Aegean Associates, Inc.

In the past several decades, microchips have transformed consumer electronics, enabling new products from digital watches and pocket-sized calculators to laptop computers and digital music players.

The next wave of this electronics revolution will involve biomedical devices, say electrical engineers in MIT’s Microsystems Technology Laboratory (MTL) who are working on tiny, low-power chips that could diagnose heart problems, monitor patients with Parkinson’s disease or predict seizures in epileptic patients. Such wearable or implantable devices could transform the way medicine is practiced and help cut the costs of expensive diagnostic tests, says Dennis Buss, former vice president of silicon technology development at Texas Instruments.

“Microelectronics have the potential to reduce the cost of health care in the same way they reduced the costs of computing in the 1980s and communications in the 1990s,” says Buss, a visiting scientist at MIT. On a limited scale, this is already taking place. For example, one of the first successful applications of microelectromechanical systems (MEMS) to medicine was the development of $10 disposable blood pressure sensors, which have been in use for over a decade and replaced sensors that cost hundreds of dollars.

Professor Charles Sodini, one of the MIT researchers involved in the effort, says the burgeoning field holds great potential for MIT and the greater Boston area because of the opportunities for collaboration between engineers, physicians and industry. “I want to see Boston become the Silicon Valley of medical electronic systems,” he says.

The market for MEMS for biomedical applications is more than $1 billion, and that could grow close to 100-fold by 2015, according to a 2006 market report from MedMarket Diligence.

Beating hearts

The key to developing small wearable and implantable medical monitors is an ultra-low-power chip for interfacing to biomedical sensors, signal processing, energy processing and communications, developed by the research group of MTL Director Anantha Chandrakasan.

Ultimately, Sodini and others at MTL hope to use that chip as the core of a device that can monitor a range of vital signs — heart rate, breathing rate, blood pressure, pulse oxygenation and temperature. For now, they’re starting with a monitor that measures and records electrocardiograms (ECGs).

An unobtrusive, comfortable ECG monitor that patients could wear as they go about their normal lives might offer a doctors a more thorough picture of heart health than the lab tests now used, says Collin Stultz, an MIT associate professor of electrical engineering and health sciences and technology and a cardiologist working on the project. Cardiologists can order up treadmill stress tests, MRIs and CT scans, among other diagnostics, but “all of these tests are done in contrived settings,” says Stultz. “Data obtained from more realistic, ‘at home’ settings may provide added information that can reveal potential problems.” Furthermore, standard tests can cost from a few hundred to a few thousand dollars.

Doctors often ask recent heart attack victims, and other patients suspected of having heart issues, to wear an ECG monitor as a Holter monitor for a few days. However, the device, which consists of several electrodes that stick to the chest, plus a bulky battery pack carried at the hip, is cumbersome and doesn’t have the memory to store much data.

In contrast, the new MIT monitor is an L-shaped device, about 4 inches along each side, that sticks to the chest and can be worn comfortably, with no external wires protruding. It can store up to two weeks of data in flash memory, and requires just two milliwatts of power. Eventually, the researchers hope to build chips that can harvest energy from the body of the person wearing the device, eliminating the need for a battery.

Doctors can use ECG data — which provides information on the electrical health of the heart — to help spot future problems. Stultz, working with MIT Professor John Guttag and recent PhD recipient Zeeshan Syed, has designed a computer algorithm that uses ECG data to assess risk of death in heart patients. They found that higher variability in heartbeat shapes in data recorded the day after a heart attack correlates with an eightfold increase in the risk of cardiac death within 90 days in some patient populations.

Currently that analysis can only be done after the data is downloaded from the chip, but eventually Stultz hopes to incorporate the algorithm into the chip itself. He envisions that the device could be equipped with an alarm that would alert the patient and/or doctor that a heart attack is imminent. It could also serve as an early detection system for longer-term problems, letting doctors know they may need to perform additional tests, alter the patient’s medication or perform surgery.

The researchers have built a prototype and plan to start testing the device in healthy subjects this spring, followed by trials in patients with cardiovascular disease.

New directions

While Stultz and colleagues are focusing on wearable devices, other MIT engineers are working on implantable electronics for medical monitoring. To do that, they need to overcome a significant challenge: how to run the device indefinitely without a battery that needs recharging. To solve that problem, Associate Professor Joel Dawson is working on a device that stores energy in an ultracapacitor, which doesn’t wear out like batteries do. He hopes to use the device, which would be about the size of a grain of rice, to measure tremors and shaking in patients with Parkinson’s disease.

Dawson is working on that project with neurologist Seward Rutkove of Beth Israel Hospital. That kind of collaboration between engineer and physician is exactly what Sodini would like to see happen with all of MTL’s biomedical projects. “We start out working with physicians so they can help define the problem, and they can start testing the devices in the clinic early in the process,” he says.

Other projects underway at MTL include tiny ultrasound devices and “lab on a chip” devices that can perform diagnostic tests on body fluids. Engineers are also working on the best ways to wirelessly transmit data from wearable or implanted devices to a cell phone or computer.

While those applications are promising, the future of biomedical electronics likely holds even more potential than we can imagine, says Buss.

“We will be using electronics in medical ways we don’t even conceive of yet,” he says. “When we started using cell phones, we had no idea we would be playing games and watching TV and surfing the Internet the way we do now.”

Once, if you wanted to become the general manager of a professional sports team, you had to have been a great athlete. For decades, sports teams were almost exclusively run by former players.

Times have changed. Today, an MBA can be a route into the NBA. Take Houston Rockets General Manager Daryl Morey, who has the height and bearing of a basketball star, but never played professionally. Instead, Morey graduated from the MIT Sloan School of Management in 2000, and parlayed his analytical skills into his current job.

“All else equal, it is preferable to have played the sport,” Morey said on Saturday, during a panel at the fourth annual MIT Sloan Sports Analytics Conference, which he co-founded. But all else is not equal: Sports are awash in misguided conventional wisdom, and scores of former players have blatantly mismanaged franchises. So Morey is in the vanguard of general managers applying the analytical techniques of academia to basketball.

In practice, that means Morey’s staff has been dissecting the sport, doing things like pinpointing the most efficient shot location (the three-pointer from the corner), and slicing defensive performance into small, measurable elements, in an attempt to quantify how effective Houston’s players are. Their forward Chuck Hayes, for example, would be considered too small for his position, at a mere 6’6,” according to the conventions of coaches and scouts, but new-school metrics indicate that Hayes is a defensive ace. “You have to have a culture where there are no bad ideas,” said Morey, meaning he encourages his staff to develop new ways of assessing talent. As a result, a year ago, unheralded Houston pushed the eventual champion Los Angeles Lakers to the seven-game limit in their playoff series.  

To be sure, the field of sports analytics has existed for years: The baseball writer Bill James’ pioneering annual book, “The Baseball Abstract,” began reaching a national audience in 1982. The subject gained new popularity through Michael Lewis’ best-seller, Moneyball (Norton, 2003), which chronicled how the Oakland Athletics were using James’ principles to find undervalued players.

The Sloan conference, which featured panels examining analytical techniques, and research papers on subjects like blocked shots in basketball (not all of them are equally valuable), reflects this wave of interest. Saturday’s event, held at the Boston Convention & Exhibition Center, drew more than 1,000 attendees, up from 400 last year; half the NBA’s teams had a representative present.

Is Plus/Minus a plus?

The current state of analytics varies widely among sports. Baseball is the most developed, because it largely consists of a series of individual confrontations between pitchers and hitters, whose results can be easily isolated. As a morning session on “Baseball Analytics” made clear, defense is the last statistical frontier of the game, and even there, statistician John Dewan estimated, observers know “60 percent” of everything they can.

Baseball analytics are so thorough, “Now I don’t think you even have to watch baseball” to dissect it, quipped ESPN.com columnist Bill Simmons during an afternoon panel. Indeed, he added, you may not even “need to know how to hold a bat.”

But other sports feature the simultaneous interaction of many athletes at once. Isolating an individual’s performance in these sports remains problematic.

“Unlike baseball where you have a lot of discrete events, in football there is a lot of interplay, so it’s more difficult to analyze,” said Parag Marathe, a San Francisco 49ers executive, at a panel on “Emerging Analytics.” Consider a 25-yard run. How much of the credit goes to the running back, his blockers, or to defense lapses? “The NFL is a little bit behind” in analytics, Marathe suggested.

To work around the problem of complex interactions in basketball, analysts are refining the concept of “Plus/Minus,” which records how many points a team scores and allows when a particular player is on the court, per 100 possessions. One winner of the conference’s research-paper contest this year attempted to improve the concept; Dallas Mavericks owner Mark Cuban has tried to use Plus/Minus, while recognizing its flaws.

“There are all these qualifications you need to keep in mind,” Cuban said in an interview with MIT News on Saturday after he spoke on two panels. A player’s Plus/Minus can depend on the quality of his teammates, the quality of opponents, the tempo of play, and more. Currently, Miami’s Dwyane Wade leads the league in Plus/Minus, but that may just mean that he has worse teammates than Cleveland’s LeBron James.

That said, Cuban thinks the metric works well in evaluating the success of different five-man lineups, not just single players. “We’ve adjusted lineups in the playoffs based on our Plus/Minus numbers,” Cuban said. In 2005, Dallas lost the first two games of its first-round series to Houston, which was using a smaller, quicker lineup. The Mavericks studied the Plus/Minus numbers, reduced lumbering center Erick Dampier’s minutes, and rallied to win the series in 7 games.

“Mark Cuban helped break me out of that mold of looking at traditional statistics,” recounted Avery Johnson, the Mavericks’ coach at the time, while speaking on a “Coaching Analytics” panel. “Using Plus/Minus helped me out a lot in terms of my substitutions.”

Well, until the team faltered. In 2007, Dallas entered the playoffs with a league-best 67-15 record. But as Johnson recounted, the numbers showed that the Mavericks fared worse against their first-round opponent, the small-but-quick Golden State Warriors, with Dampier on the court. Johnson benched Dampier, the team’s starting center, for the series’ first game. “It was the right thing to do,” Johnson said. But his players did not like the adjustment; the Warriors quickly knocked out the Mavericks in a stunning upset.

The limits of metrics

As the Mavericks’ experience suggests, analytics have limitations. General tendencies may not be borne out in specific situations. Moreover, “I think there is an onus on whoever is dispensing that information” to explain it clearly and persuasively to everyone else, asserted Simmons, whose own recent tome, The Book of Basketball (ESPN 2009), mixes empirical data and subjective impressions while judging players and teams in NBA history.

And Morey noted another problem: In a business with short careers, changing circumstances may make some sports analysis irrelevant. “I think there are fundamental things that can be solved,” said Morey. “But by the time you have enough confidence in them, the world has changed.”

What has also changed, though, is that savvy sports fans now envision a future in the business. Take Matthew Martell, a senior associate at Octothorpe Software, a Vancouver firm that designs decision-making programs. Martell, capable of talking knowledgeably about sports-analytics problems in basketball, football, and soccer, made a 12-hour trip from British Columbia on Thursday, changing planes twice, to attend the event. “This is where you want to be, to meet and see the people who really know analytics,” said Martell. “It’s incredible to be here.”

Clean-energy technologies offer the promise of revitalizing a dwindling base of manufacturing jobs in the United States while also addressing the problems of climate change and energy security, said Democratic Sen. Jeff Bingaman of New Mexico in a keynote speech at the annual MIT Energy Conference. But as great as the potential may be, it won't be realized unless substantial new policies and regulations are put in place — and the chances of that happening anytime soon are slim, he said.

While Congress and the Obama administration have taken some first steps, he said, "the policies that have been enacted to date are clearly not sufficient to establish the U.S. as the leader in clean technology." Right now, he said, "90 percent of the production capacity for new clean technology is outside the United States."

Bingaman added that "China is moving ahead very aggressively," and the United States needs to act soon to reverse the present tide. For example, while lithium-ion battery technology was developed in this country, only 1 percent of the manufacturing of these batteries — now used mostly in portable electronics devices, but seen as a key to the next generation of electric vehicles — takes place in the U.S.

That view of great potential but political stagnation was echoed by several speakers at the conference, which was held on March 6 at the Sheraton Boston. Speakers representing various levels of government, industry, academic research, international organizations, and the financial sector, among others, tended to agree that government action will play a crucial and decisive role in determining how the world responds to the challenges of growing energy demand and the risks of climate change, and how different nations' economies fare as a result.

There were two areas where clear government action was seen as being especially important: first and foremost, setting in place a clear and predictable system that puts a price on emissions of carbon, whether it be in the form of a cap-and-trade system, as the U.S. Congress has been considering, or simply a direct tax on carbon, which many consider to be a better option but not a politically feasible one; and second, offering financial support for new energy technologies, not only at the research stage but also in establishing manufacturing capacity.

Regaining the competitive edge

Globally, the trend toward non-fossil-fuel energy is clear: In 2008, Bingaman said, for the first time global investments in clean energy technology exceeded those for fossil fuel technology. But for U.S. competitiveness, the trend is not encouraging. For many years, he said, the U.S. made the technological breakthroughs, while other countries, especially Japan, provided the follow-through. But now, other countries are joining in the follow-through, and "the U.S. no longer has a monopoly on the breakthroughs."

There are ways to turn that around, suggested Bingaman, who chairs the Senate Energy and Natural Resources Committee, but only with substantial policy changes. Clean technology "offers the opportunity to revitalize our manufacturing sector," he said, but in the past the kinds of incentives the government has provided "were concentrated downstream," on the consumers or suppliers rather than on the manufacturing end of the spectrum, and the policies have tended to come and go with changing political tides, resulting in "government-driven boom-and-bust cycles."

To change that, several senators and congressmen, with President Obama's support, are urging the creation of a substantial loan-guarantee program for clean-tech manufacturing. The biggest impact of all, Bingaman said, could come from improvements in energy efficiency, which could both produce a dramatic lowering of greenhouse emissions and oil imports, and at the same time create large numbers of long-term jobs. But to make that happen requires some form of price on carbon-emitting fuels, he said.

That's not likely in this country anytime soon, he added. "Getting comprehensive climate-change legislation is not that promising this year," he said, though he still has hope for some steps in that direction. And what might be possible next year depends on the outcome of the fall elections, he said.

Looking for innovation — and consistency

"The government's role is vital, and temporary," said David Anthony, managing partner of the investment company 21Ventures, at the conference's closing panel discussion about the financing of energy technology. He stressed that the government's main role is to invest in basic research and development, at the early stages where private financing is too difficult to secure. "The government needs to fix the problem, and then get out of the way," he said.

But while the legislative process is moving slowly, many segments of industry are moving ahead. "It's easy for me to be pro-climate legislation — it is in my economic self-interest," said John Rowe, CEO of Entergy, the nation's largest electric utility company and owner of the nation's largest fleet of nuclear power plants. Rowe, one of the conference's keynote speakers, explained that the greatest danger, from a business point of view, lies in "continuing to deal with energy in ways that are haphazard," as opposed to setting a clear policy in place that businesses can base their plans on.

Rowe, whose Chicago-based electric utility holding company has already closed many of its coal-burning plants and plans to eliminate all of its 15 million tons of greenhouse gas emissions by 2020, said that addressing the problems of greenhouse-gas emissions will depend on putting a price on carbon, either through a cap-and-trade system, or a carbon tax.

"We ought to have a predictable, confident and decisive policy on climate change," he said. Public resistance is largely based on incorrect assumptions, he suggested, because polls show people oppose carbon taxes or cap-and-trade systems because they believe those will cost them money, but they support renewable energy standards — requiring utilities to provide a set percentage of their power from renewable energy — because they believe those are cost-free. That kind of free lunch, he suggested, is an illusion.

But if policies are put in place that set a realistic price on carbon emissions, he said, the marketplace will do the rest. With such a policy, "you'll be surprised at how much can happen in 10 years."

A team of scientists at MIT have discovered a previously unknown phenomenon that can cause powerful waves of energy to shoot through minuscule wires known as carbon nanotubes. The discovery could lead to a new way of producing electricity, the researchers say.

The phenomenon, described as thermopower waves, “opens up a new area of energy research, which is rare,” says Michael Strano, MIT’s Charles and Hilda Roddey Associate Professor of Chemical Engineering, who was the senior author of a paper describing the new findings that appeared in Nature Materials on March 7. The lead author was Wonjoon Choi, a doctoral student in mechanical engineering.

Like a collection of flotsam propelled along the surface by waves traveling across the ocean, it turns out that a thermal wave — a moving pulse of heat — traveling along a microscopic wire can drive electrons along, creating an electrical current.

The key ingredient in the recipe is carbon nanotubes — submicroscopic hollow tubes made of a chicken-wire-like lattice of carbon atoms. These tubes, just a few billionths of a meter (nanometers) in diameter, are part of a family of novel carbon molecules, including buckyballs and graphene sheets, that have been the subject of intensive worldwide research over the last two decades.

A previously unknown phenomenon

In the new experiments, each of these electrically and thermally conductive nanotubes was coated with a layer of a reactive fuel that can produce heat by decomposing. This fuel was then ignited at one end of the nanotube using either a laser beam or a high-voltage spark, and the result was a fast-moving thermal wave traveling along the length of the carbon nanotube like a flame speeding along the length of a lit fuse. Heat from the fuel goes into the nanotube, where it travels thousands of times faster than in the fuel itself.  As the heat feeds back to the fuel coating, a thermal wave is created that is guided along the nanotube. With a temperature of 3,000 kelvins, this ring of heat speeds along the tube 10,000 times faster than the normal spread of this chemical reaction. The heating produced by that combustion, it turns out, also pushes electrons along the tube, creating a  substantial electrical current.

Combustion waves — like this pulse of heat hurtling along a wire — “have been studied mathematically for more than 100 years,” Strano says, but he was the first to  predict that such waves could be guided by a nanotube or nanowire and that this wave of heat could push an electrical current along that wire.

In the group’s initial experiments, Strano says, when they wired up the carbon nanotubes with their fuel coating in order to study the reaction, “lo and behold, we were really surprised by the size of the resulting voltage peak” that propagated along the wire.

After further development, the system now puts out energy, in proportion to its weight, about 100 times greater than an equivalent weight of lithium-ion battery.

The amount of power released, he says, is much greater than that predicted by thermoelectric calculations. While many semiconductor materials can produce an electric potential when heated, through something called the Seebeck effect, that effect is very weak in carbon. “There’s something else happening here,” he says. “We call it electron entrainment, since part of the current appears to scale with wave velocity.”

The thermal wave, he explains, appears to be entraining the electrical charge carriers (either electrons or electron holes) just as an ocean wave can pick up and carry a collection of debris along the surface. This important property is responsible for the high power produced by the system, Strano says.

Exploring possible applications

Because this is such a new discovery, he says, it’s hard to predict exactly what the practical applications will be. But he suggests that one possible application would be in enabling new kinds of ultra-small electronic devices — for example, devices the size of  grains of rice, perhaps with sensors or treatment devices that could be injected into the body. Or it could lead to “environmental sensors that could be scattered like dust in the air,” he says.

In theory, he says, such devices could maintain their power indefinitely until used, unlike batteries whose charges leak away gradually as they sit unused. And while the individual nanowires are tiny, Strano suggests that they could be made in large arrays to supply significant amounts of power for larger devices.

The researchers also plan to pursue another aspect of their theory: that by using different kinds of reactive materials for the coating, the wave front could oscillate, thus producing an alternating current. That would open up a variety of possibilities, Strano says, because alternating current is the basis for radio waves such as cell phone transmissions, but present energy-storage systems all produce direct current. “Our theory predicted these oscillations before we began to observe them in our data,” he says.

Also, the present versions of the system have low efficiency, because a great deal of power is being given off as heat and light. The team plans to work on improving that efficiency.

Ray Baughman, director of the Nanotech Institute at the University of Texas at Dallas, who was not involved in this work, calls the research “stellar.”

The work, Baughman says, “started with a seminal initial idea, which some might find crazy, and provided exciting experimental results, the discovery of new phenomena, deep theoretical understanding, and prospects for applications.” Because it uncovered a previously unknown phenomenon, he says, it could open up “an exciting new area of investigation.”

Most polymers — materials made of long, chain-like molecules — are very good insulators for both heat and electricity. But an MIT team has found a way to transform the most widely used polymer, polyethylene, into a material that conducts heat just as well as most metals, yet remains an electrical insulator.

The new process causes the polymer to conduct heat very efficiently in just one direction, unlike metals, which conduct equally well in all directions. This may make the new material especially useful for applications where it is important to draw heat away from an object, such as a computer processor chip. The work is described in a paper published on March 7 in Nature Nanotechnology.

The key to the transformation was getting all the polymer molecules to line up the same way, rather than forming a chaotic tangled mass, as they normally do. The team did that by slowly drawing a polyethylene fiber out of a solution, using the finely controllable cantilever of an atomic force microscope, which they also used to measure the properties of the resulting fiber.

This fiber was about 300 times more thermally conductive than normal polyethylene along the direction of the individual fibers, says the team’s leader, Gang Chen, the Carl Richard Soderberg Professor of Power Engineering and director of MIT’s Pappalardo Micro and Nano Engineering Laboratories.

The high thermal conductivity could make such fibers useful for dissipating heat in many applications where metals are now used, such as solar hot water collectors, heat exchangers and electronics.

Chen explains that most attempts to create polymers with improved thermal conductivity have focused on adding in other materials, such as carbon nanotubes, but these have achieved only modest increases in conductivity because the interfaces between the two kinds of material tend to add thermal resistance. “The interfaces actually scatter heat, so you don’t get much improvement,” Chen says. But using this new method, the conductivity was enhanced so much that it was actually better than that of about half of all pure metals, including iron and platinum.

Producing the new fibers, in which the polymer molecules are all aligned instead of jumbled, required a two-stage process, explains graduate student Sheng Shen, the lead author of the paper. The polymer is initially heated and drawn out, then heated again to stretch it further. “Once it solidifies at room temperature, you can’t do any large deformation,” Shen says, “so we heat it up twice.”

Even greater gains are likely to be possible as the technique is improved, says Chen, noting that the results achieved so far already represent the highest thermal conductivity ever seen in any polymer material. Already, the degree of conductivity they produce, if such fibers could be made in quantity, could provide a cheaper alternative to metals used for heat transfer in many applications, especially ones where the directional characteristics would come in handy, such as heat-exchanger fins (like the coils on the back of a refrigerator or in an air conditioner), cell-phone casings or the plastic packaging for computer chips. Other applications might be devised that take advantage of the material’s unusual combination of thermal conductivity with light weight, chemical stability and electrical insulation.

So far, the team has just produced individual fibers in a laboratory setting, Chen says, but “we’re hoping that down the road, we can scale up to a macro scale,” producing whole sheets of material with the same properties.

Ravi Prasher, an engineer at Intel, says that “the quality of the work from Prof. Chen’s group has always been phenomenal,” and adds that “this is a very significant finding” that could have many applications in electronics. The remaining question, he says, is “how scalable is the manufacturing of these fibers? How easy is it to integrate these fibers in real-world applications?”

This work, which also included Chen’s former graduate students Asegun Henry and Jonathan Tong, was supported by the National Science Foundation and the Department of Energy’s Office of Basic Energy Sciences.

MIT researchers have built the first sensor array that can detect single molecules emitted by a living cell. Their sensor targets hydrogen peroxide and could help scientists learn more about that molecule’s role in cancer.

Hydrogen peroxide has long been known to damage cells and their DNA, but scientists have recently uncovered evidence that points to a more beneficial role: it appears to act as a signaling molecule in a critical cell pathway that stimulates cell growth, among other functions.

When that pathway goes awry, cells can grow out of control and become cancerous, so understanding hydrogen peroxide’s role could lead to new targets for potential cancer drugs, says Michael Strano, MIT associate professor of chemical engineering and leader of the research team. Strano and his colleagues describe their new sensor array, which is made of carbon nanotubes, in the March 7 online edition of Nature Nanotechnology.

Strano’s team is also working on carbon nanotube sensors for other molecules, and within the past year has successfully tested and published sensors for nitric oxide and ATP (the molecule that carries energy within a cell).

“The list of biomolecules that we can now detect very specifically and selectively is growing rapidly,” says Strano, who also points out that the ability to detect and count single molecules sets carbon nanotubes apart from many other nanosensor platforms, including electrochemical, electromechanical cantilevers and surface acoustic wave sensors.

Nanotube array

In the new study, Strano’s team used the carbon nanotube array to study the flux of hydrogen peroxide that occurs when a common growth factor called EGF activates its target, a receptor known as EGFR, which is located on cell surfaces. For the first time, the team showed that hydrogen peroxide levels more than double when EGFR is activated.

EGF and other growth factors induce cells to grow or divide through a complex cascade of reactions inside the cell. It’s still unclear exactly how hydrogen peroxide affects this process, but Strano speculates that it may somehow amplify the EGFR signal, reinforcing the message to the cell. Because hydrogen peroxide is a small molecule that doesn’t diffuse far, the signal would be limited to the cell where it was produced.

The team also found that in skin cancer cells, believed to have overactive EGFR activity, the hydrogen peroxide flux was 10 times greater than in normal cells. Because of that dramatic difference, Strano believes this technology could be useful in building diagnostic devices for some types of cancer.

“You could envision a small handheld device, for example, which your doctor could use to assay tissue in a minimally invasive manner and tell if this pathway is corrupted,” he says.

The sensor consists of a film of carbon nanotubes embedded in collagen. Cells can grow on the collagen surface, and the collagen also attracts and traps hydrogen peroxide released by the cell. When the nanotubes come in contact with the trapped hydrogen peroxide, their fluorescence flickers.  By counting the flickers, one can obtain an accurate count of the incident single molecules.

The new sensor represents “an excellent example of the application of nanotechnology to address fundamental questions in biology,” says Ravi Kane, professor of chemical and biological engineering at Rensselaer Polytechnic Institute.

Strano points out that this is the first time an array of sensors with single-molecule specificity has ever been demonstrated.  He and his colleagues derived mathematically that such an array could distinguish “near field” molecular generation from that which takes place far from the sensor surface. 

“Arrays of this type have the ability to distinguish, for example, if single molecules are coming from an enzyme located on the cell surface or from deep within the cell,” says Strano.

In future work, researchers in Strano’s lab plan to study different forms of the EGF receptor to better characterize the hydrogen peroxide flux and its role in cell signaling. They have already discovered that molecules of oxygen are consumed to generate the peroxide.

In response to the earthquake in Haiti, MIT Media Lab students have developed a service that helps communities rebuild after a crisis by indexing the skills of local residents so that NGOs like the American Red Cross and Partners In Health can quickly find and employ them.

Since January, Greg Elliott and Aaron Zinman have been developing Konbit, a free interactive communication platform that allows Haitians, their diaspora and the international community to report their skills by phone, text message or web. In anticipation of long-term rebuilding efforts, the goal of Konbit is to index everyday skills, such as language or construction skills, that aren’t currently being advertised or tracked by sites like Craigslist or Monster.

Composed of several hardware and software systems, Konbit allows people from multiple countries to work together to help disaster victims find employment and rebuild their economy. It includes software for web, text, phone and translation services, as well as servers located in Cambridge and custom phone hardware to be installed in Haiti.

Konbit is language- and medium-neutral, meaning that voice and text messages can be translated through the Konbit phone, text or web interface. The voice component is crucial to Konbit because more than 60 percent of the Haitian population is illiterate, according to UNICEF.

Messages in native Creole will be translated by volunteer Haitians and then transcribed into the database so that NGOs can search for specific skills in real-time and by location.

In addition to aiding reconstruction, another goal of Konbit is to prevent the outsourcing of labor. When aid organizations bring non-Haitians into Haiti for relief and reconstruction work, this prevents Haitians from receiving training and experience that could be valuable once the relief teams have left. It also hurts the Haitian economy, which had a 70 percent unemployment rate before the earthquake, according to the U.S. Agency for International Development.

Zinman and Elliott hope to launch a prototype of the Konbit platform in early March. While the phone and web interfaces are essentially working, Zinman and Elliott are trying to get the major telecommunications companies in Haiti to deploy the service as soon as possible.

Konbit got its start in the four-day Independent Activities Period workshop sponsored by the Media Lab and the Center for Future Civic Media that was aimed at developing innovative technologies to alleviate the crisis caused by the Jan. 12 earthquake in Haiti.

Dale Joachim, a Media Lab visiting scientist who helped run that workshop, is now teaching “New Media Projects for Haiti” with Barry Vercoe, professor of media arts and sciences in the Media Lab. The project-oriented class will explore how communications technology can help rebuilding efforts. The class has about 30 undergraduate and graduate students enrolled, as well as a handful of participants who are attending the lectures.

The first half of the class includes lectures on topics about Haitian society, such as economics, education and language, that will help student groups choose a societal problem and devise solutions. The class will travel to Haiti during the last week of April to field test and document its solutions. Each project will have its own evaluation plan that will be discussed when the class returns to MIT.

Crisis, management

The Sloan School of Management is also offering a class directly involved in Haitian relief efforts. Several students from that class, “Applications of System Dynamics: Global Challenges,” are helping the U.S. military quickly analyze data for humanitarian-relief needs.

Taught by Sloan senior lecturer Anjali Sastry, the project developed after Marc Zissman, assistant head of the Communications and Information Technology Division at MIT Lincoln Laboratory, asked for help assessing the current state of health, food, shelter and water in Haiti. Zissman is helping the U.S. military Joint Task Force in Haiti coordinate a group that will survey 288 displacement camps and neighborhoods in the country to determine the needs and supplies for the overall humanitarian effort.

On a weekly basis, Sastry’s students will analyze up-to-the-minute data collected in Haiti over the next one to four months. The results are urgently needed for planning and decision-making and will be reported to the military, NGOs and the U.N. In addition to the rapid data analysis, the students will produce a set of white papers to help frame the results within the larger picture of long-term sustainability in Haiti, and how the humanitarian efforts might play out over time.

“This is really about putting MIT skills to the test,” Sastry said, urging the various relief and reconstruction projects within the MIT community to communicate and partner with one another so there can be a better understanding of the needs and opportunities in Haiti.

That line of thinking will guide the March 8 retreat by the MIT/Haiti Response Advisory Group. Hosted by Vice Chancellor and Dean for Graduate Education Steve Lerman, the goal of the retreat is to identify viable MIT-led projects that could meet Haiti’s needs.

Behavioral economics is used to examine how consumers make decisions about everything from their life savings to which brands of jam they select in a supermarket. Hunt Allcott, a behavioral economist with a two-year appointment as the Energy and Society Fellow in MIT’s Department of Economics and the MIT Energy Initiative, wants to apply his field’s insights to the realm of energy use.

In the latest issue of Science, Allcott and co-author Sendhil Mullainathan, of Harvard, advocate passage of a bill currently in Congress that would fund more behavioral research about energy consumption. The authors also note initiatives like that of OPOWER, a Virginia company, which has found that the user-friendly energy reports it sends to consumers can influence behavior enough to reduce household energy use by 2 percent, at minimal cost (OPOWER is an affiliate of Ideas42, an MIT-linked think tank to which Allcott also belongs). MIT News spoke with Allcott about how behavioral economics addresses our energy needs.

Q. Why should we invest in behavioral research pertaining to energy efficiency, and what are the specific kinds of research we can do right now?

A. It’s an economic argument. There are lots of different technology-centered R&D investments that we can and do make: Fuel cells, hybrid vehicles, wind power. But we can also invest in new social science research that can inform policies and programs that encourage people to consume energy differently. The argument Sendhil and I make is that we have to compare across all of these classes and say, “What’s cost-effective in terms of achieving our goals?” We use the results from recent large-scale energy conservation programs that were motivated by behavioral science to show that behavioral science R&D is an underexplored and potentially cost-effective approach.

Let me give you two examples of how economics can inform energy efficiency policy. First, much of the policy-oriented research in behavioral economics has been about identifying barriers in individual decision making that keep us from making the choices that, in a perfect world, we would have wanted to make for ourselves. Perhaps the leading example of this has been helping people to make better choices about how much to invest in their retirement plans, and what funds to hold. One of the things I’m interested in is to document whether consumers make similar types of mistakes when they go to buy air conditioners, or cars. It’s a complex decision, and the benefits of energy efficiency occur incrementally and in the future, so those benefits are not very salient. Depending on the types of mistakes that consumers are making — if we conclude they are indeed making mistakes — we can design policies to nudge them in ways that they would find helpful.

Second, economists tend to think of energy consumption as driven primarily by prices. Indeed, in many domains, I think we reflexively focus on price at the expense of failing to model other important drivers of consumer choice. There’s a lot of research in behavioral economics that suggests we can influence people to conserve energy, or do other things, in many ways other than raising prices. I think an important research area is to document whether policies and programs based on these sorts of insights can increase welfare or be cost-effective in reducing carbon emissions.

Q. To what extent will consumers make different choices if they simply have the facts about energy explained to them in a clear manner?

A. The effect of clearer information is an empirical question that often has surprising answers. One example of this is from OPOWER, a company that our research group interacts with a lot. OPOWER sends home energy use reports to households that compare those households to their neighbors and give energy conservation tips. The information in these reports is very similar to what’s already on a utility bill: How much did you spend this month, how much did you spend this year, here’s where you can get compact fluorescent lightbulbs. But something about the way they’re presenting it — presumably the way they use comparisons to neighbors — seems to be very powerful. I’m not sure it would have been obvious to any of us 10 years ago or three years ago that this program would have large effects in the real world.

There was an academic study by psychologist Bob Cialdini and co-authors that helped provide the proof-of-concept for the OPOWER program. In this study, the researchers left door-hangers at a group of households in California. Some of the door-hangers said, “Save money by saving energy,” some of them said, “Save the environment,” and some said, “Here’s how much your neighbors are using.” And the ones that said, “Here’s how much your neighbors are using” had a much stronger impact on energy consumption. In the last couple of years that study in particular has had a lot of influence.

Q. Okay, so why is it that referring to neighbors is effective?

A. Psychologists have been great at documenting that if you tell people what the social norm is, people will converge to the social norm. In my mind there are two leading economic hypotheses for why this works in energy consumption. One is called “conditional cooperation.” People may be altruistic, and they view conserving energy as contributing to the public good of reducing climate change. People are typically more willing to contribute to a public good if they are informed that other people are contributing more than they are.

The other explanation is just social inference. It could be that I couldn’t care less about the environment, but I do want to save money. And if you tell me that I’m using twice as much energy as my neighbor, that lets me know that maybe I’ve been leaving a window open or that my furnace is inefficient. So that’s purely a self-interested, informational story. Testing between these two explanations is one of the research questions we’re interested in.

Today, computers can’t reliably identify the objects in digital images. But if they could, they could comb through hours of video for the two or three minutes that a viewer might be interested in, or perform web searches where the search term was an image, not a sequence of words. And of course, object recognition is a prerequisite for the kind of home assistance robot that could execute an order like “Bring me the stapler.” Now, MIT researchers have found a way to improve object recognition systems by using information about context. If the MIT system thinks it’s identified a chair, for instance, it becomes more confident that the rectangular thing nearby is a table.

A typical object recognition system will scan a digital image for groups of pixels that differ from those around them; those pixels could define an edge, a corner or some other feature of an object. Usually, the system has been trained on a set of sample images, which teaches it how to correlate feature patterns with particular objects.

Some researchers have tried to use context information to refine those correlations. But according to Myung “Jin” Choi, a grad student in MIT’s Laboratory for Information and Decision Systems and one of the leaders of the new project, those researchers were generally working with a standard training set that included examples of only about 20 different types of object. In that case, it was fairly straightforward to specify how frequently each object co-occurred with every other object in the set.

Upping the ante

A system that could recognize only 20 different objects, however, wouldn’t be very useful. And with a large number of objects, it becomes computationally impractical to consider the frequency of all possible two-object combinations. In work to be presented at the IEEE Conference on Computer Vision and Pattern Recognition this summer, Choi and her colleagues — including graduate student Joseph Lim and Professors Antonio Torralba and Alan Willsky — describe a different approach. Working with a training set that included more than 4,000 images and 107 different types of objects, they created algorithms that pored through the images and automatically constructed a hierarchical map of the object categories — kind of like the organizational chart for a large company, which shows who reports to whom. In the map, each object is connected to at most one object above it in the hierarchy (everyone in the organization reports to only one person), drastically reducing the number of connections that the system has to consider. The connection between any two objects is given a weight that indicates how often the objects appear together in the training images. The map also encodes information about the typical relative locations of two connected objects: buildings generally appear above roads, for instance, not below them.

When the system analyzes a new image, it uses standard object recognition algorithms to generate a list of candidate objects, together with each object’s “confidence score” — a statistical measure of how likely the object is to have been correctly identified. Then it revises those scores on the basis of the information encoded in the contextual map.

In experiments, Choi compared the performance of the bare object recognition algorithms with their performance when augmented by the contextual map. In both cases, she considered the three objects per image with the highest confidence scores. The bare algorithms correctly identified all three objects roughly 14 percent of the time; with the addition of the contextual map, the success rate jumped to about 25 percent.

Long row to hoe

Of course, that means that the system still failed to correctly identify three objects per image about 75 percent of the time, which shows just how difficult the problem of object recognition remains. “Context really is essential,” says Serge Belongie, an associate professor of computer science at the University of California, San Diego who has worked on both object recognition in general and context-based object recognition in particular. “It deserves a proper treatment, and Jin is doing that.” But Belongie cautions that context awareness will never be more than an augmentation of an underlying system that recognizes objects from visual features. “We absolutely cannot afford to take our eye off the ball of the component recognition systems that need to feed these context engines,” he says. And, he adds, to be useful, object recognition systems will need to be much more precise than today’s prototypes are. “Imagine that you take a picture of a wild mushroom while you’re hiking,” Belongie says, “and then you send it to the system to find out what it is. And it says, ‘Mushroom!’ You’re like, Thanks. That’s really useful. I knew that part.”

Nonetheless, Choi is continuing to improve her contextual-map system, against the day when the underlying algorithms are more reliable. The next version of the system, she says, will add entries to the map that, in effect, represent higher-level scene descriptions. Street scenes, for instance, may frequently feature sky, buildings and roads, while building interiors may frequently feature floors, walls and windows. The system won’t need to explicitly label these additional map entries, however; it will simply register them as foci around which certain types of objects regularly cluster. She’s confident that this modification will make the added benefits of context awareness even more acute.

tktkt

The role of hurricanes in the global climate system has gained interest ever since scientists suggested that strong hurricanes have become more frequent in recent decades and might continue to do so as the planet warms. Because hurricanes are known to influence the oceans and overall climate system, the consequences of the increase in the frequency of hurricanes could reach further.

When a hurricane passes over an ocean, its powerful winds stir and mix the warm surface water with the colder, deeper water. This mixing results in warm water being forced down into the deep ocean and cold water being brought to the surface layer. Scientists know that the cold water near the surface is reheated by the atmosphere to pre-hurricane temperatures within a few weeks, but they have been less clear on what happens to the warm water mixed into the deep ocean. It has been suggested that this heat is transported toward the poles by ocean currents and contributes to the ocean heat transport, the process by which oceans regulate our climate by transporting warm water away from the equator and cold water toward the equator. It has also been speculated that the heat pumped into the ocean by hurricanes strengthens subsequent storms that pass over the same part of the ocean, because ocean heat is the energy source that powers hurricanes. Stronger storms would then mix even more heat into the ocean driving a positive feedback loop for hurricane intensity.

A new MIT analysis suggests that previous studies have overestimated the amount of hurricane-induced ocean heating and its overall impact on climate. The analysis indicates that previous estimates have failed to consider how the oceans change with the seasons.

Most of the heat from the warm water that hurricanes mix deep into the oceans during the summer and early fall is returned to the atmosphere in the winter, meaning these “warm anomalies” don’t appear to affect the long-term state of the oceans, according to a paper published Feb. 10 in Geophysical Research Letters by Raffaele Ferrari, the Cecil and Ida Green Professor of Oceanography in the Department of Earth, Atmospheric and Planetary Science; graduate student Malte F. Jansen; and undergraduate student Todd Mooring.

Jim Price, a senior scientist at Woods Hole Oceanographic Institution, believes Ferrari and his students’ estimate more accurately reflects how hurricanes affect the long-term state of the oceans. “They’ve taken a step that makes the previous estimates look a little bit excessive,” he said.

By analyzing the paths of almost 1,000 tropical cyclones, as well as data about the sea surface temperature and height before and after each storm, Ferrari and his students estimate that only about one-quarter of the warm water that is mixed downward by a hurricane remains in the ocean for more than a season.

Previous estimates, including a 2001 study by Kerry Emanuel, the Breene M. Kerr Professor of Atmospheric Science, suggested that on average, hurricanes contribute about one petawatt (equal to one quadrillion watts) of heat to the estimated two to four petawatts of heat that the ocean transports out of the tropics and toward the poles each year.

Although Emanuel and a colleague concluded in 2008 that only about half of the heat pumped into the ocean by hurricanes contributes to the total ocean heat transport toward the poles, Jansen said this study did not consider the crucial role of the seasonal cycle.

Analyzing the seasonal cycle

The ocean is made up of several layers. The uppermost layer has the highest temperatures and it is called a mixed layer, because it is continuously being mixed and thus kept in contact with the atmosphere. Below this is the thermocline layer, which gets colder as depth increases. During the summer, the mixed layer is shallow, often only about 10-20 meters deep, but during the winter, cooler atmospheric temperatures and stronger winds can cause the mixed layer to expand to more than 100 meters. As a result, the mixed layer extends into what used to be the upper part of the thermocline, called the seasonal thermocline, thereby re-absorbing any warm anomaly that was deposited in this layer. Only anomalies in the part of the thermocline that remains below the mixed layer for the entire year, known as the permanent thermocline, will remain after the winter.

To calculate the fraction of heat from the mixed layer that is deposited into the permanent thermocline in the tropics, the researchers relied on theoretical models, as well as data about how the oceans changed as a result of all tropical cyclones that occurred between 1998 and 2006. Previous work, not related to Emanuel’s estimates, calculated that about half a petawatt of heat is pumped into the ocean during a hurricane. Using a formula that takes into account how the mixed layer deepens during the winter, Ferrari’s team concluded that this amount is actually between zero and .3 petawatts with a most likely estimate of .15 petawatts.

The analysis suggests that the major part of a warm anomaly sits in the seasonal thermocline until the following winter, when it is reabsorbed by the deepened mixed layer; at that point, it interacts with the atmosphere and releases its extra heat. Rather than contributing to ocean heat transport, therefore, the warm anomalies function as thermostats by transferring their heat to the atmosphere during the winter. Additional research would presumably focus on the effects of this warmer atmosphere during winter.

Emanuel praised the result, noting that “the path is different from what we initially envisioned” almost a decade ago. He explained that the interest in ocean mixing has grown recently also because of its potential biological consequences. Because cold water contains more dissolved carbon dioxide than warm water, when hurricanes bring that cold water closer to the surface, the carbon dioxide interacts with and enters the atmosphere. At the same time, however, as the nutrient-rich cold water is pushed closer to the surface and becomes exposed to sunlight, the number of tiny plant-like microorganisms called phytoplankton multiplies. When the plankton eventually die, the carbon in their shells is deposited at the bottom of the sea. Ferrari predicts that the seasonal effects of ocean mixing will become major factors for research that seeks to determine the net balance of carbon dioxide leaving and staying in the ocean.

Jansen added that a potential next step for research on tropical cyclones and ocean mixing could examine how hurricane-induced ocean heating affects El Niño, a weather pattern in the Pacific Ocean where most of the strongest tropical cyclones occur, that is related to large climate disturbances. Whether the warm water mixed deep into the Pacific during a hurricane somehow affects El Niño remains unknown, partly because there isn’t enough quality data, according to Jansen.

Santiago native Eduardo Kausel, a professor in MIT's Department of Civil and Environmental Engineering (CEE), is an expert in structural dynamics and earthquake engineering. In an interview with MIT News, Kausel explains why Chile’s stronger earthquake led to less-catastrophic damage than the earthquake that struck Haiti in January. He also explains some of the risks that could be associated with a sizable earthquake in Boston, and why the philosophy behind building codes may be changing.

Q. Even though the 8.8-magnitude earthquake that rocked Chile was 500 times stronger than the 7.0-magnitude earthquake that struck Haiti in January, the scope of damage was significantly less. Describe for us some of the differences between Chile and Haiti that helped limit loss of life and property.

A. Chile possesses an educated middle class and ranks among the most developed nations in South America whereas Haiti is the poorest country in the Western Hemisphere. Clearly, there is a strong correlation between poverty and quality of housing and infrastructure. It is this difference more than the proximity of the epicenter of the earthquake that led to massive damage in Haiti.

The most fundamental reason for the difference in damage and casualties is that Chile has been taking into account the effect of earthquakes and designing for them since at least the beginning of the 20th century. Chile's building codes are comparable to those of the United States, Japan, Turkey or Mexico, and rank among the most stringent and demanding. They have to, because strong earthquakes are a fact of life in Chile.

In Haiti, however, virtually no construction is earthquake-proof, not even government buildings or the houses of the affluent. This may relate to the fact that although Haiti is in a seismically active zone, strong earthquakes there are much less frequent than in Chile and have return periods measured in centuries, not decades.

Q. In the U.S., earthquakes tend to be associated with the West Coast, but strong earthquakes have struck the Midwest and Northeast over the last few hundred years. In your view, are these areas of the country prepared to withstand an earthquake of considerable magnitude?

A. Considerable research is being carried out at present on how to make the Midwest safe against earthquakes. Fortunately, paleoseismology seems to suggest that mammoth intra-plate earthquakes such as the four that took place in New Madrid, Missouri, between Dec. 16, 1811 and Feb. 7, 1812, which rank among the strongest in the Midwest in historic times, may be rare — although strong quakes on the order of magnitude 6 or so could be expected to occur sometime this century. Still, the problem is not only technical, but also economical, for any upgrading of the large inventory of old, low-rise, unreinforced masonry structures from Chicago through St. Louis to Memphis would entail enormous costs, which the public would have to weigh against the low risk.

A somewhat different story is that of quakes in the Northeast from Boston to Canada. Although not as strong or frequent as those in California, they may not be so rare either. For example, some strong earthquakes have recently taken place in the Quebec province, but having occurred in largely uninhabited areas, they have been inconsequential. On the other hand, a repeat of the 1755 Cape Ann earthquake some 50 miles to the northeast of Boston could conceivably produce substantial damage to the unreinforced red stone buildings of Back Bay, an area that was reclaimed from the Atlantic Ocean and has very soft ground conditions. But probably the more important risk factor there may be the gas lines embedded in that soft infill soil whose possible rupture could lead to fires.

It is worth keeping in mind that modern high rise buildings in Boston and elsewhere in the East and Midwest are very safe indeed, for they not only account for seismic considerations, but have been designed to resist the enormous overturning forces caused by strong winds or even hurricanes, which are much more frequent than earthquakes in this region.

Q. What can we to do to prevent significant damage in the U.S.? How can building codes and processes be improved further?

A. Over time, all codes continuously evolve, reflecting the lessons learned from past design mistakes, most of which were not a-priori obvious. Until recently, the goal of seismic codes was to protect human life, not the buildings themselves. This philosophy may be gradually changing now. It has been argued that the economic loss to the affected region or nation can be far greater than the aggregate of the physical losses. An example is the massive damage to the port facilities in 1995 by an earthquake in Kobe, Japan, which caused much of that port's shipping commerce to move elsewhere. Societal and economic considerations such as these may begin to affect seismic codes yet to come.

In Chile's case, the various bridges that failed on the highway to the south are not only a loss to the local municipalities or the Chilean highway administration, but the damage to the local economy may vastly exceed the cost of the bridges, which could have been made safer if constructed at a modest additional expense. The stricken area is the heart of the Chilean agriculture, akin to California's San Joaquin Valley. Much of the fruit consumed in winter in the U.S., not to mention the wine, comes from that area. If trucks cannot take these to the ports — many of which where also destroyed — then the produce cannot make it to our markets. Thus, the seismic codes that govern the infrastructure may be in need of upgrading. Expect the fruit and vegetable prices in the U.S. to rise sharply in the weeks ahead.

Phytoplankton are single-celled organisms that serve as the base of the marine food web and provide half the oxygen we breathe on Earth. They also play a key role in global climate change by removing carbon from the atmosphere and injecting it deep into the oceans.

Scientists study phytoplankton to understand how the tiny plants help transport elements like carbon through the environment. Although they understand much of what phytoplankton do, less is understood about why particular plankton live in particular environments and what maintains the diversity of phytoplankton.

Previous research has suggested that more diverse ecosystems may be more efficient at utilizing resources, meaning that the diversity of phytoplankton could be important for regulating the cycles of carbon and other elements in the ocean. But scientists need a better understanding of that diversity before they can understand how much carbon the ocean ultimately removes from the atmosphere.

Researchers from MIT’s cross-disciplinary Darwin Project, a collaboration between the Earth System Initiative (ESI) and the Computational and Systems Biology Initiative (CSBi) and funded by the Gordon and Betty Moore Foundation’s Marine Microbiology Initiative and NASA, have developed a computer model to simulate ecosystems in a virtual ocean, a model that could guide future field surveys of phytoplankton. They suggest that the diversity of phytoplankton species at a given location depends on the balance between the removal of species through competition for limited nutrient resources and their replacement by ocean currents, according to a paper published online Feb. 25 in Science Express.

In order to grow, phytoplankton need sunlight and nutrients like carbon, some of which comes from the carbon dioxide in the atmosphere. When phytoplankton die, some of their cells sink to the ocean floor, taking carbon away from the atmosphere and injecting it deep into the ocean through a process known as the “biological pump.” To understand the global scale of this process, scientists must learn more about the diversity of phytoplankton species.

“We feel this paper is a step toward understanding what the phytoplankton diversity is at different places in the ocean and what regulates that diversity,” said lead author Andrew D. Barton, a graduate student in the Department of Earth, Atmospheric and Planetary Sciences (EAPS).

Although future studies will have to make a more explicit link between phytoplankton diversity and the climate, Barton hopes that his group’s models could be used as a tool to inform future sampling surveys that try to map phytoplankton diversity in the ocean.

Building an ecosystem

Barton and his colleagues used a computer model developed in 2007 by co-author Mick Follows, a senior research scientist in EAPS, to study the distribution of particular phytoplankton types, as well as to observe how phytoplankton help move different elements through the oceans.

To study these cycles, Barton’s team plugged information about the traits of nearly 80 phytoplankton species, such as how fast they grow and what temperature they prefer to live in, into the computer model, which also simulates the physical circulation and currents of the ocean. After the computer progressed the virtual ocean forward for a decade, certain patterns began to appear, with more species appearing in the warm tropics and Gulf Stream regions than at colder, higher latitudes.

Barton’s team then hypothesized why those patterns occur, taking into account the circulation of the ocean in different regions, as well as the fact that growth rates depend on changes in temperature, light and nutrient concentration. They conclude that the amount of species in a given location is based on how rapidly species are removed because of competition for limited resources, and the rate at which species are returned to that location by the ocean’s currents — a balance that is also affected by the nature of the environment.

In the tropics, seasonal variations are weak, and different species can coexist for long periods. But there is less diversity at higher latitudes, where the changing seasons vary the amount of light and nutrients that phytoplankton can consume throughout the year. Here, a few highly specialized phytoplankton rapidly outcompete all others during the strong spring blooms, and this effect outweighs the rate at which the ocean’s currents can return species to these latitudes.

Barton and his colleagues also explain that a relatively large variety of phytoplankton coexist in the Gulf Stream and similar currents that constantly move and mix different species from different regions. In this case, the variability of the environment doesn’t matter, because the intensity of the currents prevents more dominant species from outcompeting other species for food.

Future mapping

Princeton ecologist Simon Levin called the research “highly original and exciting” for scaling microscopic details of the ocean to macroscopic patterns by combining fluid dynamics, ecology and evolutionary biology data into one model. He also thinks the research will be useful for planning future studies where phytoplankton are collected.

Barton and his colleagues hope their interpretation will help inform future mapping surveys of the ocean by guiding oceanographers where to look for particular patterns in phytoplankton diversity. They need new field data to test and refine their hypotheses and are currently speaking to scientists at Woods Hole, MIT and the University of Hawaii about collecting data on upcoming long-distance scientific explorations in the Pacific Ocean.

Barton’s next step is to evaluate the diversity patterns using a very high resolution version of the current computer model to examine how the ocean’s complex range of structures — small eddies, currents and fronts — provide small habitats that could enhance diversity.

Future research should also examine how the processes of extinction and evolution help maintain the diversity patterns, he said.

On March 17, the Federal Communications Commission will present the U.S. Congress with its National Broadband Plan, a set of recommendations for bringing high-speed Internet access to the millions of Americans who don’t yet have it. The plan is likely to determine the allocation of the $7.2 billion in stimulus money intended to bring broadband to rural and underserved areas, and many observers believe that the government is already planning to augment that investment with billions more in discretionary spending.

In anticipation of the plan’s unveiling, the FCC is holding a series of regional forums on particular aspects of the plan. On Monday evening, in an event co-hosted by the MIT Media Lab’s Center for Future Civic Media, Eugene Huang, director of government performance and civic engagement for the National Broadband Plan, spoke at MIT on the plan’s provisions for better engaging the public in the democratic process.

Huang began by describing the FCC’s attempts to engage the public in the creation of the broadband plan itself, describing the agency’s 35 public workshops on the topic, held between August of last year and the end of January, some of which had as many as 1,000 live attendees and another 5,000 online participants. He also commended the launch, last May, of data.gov, an online index of publicly available government data maintained by the White House. “But,” he added, “data.gov includes only a small amount of federal-government data, and we believe that all data and information that the government treats as public should be made available online, in machine-readable formats.” Huang said that the broadband plan recommends that data from the legislative and judiciary branches, too, be freely accessible online, pointing out, for example, that the federal courts’ Public Access to Court Electronic Records system, or PACER, charges for access to judicial decisions. In fact, Huang said, “The U.S. federal courts themselves pay private contractors $150 million annually for electronic access to judicial documents.”

Huang also said that the plan urges the creation of a public archive of historically significant video — a kind of YouTube for policy deliberations and news footage. To that end, he said, the FCC will call on Congress to revise copyright law to make it easier for news organizations to donate historical footage to the archive.

In addition, he emphasized that the federal government should make better use of social media, pointing to the success of the Centers for Disease Control in using Twitter, YouTube, podcasts, and other social-networking technologies to disseminate information about the H1N1 flu outbreak.

When he turned to the topic of how government can draw citizens into the deliberative process, rather than simply providing them with better information, Huang became a little more vague: “Government is just beginning to think about these types of issues,” he acknowledged. But in thinking about how to use digital tools to directly engage the citizenry, he said, the government is using digital tools to directly engage the citizenry. The White House’s Open Government Initiative, Huang said, has used what he described as “public brainstorming blogs, a wiki, and a collaborative drafting tool” to solicit public participation in determining just what its project should be.

Huang grew more specific, however, in describing how Internet access could facilitate the process of voter registration and make voters’ records more portable. “One recent study estimates that voter registration problems resulted in more than two millions voters’ being unable to vote in the 2008 general election,” Huang said. “Providing broadband to more Americans provides an important opportunity to fix the problems in the existing process.”

tk

It’s Wednesday evening, so Alexander Slocum is hard at work cooking up a huge batch of spaghetti, green beans and garlic bread in a cramped kitchen on the top floor of MIT’s Building 24. That might not be most people’s image of how an MIT professor spends part of his workday, but it’s something Slocum does every Wednesday for a hungry and appreciative group of a few dozen freshmen.

“Food is a great conversation catalyst,” Slocum explains with a smile as he lays out the dinner on a set of long tables and the students eagerly line up.

The students are all part of the Experimental Study Group, a program celebrating its 40th year at MIT. As an alternative that allows about 50 freshmen to satisfy their General Institute Requirements in an atmosphere of small, personalized classes and peer learning in small study groups, ESG generally provides a sense of community and camaraderie more typical of a small college. Slocum, director of the program since 2002, is also an alumnus of ESG himself.

Slocum ’82, MS ’83, PhD ’85, the Neil and Jane Pappalardo Professor of Mechanical Engineering, loves to tinker and build, but most of all he loves to ignite that same passion for creating new devices in other people. His enthusiasm is one reason he was named Massachusetts’ “professor of the year” in 2000, among many awards he has garnered for both research and teaching.

For more than a decade, his was the funny and effervescent voice announcing the play-by-play for MIT’s famed annual student competition of remotely operated robots, held every year as the culmination of class 2.007, “Introduction to Design and Manufacturing.” And when President Barack Obama came to campus last fall and toured labs, Slocum was the researcher who greeted the president wearing (as he usually does) a vibrant Hawaiian shirt and who explained to the commander-in-chief his novel concept to store some of the energy harvested by offshore wind farms.

Devices for doctors

One of the programs Slocum has been especially interested in is a class that he evolved from a class originally created by Prof. Guttag with CIMIT, the Boston-based Center for Integration of Medicine and Innovative Technology. Each year in that class (2.75, “Precision Machine Design”), clinicians and doctors from area hospitals propose to CIMIT a device they wish someone would invent to deal with problems they encounter in their practice.

Students can then pick a problem they would like to tackle, and form teams to develop devices to fill that need. Thanks to U.S. Army sponsorship, each team then gets a budget of about $5,000 to pursue the project. “I have weekly design-review and problem-solving meetings with the student/doctor teams, and the students’ task is to work with the doctor they’ve selected, create strategies and concepts, do a patent search, then do the research and engineering needed to build and test their solution,” Slocum explains.

Several devices developed in that class over the years have won awards (including the MIT $100K business plan competition), and some are being developed as commercial products. Sometimes, the students will advance their designs in a follow-up mechanical engineering class.

Each year, about eight new devices are developed as part of that class, and typically one ends up being described in a paper for a major journal article, and two or three are presented at medical device-design conferences. The class has produced “very important contributions to medical technology,” he says, including a low–cost, single-use robotic system for laparoscopic procedures, a novel needle for injections that can penetrate precisely to the right depth within a blood vessel or organ, and a pressure-sensing syringe.

Robert Sheridan, a doctor at Massachusetts General Hospital who has worked with Slocum’s class, recalls how students in one 2.75 class responded to a problem involving the need for negative-pressure wound therapy — creating a low-pressure atmosphere over a wound to promote healing — with a device that could be used in developing-world locations with limited resources.

“They grasped the clinical problem and developed a creative concept to address the challenge,” Sheridan says. “I had been working in my basement workshop on a concept device, and the students opened my mind to other potential approaches, which have led to the current prototype.” That prototype, further developed by Slocum’s graduate student and original team member Danielle Zurovcik, he says, is now undergoing tests in a rural hospital in Rwanda. In fact, after Zurovcik came back from Rwanda and the Haiti earthquake struck, she garnered support to make more devices and rushed off to help the earthquake victims.  

With more than a decade of experience with 2.75, Slocum plans to write up the lessons learned so that mechanical engineering instructors elsewhere can add elements of the class’s distinctiveness to their own. “It might be a useful model for others to adopt bits from,” he says.

The doctors have been quite enthusiastic about the process. Tom Brady, also a doctor from Mass General, says, “the magic is a combination of motivated students and clinicians, unique prototyping facilities, and Alex’s passion for solving medical problems. Alex is the creative force that makes this class work.”

Energy solutions

Slocum is also working with students to tackle one of the thorniest problems facing the world: meeting growing energy demand with renewable sources that don’t add greenhouse gases to the atmosphere. His major focus has been a proposal for incorporating an energy-storage system into the mooring structures for offshore floating windmill installations, connected by submerged cables to the grid onshore, so intermittent energy produced by variable winds can be stored on site and delivered just when it’s needed.

The concept involves huge concrete tanks — spheres or cylinders attached to the seafloor beneath the floating windmills — that pump out water when the windmill is producing excess power and then allow water to flow back in, through a generating turbine, when power is needed. It’s a classic pumped hydro system similar to those used for decades with dams and lakes. That’s the concept he demonstrated to President Obama last fall, and Slocum is hopeful that the idea will catch on.

Such a system, he says, could theoretically be used to provide enough power to meet all of the nation’s electricity needs. Offshore wind farms — far enough out to avoid objections about interfering with pristine views — could be arrayed in a few places along the coasts and in the Great Lakes, he says. Wind turbines in a total area of about 100 miles by 500 miles, he estimates, would produce as much electric power as the nation currently uses.

He’s also looking into ways of storing solar power, using molten salts to store the thermal energy for delivery later. With the right policies in place to promote the development and deployment of these technologies along with modular nuclear plants and liquefied coal or biomass for transportation fuels, he says, “The U.S. could essentially be energy independent in 20 years, and that would do more for world peace than anything else. We built the interstate highway system in the interest of national security, so we should be able to build a National Energy System.”

In addition to looking for innovative ways to teach students about science and engineering, and to apply that knowledge to important problems, Slocum is interested in helping them learn how to pass along their knowledge to others. “How do you ever teach someone to become a teacher?” he muses. One way, he suggests, is to have their education be as interactive as possible from the start. Most students who sign up for ESG, for example, come from smaller schools, where they are used to small classes and lots of interaction with their teachers.

And after their freshman year of immersion in that program, many of the students come back as teaching assistants, mentors, seminar leaders and tutors. Slocum thinks having a mix of freshmen, upperclassmen, graduate students, faculty and alums helps to foster the learning process. The more different kinds of people you have around, “the more serendipity can take hold,” he says.

The students tend to appreciate that kind of openness. Nevan Hanumara SM ’06, took Slocum’s 2.75 class a few years ago and now, while working on his doctorate at MIT, has returned as one of the class’s two Teaching Assistants. “He’s a sharer,” he says, “he does not hoard ideas. That’s unusual.” And Slocum encourages an open, exploratory approach: “He’s extremely creative, and also extremely nonlinear and chaotic. You can try stuff really easily.” Those who have taken his classes tend to stay in touch, Hanumara says. “He’s very much interested in our development as people, not just as students.”

One result, he says, is that Slocum has created “a lab full of people who play well together.”

Mechanical engineering graduate student Conor Walsh says simply that 2.75 was “the most interesting class I took at MIT,” and he has continued to work on the project he started in that class as part of his thesis. Working with Slocum and the doctor who suggested that device, he says, “we are developing lots of other devices in the same area as the original product. So in a way, the class spawned a new research area!”

The classes he teaches, Slocum says, where students get a chance to work on something whose real-world application to meet a significant need is clear from the outset, are based on “hard-core mechanical design.” As he finishes scarfing down a plateful of the spaghetti he just prepared, he explains the importance of this kind of hands-on approach to learning: “You can read about cooking, but then you’re still hungry,” he says. “To really satisfy your hunger, you need to get in the kitchen.”

For neuroscientists, one of the best ways to study brain activity is with a scanning technique called functional magnetic resonance imaging (fMRI), which reveals blood flow in the brain.

However, although fMRI is a powerful tool for identifying brain regions that are active during a particular task, it offers only an indirect view of what’s happening. Measuring a more direct indicator of neural activity, such as concentrations of neurotransmitters (brain chemicals that carry messages between neurons) could be much more valuable.

Now, for the first time, MIT and Caltech researchers have come up with a new type of fMRI sensor that can do just that. The two sensors, described in the Feb. 28 online edition of Nature Biotechnology, detect dopamine — a neurotransmitter involved in learning, movement control and many other brain processes.

“This new tool connects molecular phenomena in the nervous system with whole-brain imaging techniques, allowing us to probe very precise processes and relate them to the overall function of the brain and of the organism,” says Alan Jasanoff, an associate professor of biological engineering at MIT and senior author of the paper.

Dopamine holds particular interest for neuroscientists because of its role in motivation, reward, addiction and several neurodegenerative conditions, including Parkinson’s disease. The new sensors could help scientists learn more about how dopamine acts in the brain and in other organs, says Andrew Alexander, co-director of the Brain Imaging Core at the University of Wisconsin at Madison.

“Previously we really haven’t had specific biomarkers for looking at things like dopamine or other chemical neurotransmitters” with MRI, says Alexander.

Designing a new sensor

Conventional fMRI measures blood flow in the brain by tracking hemoglobin, the molecule that carries oxygen. Hemoglobin has an iron atom at its core that binds to oxygen. When bound to oxygen, hemoglobin’s magnetic properties change in a way that can be detected with MRI.

“fMRI is an extremely powerful technique for studying how the brain functions, and it’s the only way to obtain spatial information and information about when things are happening,” says Jasanoff, who also has appointments in the Departments of Brain and Cognitive Sciences and Nuclear Science and Engineering, and in the McGovern Institute for Brain Research at MIT.

However, the spatial and temporal information is imprecise. Researchers can detect increased activity in a certain area, but they can’t see what the activity is, nor can they get a high-resolution picture of which neurons are involved.

A more detailed picture of brain activity could emerge with MRI sensors specific to particular neurotransmitters. The MIT team designed sensors specifically for dopamine, but their technique could be used to create sensors for other neurotransmitters or even unrelated molecules of biological interest.

To build the new sensors, the MIT team worked with chemical engineers at Caltech, using an approach called “directed evolution.” They started with a protein called cytochrome P450, an enzyme found in most organisms that is paramagnetic (meaning it can become weakly magnetic when exposed to a magnetic field). Using a technique called error-prone PCR, which is a faulty version of the way cells naturally replicate their genes, they generated a large collection of different mutated forms of the gene.

Each mutated gene was placed into an E. coli bacterium, which produced the mutated protein. The researchers then tested each protein for its ability to bind dopamine. At the end of each round, they took the best candidate and mutated it again for a new round of improvement. At the end of five rounds, they had two sensors that would bind strongly to dopamine but not to other neurotransmitters.

“You want it to be specific to dopamine — you don’t want it to bind to dopamine and half a dozen other things,” says Jasanoff.

In studies of rats, the researchers showed that the sensor can effectively detect dopamine in the brain. However, in its current form, the dopamine probe must be injected into the brain, and the imaging is limited to the site of injection.

Bruce Jenkins, director of neurochemical imaging at the Martinos Center for Biomedical Imaging at MGH, says the new probe is “very cleverly designed,” but points out that an important challenge is yet to come: getting the molecule to cross the layer of cells that separates the brain from circulating blood. “Trying to get a charged protein across the blood-brain barrier is very tricky,” he says.

The MIT team hopes to overcome that obstacle by applying barrier disruption techniques used historically to deliver chemotherapeutic agents to the brain. They will also try to genetically program brain cells to express the sensor, so it doesn’t have to be injected.

They plan to adapt the directed evolution strategy to look for sensors for other neurotransmitters as well. If successful, that could help researchers in Jasanoff’s lab and elsewhere create a better wiring diagram of how different brain regions and neurotransmitters work together to yield behavior such as learning, memory, addiction and movement.

“We hope to develop probes that target different parts of the mechanism, allowing us to piece these systems together in a way that’s noninvasive,” says Jasanoff.

What difference does a great doctor make to your health? Patients everywhere would love to know the answer.

A recent study co-authored by Joseph Doyle, an economist at the MIT Sloan School of Management, offers a subtle conclusion to this question. Treatment by a highly rated physician does not necessarily change the outcome of a serious medical problem. Instead, the best doctors typically offer an accurate diagnosis more quickly than moderately rated doctors, leading to hospital stays for patients that are 10 percent shorter and less expensive — an average that increases to 25 percent for certain medical specialties.

“As a patient myself, I always hope to go to a prestigious hospital, but I wonder how much more of an advantage that is,” says Doyle. “It turns out that if you don’t have access to the most prestigious teams, the less prestigious ones will eventually make the same types of interventions, but it just takes them longer to get there, and it’s more costly.” These findings figure to resonate at a time when the cost of health care is a major political preoccupation.

To reach this conclusion, Doyle — along with his colleagues Steven Ewer of the University of Wisconsin and Todd Wagner of Stanford University — examined roughly 70,000 treatment episodes involving 30,000 patients, spread over 13 years, at a Veterans Affairs hospital in a large city in the United States. The hospital’s practices naturally lent themselves to a comparison of doctor quality since the institution randomly assigned patients to two separate teams of physicians and residents, which had markedly different medical backgrounds.

One of these teams (dubbed “Program A” by the researchers) consisted of members trained at an elite U.S. medical school, which sometimes boasts the nation’s highest average MCAT scores among its incoming students. The other group (“Program B”) has members trained at a middle-ranked medical school. Medical residents with Program A had medical board-certification scores that on average placed them in the top quarter of the national results, while the Program B doctors had scores placing them in the bottom fifth of U.S. residency programs. (The researchers agreed to keep the identities of the VA hospital and medical schools anonymous.)

Money, not mortality

Despite these differences, in some ways the bottom-line results for patients were similar regardless of whether they were treated by doctors in Program A or Program B. The mortality rates for the two programs were within a percentage point of each other, as measured over 30 days, one year and five years from the time each patient was treated.

“I find that to be a feel-good result,” Doyle says.

As Doyle, Ewer, and Wagner see it, the major difference between the teams involved the ease and confidence with which the more highly regarded doctors made their diagnoses. The doctors in Program B, the ones from the lower-ranked medical school, ordered 8 percent more tests than their counterparts in Program A, and on average took 8 percent longer to request an initial test for a patient. These differences were more pronounced within certain specialties. For instance, the Program B doctors took 21 percent longer to order heart exams, 51 percent longer to request an angiography, and 32 percent longer to order a cardiac stress test, for patients with congestive heart failure. Such delays have a direct impact on the overall cost of treatment, since they result in longer hospital stays for patients. Moreover, laboratory expenditures were 13 percent higher or patients in program B.

The doctors in Program B also consulted with specialists more often, which can also prolong the duration of a hospital visit. “Sometimes people look at the use of specialists as waste or excessive cost,” says Doyle. “But maybe these [lesser-ranked] physicians need specialists to achieve the same outcomes.”

Other health-care economists find the study a useful look at a complex question. “I think this has been on the minds of people trying to fix health care: What goes on in that black box inside the heads of doctors?” says Jonathan Skinner, an economist who teaches at both Dartmouth College and Dartmouth Medical School. “It speaks more broadly to why we see greater medical costs in some areas — it may be the difficulties physicians are having making a diagnosis.”

Only so much room at the top

To be sure, there will always be differences among doctors; in any group of physicians, some will have better training or be more highly regarded than others. By quantifying the disparities among doctors, however, and linking them to particular practices, Doyle’s work provides a yardstick for medical professionals who would like to reduce the gap between excellent and average doctors in absolute terms.

As Doyle notes, a basic caveat to the study is that it examines just one VA hospital (the researchers are looking for others featuring the same arrangements). Nonetheless, Skinner, for one, thinks the paper will have a significant impact among policymakers. “The larger point this speaks to in terms of policy is that measurement is really important in understanding what physicians do,” he says. “If you take a bunch of physicians and train them, you cannot assume they will all do the same things. It would be nice to have feedback mechanisms where doctors and residents could sit down and observe what is going on. They might change what they do.”

A question central to research on global warming is how warmer temperatures caused by increased greenhouse gases could influence climate. Probing the past for clues about this potential effect, MIT and Yale climate scientists examined the Pliocene period, which began five million years ago and which some consider to be a potential analog to modern greenhouse conditions. They found that hurricanes influenced by weakened atmospheric circulation — possibly related to high levels of carbon dioxide — contributed to very warm temperatures in the Pacific Ocean, which in turn led to more frequent and intense hurricanes. The research indicates that Earth’s climate may have multiple states based on this feedback cycle, meaning that the climate could change qualitatively in response to the effects of global warming.

Although scientists know that the early Pliocene had carbon dioxide concentrations similar to those of today, it has remained a mystery what caused the high levels of greenhouse gas and how the Pliocene’s warm conditions, including an extensive warm pool in the Pacific Ocean and temperatures that were roughly 4 degrees C higher than today’s, were maintained.

In a paper published Feb. 25 in Nature, Kerry Emanuel, the Breene M. Kerr Professor of Atmospheric Science in the Department of Earth, Atmospheric and Planetary Science, and two colleagues from Yale University’s Department of Geology and Geophysics suggest that a positive feedback between tropical cyclones — commonly called hurricanes and typhoons — and the circulation in the Pacific could have been the mechanism that enabled the Pliocene’s warm climate.

The Pliocene ended around three million years ago with the onset of large ice sheets in the Northern Hemisphere. There has been a slow reduction in carbon dioxide levels in the atmosphere for about 15 million years, and it is thought that the start of the glacial cycles was the climate’s response once those levels reached a certain threshold, according to co-author Chris Brierley. While that level remains unknown, this research indicates that by increasing carbon dioxide levels, humans could reach the threshold that would induce a Pliocene-like climate.

By combining a hurricane model and coupled ocean-atmosphere general circulation model to investigate the early Pliocene, Emanuel, Brierley and co-author Alexey Fedorov observed how vertical ocean mixing by hurricanes near the equator caused shallow parcels of water to heat up and later resurface in the eastern equatorial Pacific as part of the ocean wind-driven circulation. The researchers conclude from this pattern that frequent hurricanes in the central Pacific likely strengthened the warm pool in the eastern equatorial Pacific, which in turn increased hurricane frequency — an interaction described by Emanuel as a “two-way feedback process.”

The researchers believe that in addition to creating more hurricanes, the intense hurricane activity likely created a permanent El Niño-like state in which very warm water in the eastern Pacific near the equator extended to higher latitudes. The El Niño weather pattern, which is caused when warm water replaces cold water in the Pacific, can impact the global climate by intermittently altering atmospheric circulation, temperature and precipitation patterns.

The research suggests that Earth’s climate system may have at least two states — the one we currently live in that has relatively few tropical cyclones and relatively cold water, including in the eastern part of the Pacific, and the one during the Pliocene that featured warm sea surface temperatures, permanent El Niño conditions and high tropical cyclone activity.

Although the paper does not suggest a direct link with current climate models, Fedorov said it is possible that future global warming could cause Earth to transition into a different equilibrium state that has more hurricanes and permanent El Niño conditions. “So far, there is no evidence in our simulations that this transition is going to occur at least in the next century. However, it’s still possible that the condition can occur in the future.”

Whether our future world is characterized by a mean state that is more El Niño-like remains one of the most important unanswered questions in climate dynamics, according to Matt Huber, a professor in Purdue University’s Department of Earth and Atmospheric Sciences. He praised the research, saying it is the “very specific predictions it makes about how cyclones can warm the eastern equatorial Pacific that is the most unique and exciting.”

Reconstructing the Pliocene

To investigate the hurricane activity of the Pliocene, the researchers relied on proxy ocean temperatures based on variations in several chemical tracers in drill cores of the ocean floor. The exact value of these chemical properties is known to correlate well with sea surface temperatures in the modern ocean and are used as proxies for past temperatures. Using these proxy temperatures, the researchers reconstructed the global distribution of sea surface temperatures. They then plugged these sea surface temperatures into an ocean-atmosphere general circulation model, which is a computer-based mathematical model used for climate forecasting, to determine the Pliocene’s atmospheric conditions. This showed that the early Pliocene had a weakened atmospheric circulation, and therefore, reduced vertical wind shear, which is favorable for tropical cyclone growth.

Next, Emanuel entered data from the large-scale climate model of the Pliocene into a Statistical DownScaling Model (SDSM), which is software used to derive regional climate information, such as hurricane activity, based on global climate data. By producing synthetic hurricane tracks with the SDSM, researchers can study the effects of hurricanes on ocean temperatures in different regions.

Their observations included nearly twice the number of tropical cyclones than  occur in our current climate system, including storms with lifespans that averaged two to three days longer than our current system. The hurricanes appeared in places, such as Hawaii, that differ from where they typically occur today, and also occurred throughout the seasons.

The researchers traced this storm activity to the expansion of the warm pool in the eastern equatorial Pacific that resulted from resurfaced parcels of warm water created by the hurricanes — essentially, both the cause and effect of the observed increased tropical cyclones.

Fine-tuning the theory

Additional research will focus on why the Pliocene was so warm at higher latitudes, including an iceless North Pole, and whether this resulted from moisture produced by the tropical cyclones, Fedorov said.

Brierley hopes to develop an interactive model to strengthen the group’s theory. Rather than examining individual components, such as sea surface temperatures, and then imposing that data onto a model to figure out potential ocean mixing and hurricane activity, the researchers would like to include everything in the same interactive model.

Resolving other issues, such as how to more precisely estimate the contribution of tropical cyclones to ocean mixing, will not only help improve the early Pliocene climate model, but also help predict future climate change for which the feedback between hurricanes and the ocean circulation could be crucial.

Microelectromechanical devices — tiny machines with moving parts — are everywhere these days: they monitor air pressure in car tires, register the gestures of video game players, and reflect light onto screens in movie theaters. But they’re manufactured the same way computer chips are, in facilities that can cost billions of dollars, and their rigidity makes them hard to wrap around curved surfaces.

MIT researchers have discovered a way to make microelectromechanical devices, or MEMS, by stamping them onto a plastic film. That should significantly reduce their cost, but it also opens up the possibility of large sheets of sensors that could, say, cover the wings of an airplane to gauge their structural integrity. The printed MEMS are also flexible, so they could be used to make sensors with irregular shapes. And since the stamping process dispenses with the harsh chemicals and high temperatures ordinarily required for the fabrication of MEMS, it could allow MEMS to incorporate a wider range of materials.

Conventional MEMS are built through a process called photolithography, in which different layers of material are chemically deposited on a substrate — usually a wafer of some semiconducting material — and etched away to form functional patterns. Since a wafer is at most 12 inches across, arranging today’s MEMS into large arrays requires cutting them out and bonding them to some other surface.

Instead of using a wafer, the MIT researchers begin with a grooved sheet of a rubbery plastic, which is coated with the electrically conductive material indium tin oxide. The researchers use what they call a “transfer pad” to press a thin film of metal against the grooved plastic. Between the metal film and the pad is a layer of organic molecules that weaken the metal’s adhesion to the pad. If the researchers pull the pad away fast enough, the metal remains stuck to the plastic.

“It’s kind of similar to if you have Scotch tape on a piece of paper,” says Corinne Packard, a postdoc in the Research Lab of Electronics at MIT who led the work, along with professors of electrical engineering Vladimir Bulović and Martin Schmidt. “If you peel it off slowly, you can delaminate the tape very easily. But if you peel fast, you’ll rip the paper.”

Once the transfer pad has been ripped away, the metal film is left spanning the grooves in the plastic like a bridge across a series of ravines. Applying a voltage between the indium-tin-oxide coating and the film can cause it to bend downward, into the groove in the plastic: the film becomes an “actuator” — the moving part in a MEMS. Varying the voltage would cause the film to vibrate, like the diaphragm of a loudspeaker; selectively bending different parts of the film would cause them to reflect light in different ways; and dramatically bending the film could turn a smooth surface into a rough one. Similarly, if pressure is applied to the metal film, it will generate an electric signal that the researchers can detect. The film is so thin that it should be able to register the pressure of sound waves.

Serendipity

The discovery of the manufacturing technique, which the MIT team describes in a forthcoming issue of the journal Advanced Materials, was a happy accident. The researchers were actually trying to use a printing technique to build an electrical circuit. They had created a plastic stamp with a pattern molded into it and were trying to transfer that pattern to a thin sliver film. They had expected that the plastic would pull away the silver it made contact with, leaving behind an electrode that could control an organic light-emitting diode.

Instead, however, the stamp kept pulling away the entire silver film. “The first couple times we did this, we were like, ‘Ah! Bummer, man,’” says Bulović. “And then a light bulb went off, and we said, ‘Well, but we just made the world’s first printed MEM.’” The stamp was intended as a means of creating an electronic device; instead, it ended up serving as the basis for a device itself. The researchers’ ensuing work was on the ideal architecture for the device and on ways to minimize the metal film’s adhesion to the transfer pad and maximize its adhesion to the grooved plastic.

Because the researchers hadn’t set out to make MEMS, and because, to their knowledge, their films constitute the first stamped MEMS devices, they’re still trying to determine the ideal application of the technology. Sheets of sensors to gauge the structural integrity of aircraft and bridges are one possibility; but the MEMS could also change the physical texture of the surfaces they’re applied to, altering the airflow over a wing, or modifying the reflective properties of a building’s walls or windows. A sheet of thousands of tiny microphones could determine, from the difference in the time at which sound waves arrive at different points, where a particular sound originated. Such a system could filter out extraneous sounds in a noisy room, or even perform echolocation, the way bats do. The same type of sheet could constitute a paper-thin loudspeaker; the vibrations of different MEMS might even be designed to interfere with each other, so that transmitted sounds would be perfectly audible at some location but inaudible a few feet away. The technology could also lead to large digital displays that could be rolled up when not in use.

John Rogers, a researcher at the University of Illinois at Urbana-Champaign who has pioneered techniques for printable electrical circuits, is particularly intrigued by the idea that printable MEMS could incorporate materials that are incompatible with existing MEMS manufacturing processes. “The ability to do heterogeneous integration of different material types into micromachines is a neat capability that would be enabled by this form of manufacturing,” Rogers says. “It opens up new design opportunities because it relaxes constraints on choices of materials.” And in general, Rogers says, the idea of printing MEMS is “cool.” “What they’ve done in this paper is demonstrated, for the first time, to my knowledge, this kind of concept.”