above Earth.

The students gathered in MIT’s 10-250 lecture hall. Many sported uniforms, including one team clad in NASA’s “pumpkin suit” orange. Once NASA successfully connected MIT with the ISS, a large screen at the front of the hall projected a live view from inside the station — and 10-250 erupted in cheers. Astronaut Don Pettit, onboard the ISS, responded, “We hear you loud and clear, MIT.”

Pettit and his colleague Andre Kuipers served as referees during the game, monitoring the robots, keeping score and occasionally pausing the game to refuel: Several times, the robots literally ran out of gas, and the astronauts rummaged through the ISS module’s storage bins for extra tanks of carbon dioxide.

Despite a few glitches in the video feed, most teams were able to watch their algorithms in action. During each match, both a red and a blue SPHERE began their missions simultaneously, slowly circling each other and moving about the module according to their preprogrammed trajectories.

“The coolest thing is not just that they’re controlling these robots onboard the ISS,” Melvin said. “In the future they may be developing algorithms to do inspections on the outside of the space station, or they may program robots to help vehicles dock. These are some really good skills that will help us down the road with NASA’s future missions.”

During the championship round, the video connection blinked out, leaving only a live audio feed. The astronauts continued running the final teams’ codes, and Pettit offered to narrate each round. 

“They’re stabilizing at 50 centimeters and closing,” Pettit reported to the earthbound crowd. “Are they going to crash? They just missed each other! It looks like they’ve completed … and we have a score!”

The winning team — dubbed “Team Rocket” — comprised representatives of three high schools in Maryland, New Jersey and Florida. The team’s mentor, astronaut John Grunsfeld, said the win was a surprise for the students, many of whom had never worked with this particular programming language before this challenge. 

“This is just fantastic to see all the students from all the schools and how well they did,” Grunsfeld said. “Hopefully some of them will come to MIT.”

As the United States seeks to reinvigorate its job market and move past economic recession, MIT News examines manufacturing’s role in America’s economic future through this series on work at the Institute around manufacturing.

Not long ago, MIT political scientist Suzanne Berger was visiting a factory in western Massachusetts, a place that produces the plastic jugs you find in grocery stores. As she saw on the factory floor, the company has developed an innovative automation system that has increased its business: Between 2004 and 2008, its revenues doubled, and its workforce did, too. Moreover, the firm has found a logical niche: Since plastic jugs are both bulky and inexpensive, it’s not economical to produce them overseas and ship them to the United States, simply to fill them with, say, milk or syrup.

“Is this just an odd little story?” Berger asks. “Actually, no.” While the decline of American manufacturing has been widely trumpeted — manufacturing jobs in the United States have dropped from 20 million in 1979 to about 12 million today — conglomerates such as Procter & Gamble and high-tech firms such as Dow Corning have kept significant amounts of manufacturing in the country. Moreover, 3,500 manufacturing companies across the United States — not just the jug-making firm in Massachusetts — doubled their revenues between 2004 and 2008. With that in mind, Berger asks, “How can we imagine enabling these firms to branch out into more innovative activities as well?”

That is the kind of problem Berger and 19 of her faculty colleagues at MIT are now studying as part of a two-year Institute-wide research project called Production in the Innovation Economy (PIE), which is focused on renewing American manufacturing. The guiding premise of PIE is that the United States still produces a great deal of promising basic research and technological innovation; what is needed is a better sense of how to translate those advances into economic growth and new jobs.

As Berger puts it, “The single most important question in the study is: What kind of manufacturing do we need in order to get full value out of our innovation strengths?”

That question is currently at the forefront of MIT’s concerns as well. Institute President Susan Hockfield is serving as a co-chair of the steering committee of President Barack Obama’s Advanced Manufacturing Partnership (AMP), which in June will give policy recommendations to the White House about renewing American manufacturing. PIE is not a subset of AMP, but arises from similar concerns about applying technology in the national interest. 

• Video: Watch the recorded webcast from the AMP regional meeting held at MIT

Made in America in the 21st century

In the course of conducting its research, PIE will issue an interim report later this spring; publish a final report in 2013; create a film on manufacturing; host a lecture series; and issue a working-paper series of research findings from the professors on the team.

The co-chairs of the PIE Commission are Berger, the Raphael Dorman-Helen Starbuck Professor of Political Science, and biologist and Institute Professor Phillip Sharp. Olivier de Weck, who has served as the associate head of the Engineering Systems Division and also holds a dual appointment with the Department of Aeronautics and Astronautics, is serving as PIE’s executive director.

To a significant extent, PIE is modeled on the MIT Commission on Industrial Productivity, a 1980s group that conducted a similarly long-term study on the American economy and wrote the highly influential book Made in America. Published in 1989, Made in America has sold more than 300,000 copies and influenced public discussion about changes needed to improve America’s industrial productivity, such as greater flexibility in production processes and policies to help firms make capital investments.

However, PIE’s research interests differ from those of the Made in America group in substantial ways. The 1980s research project was organized around the performance of U.S. firms in several major industries then experiencing intensified competition, from automakers to consumer electronics companies. PIE focuses on specific questions that may cut across a multitude of industrial sectors, and has organized its work into eight distinct “modules” that cover a diverse set of issues, ranging from the challenges of scaling up small startups to the problems of training workers.

Like IT or not?

In so doing, PIE is also broadly scrutinizing a common assumption of the last quarter-century: that the information technology industry is the basic paradigm for innovation-based manufacturing in the United States. “Some people think we can just do the innovation, and then license and sell and outsource it,” Berger notes. By contrast, she says, “those of us in the PIE study think it’s an open question whether a similar model works elsewhere, particularly in the new emerging-technology areas.”

Information technology companies often have low startup costs covered by venture capital, and their production tasks lend themselves to being handled overseas. But in other areas with advanced-manufacturing potential, such as energy, advanced materials or biotechnology, “you’re going to need far heavier capital investment,” Berger says. It’s not obvious how such companies can best finance the development and commercialization of their products.

One of the PIE modules will also examine the effects of manufacturing — and the loss of manufacturing jobs — on other industries. Manufacturing is widely viewed as an industry that creates additional jobs besides those on the production lines; factories create a need for additional service-industry workers. Additionally, the income earned by manufacturing workers creates demand for still more goods and services.

The factory visits that Berger and her colleagues have been making for PIE underscore that point. On a recent visit to a company that makes equipment pipes and tanks for biotechnology companies, she found that a quarter of the company’s revenue comes from repairing and servicing the equipment. “What we’re discovering is that this connection between manufacturing and services is an integral one,” Berger says. “A set of capabilities is gained in making products that then get redeployed in the service part of a business.”

Ultimately, the PIE researchers may have many more such discoveries ahead of them — and may need them, to help chart the possible paths for new success in manufacturing.

The design of aromas — the flavors of packaged food and drink and the scents of cleaning products, toiletries and other household items — is a multibillion-dollar business. The big flavor companies spend tens of millions of dollars every year on research and development, including a lot of consumer testing.

But making sense of taste-test results is difficult. Subjects’ preferences can vary so widely that no clear consensus may emerge. Collecting enough data about each subject would allow flavor companies to filter out some of the inconsistencies, but after about 40 flavor samples, subjects tend to suffer “smell fatigue,” and their discriminations become unreliable. So companies are stuck making decisions on the basis of too little data, much of it contradictory.

One of the biggest flavor companies in the world has turned to researchers in MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) for help. To analyze taste-test results, the CSAIL researchers are using genetic programming, in which mathematical models compete with each other to fit the available data and then cross-pollinate to produce models that are more accurate still.

The Swiss flavor company Givaudan asked CSAIL principal research scientist Una-May O’Reilly, postdoc Kalyan Veeramachaneni and the University of Antwerp’s Ekaterina Vladislavleva to help interpret the results of tests in which 69 subjects evaluated 36 different combinations of seven basic flavors, assigning each a score according to its olfactory appeal.

For each subject, O’Reilly and her colleagues randomly generate mathematical functions that predict scores according to the concentrations of different flavors. Each function is assessed according to two criteria: accuracy and simplicity. A function that, for example, predicts a subject’s preferences fairly accurately using a single factor — say, concentration of butter — could prove more useful than one that yields a slightly more accurate prediction but requires a complicated mathematical manipulation of all seven variables.

After all the functions have been assessed, those that provide poor predictions are winnowed out. Elements of the survivors are randomly recombined to produce a new generation of functions; those are then evaluated for accuracy and simplicity. The whole process is repeated about 30 times, until it converges on a set of functions that accord well with the preferences of a single subject.

Because O’Reilly and her colleagues’ method produces profiles of individual test subjects’ tastes, it can sort them into distinct groups. It could be, for instance, that test subjects tend to have strong preferences for either cinnamon or nutmeg but not both. By marketing one product to cinnamon lovers and another to nutmeg lovers, a company could do much better than by marketing one product to both. “For every one of these 36 flavors, someone hated it and someone liked it,” O’Reilly says. “If you try to identify a flavor that the whole panel likes, you end up settling for a little bit less.”

O’Reilly and her colleagues haven’t had an opportunity to empirically determine whether their models correctly predict subjects’ responses to new flavors. So to try to establish their model’s accuracy, they instead built another model. First, they developed a set of mathematical functions that represent subjects’ true taste preferences. Then they showed that, given the limitations of particular test designs, their algorithms could still divine those preferences. Although they developed the model purely to validate their approach, O’Reilly says, flavor researchers were intrigued by the possibility of using it to develop more accurate and efficient test protocols.

“People have been playing with these [evolutionary] techniques for decades,” says Lee Spector, a professor of computer science at Hampshire College and editor-in-chief of the journal Genetic Programming and Evolvable Machines, where the MIT researchers’ latest paper appears. “One of the reasons that they haven’t made a big splash until recently is that people haven’t really figured out, I think, where they can pay off big.” Taste preference, Spector says, “is a pretty brilliant area in which to apply the evolutionary methods — and it looks as though they’re working, also, so that’s exciting.”

MIT postdoc Emile Bruneau has long been drawn to conflict — not as a participant, but an observer. In 1994, while doing volunteer work in South Africa, he witnessed firsthand the turmoil surrounding the fall of apartheid; during a 2001 trip to visit friends in Sri Lanka, he found himself in the midst of the violent conflict between the Tamil Tigers and the Sri Lankan military.

Those chance experiences got Bruneau, who taught high school science for several years, interested in the psychology of human conflict. While teaching, he also volunteered as counselor for a conflict-resolution camp in Ireland that brought Catholic and Protestant children together. At MIT, Bruneau is now working with associate professor of cognitive neuroscience Rebecca Saxe to figure out why empathy — the ability to feel compassion for another person’s suffering — often fails between members of opposing conflict groups.

“What are the psychological barriers that are put up between us in these contexts of intergroup conflict, and then, critically, what can we do to get past them?” Bruneau asks.

Bruneau and Saxe are also trying to locate patterns of brain activity that correlate with empathy, in hopes of eventually using such measures to determine how well people respond to reconciliation programs aimed at boosting empathy between groups in conflict.

“We’re interested in how people think about their enemies, and whether there are brain measures that are reliable readouts of that,” says Saxe, who is an associate member of MIT’s McGovern Institute for Brain Research. “This is a huge vision, of which we are at the very beginning.”

Before researchers can use tools such as magnetic resonance imaging (MRI) to evaluate whether conflict-resolution programs are having any effect, they need to identify brain regions that respond to other people’s emotional suffering. In a study published Dec. 1 in Neuropsychologia, Saxe and Bruneau scanned people’s brains as they read stories in which the protagonist experienced either physical or emotional pain. The brain regions that responded uniquely to emotional suffering overlapped with areas known to be involved in the ability to perceive what another person is thinking or feeling.

Failures of empathy

Hoping to see a correlation between empathy levels and amount of activity in those brain regions, the researchers then recruited Israelis and Arabs for a study in which subjects read stories about the suffering of members of their own groups or that of conflict-group members. The study participants also read stories about a distant, neutral group — South Americans.

As expected, Israelis and Arabs reported feeling much more compassion in response to the suffering of their own group members than that of members of the conflict group. However, the brain scans revealed something surprising: Brain activity in the areas that respond to emotional pain was identical when reading about suffering by one’s own group or the conflict group. Also, those activity levels were lower when Arabs or Israelis read about the suffering of South Americans, even though Arabs and Israelis expressed more compassion for South Americans’ suffering than for that of the conflict group.

Those findings, published Jan. 23 in Philosophical Transactions of the Royal Society: Biological Sciences, suggest that those brain regions are sensitive to the importance of the opposing group, not whether or not you like them.

Joan Chiao, an assistant professor of psychology at Northwestern University, says those brain regions may be acting as a “thermometer” for conflict. “It’s a really fascinating study because it’s the first to examine the neural basis of people’s behavior in longstanding conflicts, as opposed to groups that are distant and don’t have a long history of intergroup strife,” says Chiao, who was not involved in the research.

However, because the study did not reveal any correlation between the expression of empathy and the amount of brain activity, more study is needed before MRI can be used as a reliable measure of empathy levels, Saxe says.

“We thought there might be brain regions where the amount of activity was just a simple function of the amount of empathy that you experience,” Saxe says. “Since that’s not what we found, we don’t know what the amount of activity in these brain regions really means yet. This is basically a first baby step, and one of the things it tells us is that we don’t know enough about these brain regions to use them in the ways that we want to.”

Bruneau is now testing whether these brain regions send messages to different parts of the brain depending on whether the person is feeling empathy or not. He hypothesizes that when someone reads about the suffering of an in-group member, the brain regions identified in this study send information to areas that process unpleasant emotions, while stories about suffering of a conflict-group member activate an area called the ventral striatum, which has been implicated in schadenfreude — taking pleasure in the suffering of others.

The northern goshawk is one of nature’s diehard thrill-seekers. The formidable raptor preys on birds and small mammals, speeding through tree canopies and underbrush to catch its quarry. With reflexes that rival a fighter pilot’s, the goshawk zips through a forest at high speeds, constantly adjusting its flight path to keep from colliding with trees and other obstacles.

While speed is a goshawk’s greatest asset, researchers at MIT say the bird must observe a theoretical speed limit if it wants to avoid a crash. The researchers found that, given a certain density of obstacles, there exists a speed below which a bird — and any other flying object — has a fair chance of flying collision-free. Any faster, and a bird or aircraft is sure to smack into something, no matter how much information it has about its environment. A paper detailing the results has been accepted to the IEEE Conference on Robotics and Automation.

• Video: Watch a hawk fly through a dense forest

These findings may not be news to the avian world, but Emilio Frazzoli, an associate professor of aeronautics and astronautics at MIT, says knowing how fast to fly can help engineers program unmanned aerial vehicles (UAVs) to fly at high speeds through cluttered environments such as forests and urban canyons.

Frazzoli is part of an interdisciplinary team that includes biologists at Harvard University, who are observing flying behaviors in goshawks and other birds, and roboticists at MIT, who are engineering birdlike UAVs. With Frazzoli’s mathematical contributions, the team hopes to build fast, agile UAVs that can move through cluttered environments — much like a goshawk streaking through the forest.

Speedy intuition

Most UAVs today fly at relatively slow speeds, particularly if navigating around obstacles. That’s mainly by design: Engineers program a drone to fly just fast enough to be able to stop within the field of view of its sensors.

“If I can only see up to five meters, I can only go up to a speed that allows me to stop within five meters,” Frazzoli says. “Which is not very fast.”

If the northern goshawk flew at speeds purely based on what it could immediately see, Frazzoli conjectures that the bird would not fly as fast. Instead, the goshawk likely gauges the density of trees, and speeds past obstacles, knowing intuitively that, given a certain forest density, it can always find an opening through the trees.

Frazzoli points out that a similar intuition exists in downhill skiing.

“When you go skiing off the path, you don’t ski in a way that you can always stop before the first tree you see,” Frazzoli says. “You ski and you see an opening, and then you trust that once you go there, you’ll be able to see another opening and keep going.”

Frazzoli says that in a way, robots may be programmed with this same speedy intuition. Given some general information about the density of obstacles in a given environment, a robot could conceivably determine the maximum speed below at it can safely fly.

Forever flying

Toward this end, Frazzoli and PhD student Sertac Karaman developed mathematical models of various forest densities, calculating the maximum speed possible in each obstacle-filled environment.

The researchers first drew up a differential equation to represent the position of a bird in a given location at a given speed. They then worked out what’s called an ergodic model representing a statistical distribution of trees in the forest — similar to those commonly used by ecologists to characterize the density of a forest. In an ergodic forest, while the size, shape and spacing of individual trees may vary, their distribution in any given area is the same as any other area. Such models are thought to be a fair representation of most forests in the world.

Frazzoli and Karaman adjusted the model to represent varying densities of trees, and calculated the probability that a bird would collide with a tree while flying at a certain speed. The team found that, for any given forest density, there exists a critical speed above which there is no “infinite collision-free trajectory.” In other words, the bird is sure to crash. Below this speed, a bird has a good chance of flying without incident.

“If I fly slower than that critical speed, then there is a fair possibility that I will actually be able to fly forever, always avoiding the trees,” Frazzoli says.

The team’s work establishes a theoretical speed limit for any given obstacle-filled environment. For UAVs, this means that no matter how good robots get at sensing and reacting to their environments, there will always be a maximum speed they will need to observe to ensure survival.

Steven LaValle, professor of computer science at the University of Illinois at Urbana-Champaign, says knowing where to cap a UAV’s speed can help engineers like himself design more agile robots.

“Rather than trying to optimize robot speed, we might be able to [design] the robot at 95 percent of that speed, and achieve must simpler strategies that are also much safer to execute,” says LaValle, who did not contribute to the research.

The researchers are now seeing if the theory bears out in nature. Frazzoli is collaborating with scientists at Harvard, who are observing how birds fly through cluttered environments — in particular, whether a bird will choose not to fly through an environment that is too dense. The team is comparing the birds’ behavior with what Frazzoli’s model can predict. So far, Frazzoli says preliminary results in pigeons are “very encouraging.”

In the coming months, Frazzoli also wants to see how close humans can come to such theoretical speed limits. He and his students are developing a first-person flying game to test how well people can navigate through a simulated forest at high speeds.

“What we want to do is have people play, and we’ll just collect statistics,” Frazzoli says. “And the question is, how close to the theoretical limit can we get?”

Why did language evolve? While the answer might seem obvious — as a way for individuals to exchange information — linguists and other students of communication have debated this question for years. Many prominent linguists, including MIT’s Noam Chomsky, have argued that language is, in fact, poorly designed for communication. Such a use, they say, is merely a byproduct of a system that probably evolved for other reasons — perhaps for structuring our own private thoughts.

As evidence, these linguists point to the existence of ambiguity: In a system optimized for conveying information between a speaker and a listener, they argue, each word would have just one meaning, eliminating any chance of confusion or misunderstanding. Now, a group of MIT cognitive scientists has turned this idea on its head. In a new theory, they claim that ambiguity actually makes language more efficient, by allowing for the reuse of short, efficient sounds that listeners can easily disambiguate with the help of context.

“Various people have said that ambiguity is a problem for communication,” says Ted Gibson, an MIT professor of cognitive science and senior author of a paper describing the research to appear in the journal Cognition. "But the fact that context disambiguates has important ramifications for the re-use of potentially ambiguous forms. Ambiguity is no longer a problem — it's something that you can take advantage of, because you can reuse easy [words] in different contexts over and over again."

Lead author of the paper is Steven Piantadosi PhD ’11; Harry Tily, a postdoc in the Department of Brain and Cognitive Sciences, is another co-author.

What do you ‘mean’?

For a somewhat ironic example of ambiguity, consider the word “mean.” It can mean, of course, to indicate or signify, but it can also refer to an intention or purpose (“I meant to go to the store”); something offensive or nasty; or the mathematical average of a set of numbers. Adding an ‘s’ introduces even more potential definitions: an instrument or method (“a means to an end”), or financial resources (“to live within one’s means”).

But virtually no speaker of English gets confused when he or she hears the word “mean.” That’s because the different senses of the word occur in such different contexts as to allow listeners to infer its meaning nearly automatically.

Given the disambiguating power of context, the researchers hypothesized that languages might harness ambiguity to reuse words — most likely, the easiest words for language processing systems. Building on observation and previous studies, they posited that words with fewer syllables, high frequency and the simplest pronunciations should have the most meanings.

To test this prediction, Piantadosi, Tily and Gibson carried out corpus studies of English, Dutch and German. (In linguistics, a corpus is a large body of samples of language as it is used naturally, which can be used to search for word frequencies or patterns.) By comparing certain properties of words to their numbers of meanings, the researchers confirmed their suspicion that shorter, more frequent words, as well as those that conform to the language’s typical sound patterns, are most likely to be ambiguous — trends that were statistically significant in all three languages.

To understand why ambiguity makes a language more efficient rather than less so, think about the competing desires of the speaker and the listener. The speaker is interested in conveying as much as possible with the fewest possible words, while the listener is aiming to get a complete and specific understanding of what the speaker is trying to say. But as the researchers write, it is “cognitively cheaper” to have the listener infer certain things from the context than to have the speaker spend time on longer and more complicated utterances. The result is a system that skews toward ambiguity, reusing the “easiest” words. Once context is considered, it’s clear that “ambiguity is actually something you would want in the communication system,” Piantadosi says.

Tom Wasow, a professor of linguistics and philosophy at Stanford University, calls the paper “important and insightful.”

“You would expect that since languages are constantly changing, they would evolve to get rid of ambiguity,” Wasow says. “But if you look at natural languages, they are massively ambiguous: Words have multiple meanings, there are multiple ways to parse strings of words. … This paper presents a really rigorous argument as to why that kind of ambiguity is actually functional for communicative purposes, rather than dysfunctional.”

Implications for computer science

The researchers say the statistical nature of their paper reflects a trend in the field of linguistics, which is coming to rely more heavily on information theory and quantitative methods.

“The influence of computer science in linguistics right now is very high,” Gibson says, adding that natural language processing (NLP) is a major goal of those operating at the intersection of the two fields.

Piantadosi points out that ambiguity in natural language poses immense challenges for NLP developers. “Ambiguity is only good for us [as humans] because we have these really sophisticated cognitive mechanisms for disambiguating,” he says. “It’s really difficult to work out the details of what those are, or even some sort of approximation that you could get a computer to use.”

But, as Gibson says, computer scientists have long been aware of this problem. The new study provides a better theoretical and evolutionary explanation of why ambiguity exists, but the same message holds: “Basically, if you have any human language in your input or output, you are stuck with needing context to disambiguate,” he says.