Artificial Intelligence

Deep neural networks excel at finding patterns in datasets too vast for the human brain to pick apart. That ability has made deep learning indispensable to just about anyone who deals with data. This year, the MIT Quest for Intelligence and the MIT-IBM Watson AI Lab sponsored 17 undergraduates to work with faculty on yearlong research projects through MIT’s Advanced Undergraduate Research Opportunities Program (SuperUROP).

Students got to explore AI applications in climate science, finance, cybersecurity, and natural language processing, among other fields. And faculty got to work with students from outside their departments, an experience they describe in glowing terms. “Adeline is a shining testament of the value of the UROP program,” says Raffaele Ferrari, a professor in MIT’s Department of Earth and Planetary Sciences, of his advisee. “Without UROP, an oceanography professor might have never had the opportunity to collaborate with a student in computer science.”

Highlighted below are four SuperUROP projects from this past year.

A faster algorithm to manage cloud-computing jobs

The shift from desktop computing to far-flung data centers in the “cloud” has created bottlenecks for companies selling computing services. Faced with a constant flux of orders and cancellations, their profits depend heavily on efficiently pairing machines with customers.

Approximation algorithms are used to carry out this feat of optimization. Among all the possible ways of assigning machines to customers by price and other criteria, they find a schedule that achieves near-optimal profit.​ For the last year, junior Spencer Compton worked on a virtual whiteboard with MIT Professor Ronitt Rubinfeld and postdoc Slobodan Mitrović to find a faster scheduling method.

“We didn’t write any code,” he says. “We wrote proofs and used mathematical ideas to find a more efficient way to solve this optimization problem. The same ideas that improve cloud-computing scheduling can be used to assign flight crews to planes, among other tasks.”

In a pre-print paper on arXiv, Compton and his co-authors show how to speed up an approximation algorithm under dynamic conditions. They also show how to locate machines assigned to individual customers without computing the entire schedule.

A big challenge was finding the crux of the project, he says. “There’s a lot of literature out there, and a lot of people who have thought about related problems. It was fun to look at everything that’s been done and brainstorm to see where we could make an impact.”​

How much heat and carbon can the oceans absorb?

Earth’s oceans regulate climate by drawing down excess heat and carbon dioxide from the air. But as the oceans warm, it’s unclear if they will soak up as much carbon as they do now. A slowed uptake could bring about more warming than what today’s climate models predict. It’s one of the big questions facing climate modelers as they try to refine their predictions for the future.

The biggest obstacle in their way is the complexity of the problem: today’s global climate models lack the computing power to get a high-resolution view of the dynamics influencing key variables like sea-surface temperatures. To compensate for the lost accuracy, researchers are building surrogate models to approximate the missing dynamics without explicitly solving for them.

In a project with MIT Professor Raffaele Ferrari and research scientist Andre Souza, MIT junior Adeline Hillier is exploring how deep learning solutions can be used to improve or replace physical models of the uppermost layer of ocean, which drives the rate of heat and carbon uptake. “If the model has a small footprint and succeeds under many of the physical conditions encountered in the real world, it could be incorporated into a global climate model and hopefully improve climate projections,” she says.

In the course of the project, Hillier learned how to code in the programming language Julia. She also got a crash course in fluid dynamics. “You’re trying to model the effects of turbulent dynamics in the ocean,” she says. “It helps to know what the processes and physics behind them look like.”

In search of more efficient deep learning models

There are thousands of ways to design a deep learning model to solve a given task. Automating the design process promises to narrow the options and make these tools more accessible. But finding the optimal architecture is anything but simple. Most automated searches pick the model that maximizes validation accuracy without considering the structure of the underlying data, which may suggest a simpler, more robust solution. As a result, more reliable or data-efficient architectures are passed over.

“Instead of looking at the accuracy of the model alone, we should focus on the structure of the data,” says MIT senior Kristian Georgiev. In a project with MIT Professor Asu Ozdaglar and graduate student Alireza Fallah, Georgiev is looking at ways to automatically query the data to find the model that best suits its constraints. “If you choose your architecture based on the data, you’re more likely to get a good and robust solution from a learning theory perspective,” he says.

The hardest part of the project was the exploratory phase at the start, he says. To find a good research question he read through papers ranging from topics in autoML to representation theory. But it was worth it, he says, to be able to work at the intersection of optimization and generalization. “To make good progress in machine learning you need to combine both of these fields.”

What makes humans so good at recognizing faces?

Face recognition comes easily to humans. Picking out familiar faces in a blurred or distorted photo is a cinch. But we don’t really understand why or how to replicate this superpower in machines. To home in on the principles important to recognizing faces, researchers have shown headshots to human subjects that are progressively degraded to see where recognition starts to break down. They are now performing similar experiments on computers to see if deeper insights can be gained

In a project with MIT Professor Pawan Sinha and the MIT Quest for Intelligence, junior Ashika Verma applied a set of filters to a dataset of celebrity photos. She blurred their faces, distorted them, and changed their color to see if a face-recognition model could pick out photos of the same face. She found that the model did best when the photos were either natural color or grayscale, consistent with the human studies. Accuracy slipped when a color filter was added, but not as much as it did for the human subjects — a wrinkle that Verma plans to investigate further.

The work is part of a broader effort to understand what makes humans so good at recognizing faces, and how machine vision might be improved as a result. It also ties in with Project Prakash, a nonprofit in India that treats blind children and tracks their recovery to learn more about the visual system and brain plasticity. “Running human experiments takes more time and resources than running computational experiments,” says Verma’s advisor, Kyle Keane, a researcher with MIT Quest. “We’re trying to make AI as human-like as possible so we can run a lot of computational experiments to identify the most promising experiments to run on humans.”

Degrading the images to use in the experiments, and then running them through the deep nets, was a challenge, says Verma. “It’s very slow,” she says. “You work 20 minutes at a time and then you wait.” But working in a lab with an advisor made it worth it, she says. “It was fun to dip my toes into neuroscience.”

SuperUROP projects were funded, in part, by the MIT-IBM Watson AI Lab, MIT Quest Corporate, and by Eric Schmidt, technical advisor to Alphabet Inc., and his wife, Wendy. More

Not so long ago, watching a movie on a smartphone seemed impossible. Vivienne Sze was a graduate student at MIT at the time, in the mid 2000s, and she was drawn to the challenge of compressing video to keep image quality high without draining the phone’s battery. The solution she hit upon called for co-designing energy-efficient circuits with energy-efficient algorithms.

Sze would go on to be part of the team that won an Engineering Emmy Award for developing the video compression standards still in use today. Now an associate professor in MIT’s Department of Electrical Engineering and Computer Science, Sze has set her sights on a new milestone: bringing artificial intelligence applications to smartphones and tiny robots.

Her research focuses on designing more-efficient deep neural networks to process video, and more-efficient hardware to run those applications. She recently co-published a book on the topic, and will teach a professional education course on how to design efficient deep learning systems in June.

On April 29, Sze will join Assistant Professor Song Han for an MIT Quest AI Roundtable on the co-design of efficient hardware and software moderated by Aude Oliva, director of MIT Quest Corporate and the MIT director of the MIT-IBM Watson AI Lab. Here, Sze discusses her recent work.

Q: Why do we need low-power AI now?

A: AI applications are moving to smartphones, tiny robots, and internet-connected appliances and other devices with limited power and processing capabilities. The challenge is that AI has high computing requirements. Analyzing sensor and camera data from a self-driving car consumes about 2,500 watts, but the computing budget of a smartphone is just about a single watt. Closing this gap requires rethinking the entire stack, a trend that will define the next decade of AI.

Q: What’s the big deal about running AI on a smartphone?

A: It means that the data processing no longer has to take place in the “cloud,” on racks of warehouse servers. Untethering compute from the cloud allows us to broaden AI’s reach. It gives people in developing countries with limited communication infrastructure access to AI. It also speeds up response time by reducing the lag caused by communicating with distant servers. This is crucial for interactive applications like autonomous navigation and augmented reality, which need to respond instantaneously to changing conditions. Processing data on the device can also protect medical and other sensitive records. Data can be processed right where they’re collected.

Q: What makes modern AI so inefficient?

A: The cornerstone of modern AI — deep neural networks — can require hundreds of millions to billions of calculations — orders of magnitude greater than compressing video on a smartphone. But it’s not just number crunching that makes deep networks energy-intensive — it’s the cost of shuffling data to and from memory to perform these computations. The farther the data have to travel, and the more data there are, the greater the bottleneck.

Q: How are you redesigning AI hardware for greater energy efficiency?

A: We focus on reducing data movement and the amount of data needed for computation. In some deep networks, the same data are used multiple times for different computations. We design specialized hardware to reuse data locally rather than send them off-chip. Storing reused data on-chip makes the process extremely energy-efficient.  

We also optimize the order in which data are processed to maximize their reuse. That’s the key property of the Eyeriss chip that I co-designed with Joel Emer. In our followup work, Eyeriss v2, we made the chip flexible enough to reuse data across a wider range of deep networks. The Eyeriss chip also uses compression to reduce data movement, a common tactic among AI chips. The low-power Navion chip that I co-designed with Sertac Karaman for mapping and navigation applications in robotics uses two to three orders of magnitude less energy than a CPU, in part by using optimizations that reduce the amount of data processed and stored on-chip. 

Q: What changes have you made on the software side to boost efficiency?

A: The more that software aligns with hardware-related performance metrics like energy efficiency, the better we can do. Pruning, for example, is a popular way to remove weights from a deep network to reduce computation costs. But rather than remove weights based on their magnitude, our work on energy-aware pruning suggests you can remove the more energy-intensive weights to improve overall energy consumption. Another method we’ve developed, NetAdapt, automates the process of adapting and optimizing a deep network for a smartphone or other hardware platforms. Our recent followup work, NetAdaptv2, accelerates the optimization process to further boost efficiency.

Q: What low-power AI applications are you working on?

A: I’m exploring autonomous navigation for low-energy robots with Sertac Karaman. I’m also working with Thomas Heldt to develop a low-cost and potentially more effective way of diagnosing and monitoring people with neurodegenerative disorders like Alzheimer’s and Parkinson’s by tracking their eye movements. Eye-movement properties like reaction time could potentially serve as biomarkers for brain function. In the past, eye-movement tracking took place in clinics because of the expensive equipment required. We’ve shown that an ordinary smartphone camera can take measurements from a patient’s home, making data collection easier and less costly. This could help to monitor disease progression and track improvements in clinical drug trials.

Q: Where is low-power AI headed next?

A: Reducing AI’s energy requirements will extend AI to a wider range of embedded devices, extending its reach into tiny robots, smart homes, and medical devices. A key challenge is that efficiency often requires a tradeoff in performance. For wide adoption, it will be important to dig deeper into these different applications to establish the right balance between efficiency and accuracy. More

Following a year that demonstrated the importance and practical applications of scientific advancement and invention, the Lemelson-MIT Program announced seven winners of its annual 2021 Lemelson-MIT Student Prize on April 26, World Intellectual Property Day. The program awarded a total of $90,000 to four graduate students and three undergraduate teams from across the country. The majority of winners have filed for patents, while others have been awarded full or provisional patents. Their inventions range from an innovative approach to plastic pollution in Uganda to self-driving wheelchair technology.

“We are thrilled with and inspired by the quality of inventions this year,” says Michael J. Cima, faculty director of the Lemelson-MIT Program and associate dean of innovation at the MIT School of Engineering. “This group of students has performed tremendous work amidst difficult circumstances, often working remotely, knowing their research is too important to slow down. Science and technology have been at the forefront of conversation over the past year, and this diverse group of students is well-positioned to lead us toward great advances for years to come,” Cima says.

Supported by The Lemelson Foundation and administered by the School of Engineering, the Lemelson-MIT Student Prize recognizes and provides catalyst funding to young inventors who have dedicated themselves to providing scalable solutions to real-world problems around the globe. This year’s winners have invented solutions that address pregnancy-related complications, market losses in the agricultural industry, obstacles impeding smooth patient recoveries, and other pressing problems in society. Recipients were selected from a diverse and highly competitive pool of hundreds of applicants from colleges and universities across the United States. 

“Congratulations to this year’s winners for their remarkable achievements and dedication to solving some of the biggest challenges facing society today,” says Carol Dahl, executive director of the Lemelson Foundation. “It’s particularly exciting to see this year’s cohort of graduate winners is all women, given the fact that a large gender disparity exists in patenting. More inventors are needed from communities historically underrepresented in invention, including women, if we are going to effectively solve the challenges of today and tomorrow.”

2021 Lemelson-MIT Student Prize winners were selected based on the overall inventiveness of their work, the invention’s potential for scalable commercialization or adoption, and youth mentorship experience. They are:

The “Cure it!” Lemelson-MIT Student Prize: Rewarding technology-based inventions that involve health care.

•    Nicole Black of Harvard University, $15,000 Graduate Winner The eardrum often becomes damaged through traumatic head injuries, blast injuries, chronic ear infections, and other incidents, affecting millions of people worldwide every year. Current eardrum graft materials are tissues taken from other parts of the body. These current grafts intend to repair damage, yet do not integrate well with the eardrum and surrounding tissue, resulting in poor healing and hearing outcomes that often require further surgery. Using novel biodegradable materials and 3D printing techniques, Black invented a tunable, biomimetic eardrum graft called PhonoGraft. Because PhonoGraft is able to retain the circular and radial structure of the eardrum, its sound-induced motion is similar to that of original eardrum tissue. Additionally, PhonoGraft acts as a kind of scaffolding that bridges the hole and becomes part of the native tissue, allowing the eardrum to essentially heal itself and restore hearing more effectively.

•    Mira Moufarrej of Stanford University, $15,000 Graduate WinnerPregnancy-related complications like preeclampsia and preterm delivery pose significant risks to both fetal and maternal health and are often difficult to detect in time for effective medical intervention. Moufarrej developed three novel liquid biopsy tests that monitor prenatal health and identify high-risk pregnancies by more accurately predicting due date, risk of preeclampsia, and likelihood of preterm delivery, making assessments possible well in advance of the mother becoming symptomatic. Following preclinical validation, these affordable, simple, and reliable maternal blood tests may change the standard of care for preeclampsia and preterm delivery — risks that no other test can currently predict early enough to allow for meaningful clinical intervention.

•    Innerva: Bruce Enzmann, Michael Lan, and Anson Zhou of Johns Hopkins University, $10,000 Undergraduate Team WinnerTargeted muscle reinnervation (TMR), a procedure to connect severed nerves to smaller motor nerves, is an increasingly popular method for treating peripheral nerve injuries, as it partially guides nerve regeneration and makes it possible for amputees to more effectively operate prosthetic devices. About 30 percent of TMR patients, however, experience pain due to nerve tumors, or neuromas, that result from the inherent differences in size between the newly connected nerves. Innerva’s invention is a nerve conduit that creates an interface between the different sized nerves connected during TMR, modulating nerve regeneration and preventing the formation of neuromas.

The “Eat it!” Lemelson-MIT Student Prize: Rewarding technology-based inventions that involve food/water and agriculture.

•    Hilary Johnson of MIT, $15,000 Graduate WinnerCentrifugal pumps are integral drivers in many fluid systems, such as clean water distribution, wastewater treatment, crop irrigation, oil and gas production, and pumped hydro energy storage. Requiring significant energy to operate, collectively these pumps consume 6 percent of annual U.S. electricity. Hilary’s invention is a variable volute pump, a new category of centrifugal pumps that mechanically adapts the hydraulic chamber to adjust to fluctuating system demand. Variable volute pumps show the potential to significantly improve efficiency and operating range across applications by adjusting the spiral fluid passages to match the flow rate.

•    Grain Weevil: Benjamin Johnson and Zane Zents of the University of Nebraska at Omaha, $10,000 Undergraduate Team WinnerLarge grain bins are used to store surplus grain supplies and allow farmers to hold their yield for higher prices. Managing grain condition and extraction require farmers to physically enter the grain bin, which is difficult and dangerous, often trapping and even killing farmers. A lack of proper management and extraction systems cause a 30 percent loss in cereal grain value worldwide. The Grain Weevil is a grain extraction and bin management robot that scurries across the top of the grain within a bin, smoothing out clumps so that the grain can be properly aerated and easily extracted from the bin. This device helps farmers safely and efficiently manage the extraction of grain from the bin, as well as maintain grain quality while in storage.

The “Move it!” Lemelson-MIT Student Prize: Rewarding technology-based inventions that involve transportation and mobility.

•    Adventus Robotics: Maya Burhanpurkar and Seung Hwan An of Harvard University, $10,000 Undergraduate Team WinnerPower wheelchairs present formidable barriers to mobility for users unable to operate a joystick, and manual wheelchairs operated by porters within hospitals can increase the potential for disease transmission between patients and staff. To solve these issues, the Adventus team developed a hardware and software kit that can be retrofitted to power wheelchairs already on the market to convert them into Level 5 (fully autonomous) self-driving wheelchairs. Adventus’ system transcends existing assistive technologies by using artificial intelligence and fail-safe sensors for edge detection and collision prevention. In light of Covid-19, the team’s technology has the potential to be used in a variety of other applications like autonomous floor cleaning and disinfecting.

The “Use it!” Lemelson-MIT Student Prize: Rewarding technology-based inventions that involve consumer devices and products.

•    Paige Balcom of the University of California at Berkeley, $15,000 Graduate WinnerTakataka Plastics is a technology and systems-level solution for plastic waste in Uganda that locally recycles plastic waste and creates jobs for vulnerable youth. Paige developed small-scale, locally built, low-cost machines to transform plastic waste into saleable products such as wall tiles for buildings, personal protective equipment, and consumer goods. This technology is especially innovative for PET waste because PET plastic (water and soda bottles) currently cannot be recycled anywhere in Uganda, and exporting the waste is difficult and inaccessible to most local recyclers.

Collegiate inventors interested in applying for the 2022 Lemelson-MIT Student Prize can find more information via the Lemelson-MIT Program. The 2022 Student Prize application will open in late spring 2021. More

There is a lot of activity beneath the vast, lonely expanses of ice and snow in the Arctic. Climate change has dramatically altered the layer of ice that covers much of the Arctic Ocean. Areas of water that used to be covered by a solid ice pack are now covered by thin layers only 3 feet deep. Beneath the ice, a warm layer of water, part of the Beaufort Lens, has changed the makeup of the aquatic environment.    

For scientists to understand the role this changing environment in the Arctic Ocean plays in global climate change, there is a need for mapping the ocean below the ice cover.

A team of MIT engineers and naval officers led by Henrik Schmidt, professor of mechanical and ocean engineering, is trying to understand environmental changes, their impact on acoustic transmission beneath the surface, and how these changes affect navigation and communication for vehicles traveling below the ice.

“Basically, what we want to understand is how does this new Arctic environment brought about by global climate change affect the use of underwater sound for communication, navigation, and sensing?” explains Schmidt.

To answer this question, Schmidt traveled to the Arctic with members of the Laboratory for Autonomous Marine Sensing Systems (LAMSS) including Daniel Goodwin and Bradli Howard, graduate students in the MIT-Woods Hole Oceanographic Institution Joint Program in oceanographic engineering.

With funding from the Office of Naval Research, the team participated in ICEX — or Ice Exercise — 2020, a three-week program hosted by the U.S. Navy, where military personnel, scientists, and engineers work side-by-side executing a variety of research projects and missions.

Play video

Understanding the Arctic | MIT MechE

A strategic waterway

The rapidly changing environment in the Arctic has wide-ranging impacts. In addition to giving researchers more information about the impact of global warming and the effects it has on marine mammals, the thinning ice could potentially open up new shipping lanes and trade routes in areas that were previously untraversable.

Perhaps most crucially for the U.S. Navy, understanding the altered environment also has geopolitical importance.

“If the Arctic environment is changing and we don’t understand it, that could have implications in terms of national security,” says Goodwin.

Several years ago, Schmidt and his colleague Arthur Baggeroer, professor of mechanical and ocean engineering, were among the first to recognize that the warmer waters, part of the Beaufort Lens, coupled with the changing ice composition, impacted how sound traveled in the water.

To successfully navigate throughout the Arctic, the U.S. Navy and other entities in the region need to understand how these changes in sound propagation affect a vehicle’s ability to communicate and navigate through the water.

Using an unpiloted, autonomous underwater vehicle (AUV) built by General Dynamics-Mission Systems (GD-MS), and a system of sensors rigged on buoys developed by the Woods Hole Oceanographic Institution, Schmidt and his team, joined by Dan McDonald and Josiah DeLange of GD-MS, set out to demonstrate a new integrated acoustic communication and navigation concept.

The framework, which was also supported and developed by LAMSS members Supun Randeni, EeShan Bhatt, Rui Chen, and Oscar Viquez, as well as LAMSS alumnus Toby Schneider of GobySoft LLC, would allow vehicles to travel through the water with GPS-level accuracy while employing oceanographic sensors for data collection.

“In order to prove that you can use this navigational concept in the Arctic, we have to first ensure we fully understand the environment that we’re operating in,” adds Goodwin.

Understanding the environment belowAfter arriving at the Arctic Submarine Lab’s ice camp last spring, the research team deployed a number of conductivity-temperature-depth probes to gather data about the aquatic environment in the Arctic.

“By using temperature and salinity as a function of depth, we calculate the sound speed profile. This helps us understand if the AUV’s location is good for communication or bad,” says Howard, who was responsible for monitoring environmental changes to the water column throughout ICEX.

Because of the way sound bends in water, through a concept known as Snell’s Law, sine-like pressure waves collect in some parts of the water column and disperse in others. Understanding the propagation trajectories is key to predicting good and bad locations for the AUV to operate.  

To map the areas of the water with optimal acoustic properties, Howard modified the traditional signal-to-noise-ratio (SNR) by using a metric known as the multi-path penalty (MPP), which penalizes areas where the AUV receives echoes of the messages. As a result, the vehicle prioritizes operations in areas with less reverb.

These data allowed the team to identify exactly where the vehicle should be positioned in the water column for optimal communications which results in accurate navigation.

While Howard gathered data on how the characteristics of the water impact acoustics, Goodwin focused on how sound is projected and reflected off the ever-changing ice on the surface.

To get these data, the AUV was outfitted with a device that measured the motion of the vehicle relative to the ice above. That sound was picked up by several receivers attached to moorings hanging from the ice.

The data from the vehicle and the receivers were then used by the researchers to compute exactly where the vehicle was at a given time. This location information, together with the data Howard gathered on the acoustic environment in the water, offer a new navigational concept for vehicles traveling in the Arctic Sea.

Protecting the Arctic

After a series of setbacks and challenges due to the unforgiving conditions in the Arctic, the team was able to successfully prove their navigational concept worked. Thanks to the team’s efforts, naval operations and future trade vessels may be able to take advantage of the changing conditions in the Arctic to maximize navigational accuracy and improve underwater communications.

“Our work could improve the ability for the U.S. Navy to safely and effectively operate submarines under the ice for extended periods,” Howard says.

Howard acknowledges that in addition to the changes in physical climate, the geopolitical climate continues to change. This only strengthens the need for improved navigation in the Arctic.

“The U.S. Navy’s goal is to preserve peace and protect global trade by ensuring freedom of navigation throughout the world’s oceans,” she adds. “The navigational concept we proved during ICEX will serve to help the Navy in that mission.” More

MIT students Spencer Compton, Karna Morey, Tara Venkatadri, and Lily Zhang have been selected to receive a Barry Goldwater Scholarship for the 2021-22 academic year. Over 5,000 college students from across the United States were nominated for the scholarships, from which only 410 recipients were selected based on academic merit. 

The Goldwater scholarships have been conferred since 1989 by the Barry Goldwater Scholarship and Excellence in Education Foundation. These scholarships have supported undergraduates who go on to become leading scientists, engineers, and mathematicians in their respective fields. All of the 2021-22 Goldwater Scholars intend to obtain a doctorate in their area of research, including the four MIT recipients. 

Spencer Compton

A junior majoring in computer science and engineering, Compton is set to graduate next year with both his undergraduate and master’s degrees. For Compton, solving advanced problems is as fun as it is challenging — he’s been involved in algorithm competitions since high school, where, on the U.S. team for the 2018 International Olympiad in Informatics, Compton won gold. “I still participate — there’s a college equivalent, the Intercollegiate Programming Contest or ICPC, and I’m on last year’s MIT team that won first in North America,” reports Compton. “We were supposed to represent MIT in the World Finals in Russia last summer, but it’s been postponed due to Covid.” Compton brings his competitive and enthusiastic mindset to his areas of research, including his collaboration on causal inference with the MIT-IBM Watson AI Lab, and his work on approximation algorithms and scheduling with professor of electrical engineering and computer science Ronitt Rubinfeld and postdoc Slobodan Mitrović​.

In her recommendation letter, Rubinfeld, a member of the Computer Science and Artificial Intelligence Laboratory, spoke at length about Compton’s aptitude as a student but she also left a glowing review as to Compton’s character. “Spencer is extraordinarily pleasant to work with. He is kind and caring when he interacts with younger students. I once had a high school student follow me for a day on which I happened to have a meeting with Spencer ­­— she was so impressed with him that he became a role model for her,” wrote Rubinfeld. Following the completion of his current degrees at MIT, Compton plans to obtain his PhD in computer science, continue his research in algorithms, and teach at the university level.

Karna Morey

Morey is a third-year majoring in physics with a minor in Spanish. He got interested in physics while reading Albert Einstein’s biography in the seventh grade, and performed research for two years in high school on gravitational wave physics of a body falling into a black hole. On campus, he has been involved in physics research in theoretical and observational astrophysics, as well as in condensed matter experiments. He recently authored an accepted paper on measuring the lifetime of high-redshift quasars to better understand the ways that supermassive black holes grow. Currently, he is working in the Gedik group, exploring quantum materials using second harmonic generation. Morey plans on pursuing a PhD in physics and one day conduct research at the university level.

“It was a great experience working with Karna. He was the first student I worked with and he set the bar very high for any future students,” said Christina Eilers, a Pappalardo Fellow in the MIT Department of Physics; Eilers supervised Morey’s research estimating the timescales of supermassive black holes in the early universe and was extremely impressed by his coding skills and confidence as a researcher. Morey is also heavily involved in diversity, equity, and inclusion efforts in the physics department and in the broader field, where he serves as one of the co-chairs of the cross-constituency Physics Values Committee, which seeks to work with department leadership and stakeholders to improve the climate and culture of the physics department. He hopes to make meaningful contributions not only to further scientific discoveries, but also to making science more inclusive.

Tara Venkatadri

A fourth-generation engineer and junior at MIT, Venkatadri is following her passion for space exploration, majoring in aeronautical and astronautical engineering with a minor in Earth, atmospheric, and planetary sciences. During her time at MIT, Venkatadri became interested in aerospace structures, pointing out that the unforgiving space environment places unique spacecraft constraints, especially for crewed missions. “As we go deeper into outer space and send humans to other planets, we need to design new methods and materials to ensure the safety of astronauts when pursuing increasingly ambitious space exploration,” she said.

Her interest in aerospace structures eventually landed her in the lab of Professor Tal Cohen, the Robert N. Noyce Career Development Professor and assistant professor of civil and environmental engineering and mechanical engineering. Venkatadri is trying to understand how adhesive materials deform under torsion in order to use them safely and efficiently in real-world structures, such as spacecraft. There has been increasing interest in adhesives across many industries because they can bond dissimilar materials together without welding and do not concentrate stress on the materials the way mechanical fastenings like bolts and rivets do. In his letter of recommendation, Olivier de Weck, a professor of aeronautics and astronautics and of engineering systems at MIT, cited Venkatadri’s research rigor, academic scholarship, and significant acts of service to the department, noting “without hesitation that Tara is the most impressive undergraduate student I have seen in our department over the last decade.”

Lily Zhang

Zhang is a junior double-majoring in Earth, atmospheric, and planetary sciences as well as physics, with minors in public policy and math. Zhang has a passion for climate science, something she’s known since she first viewed Al Gore’s “An Inconvenient Truth” as a child. That passion was encouraged by her father, a professor of meteorology. “He was really passionate about his research and loved his job, which helped me develop my own appreciation for science and academia,” says Zhang. Though her father passed away in 2019, Zhang says he remains a major inspiration on her life.

At MIT, Zhang is now in the finishing stages of two of her own research projects, including using satellite observations to fill in the historic Halley ozone record with Professor Susan Solomon, the Lee and Geraldine Martin Professor of Environmental Studies in the Department of Earth, Atmospheric and Planetary Sciences. “Lily never ceases to astonish me with her ability to tackle research questions and come up with clever solutions. The Goldwater scholarship is fitting recognition of her enormous potential,” said Solomon. Zhang is thankful to all of her mentors, both past and present, and says that the opportunity to work alongside them and observe their research approaches first-hand has been a dream. After finishing her undergraduate degree, Zhang aims to obtain her PhD and bring her zest for education and research as a professor in climate science.

The Barry Goldwater Scholarship and Excellence in Education Program was established by Congress in 1986 to honor Senator Barry Goldwater, a soldier and national leader who served the country for 56 years. Awardees receive scholarships of up to $7,500 a year to cover costs related to tuition, room and board, fees, and books. More

Isaac Newton may have met his match.

For centuries, engineers have relied on physical laws — developed by Newton and others — to understand the stresses and strains on the materials they work with. But solving those equations can be a computational slog, especially for complex materials.

MIT researchers have developed a technique to quickly determine certain properties of a material, like stress and strain, based on an image of the material showing its internal structure. The approach could one day eliminate the need for arduous physics-based calculations, instead relying on computer vision and machine learning to generate estimates in real time.

The researchers say the advance could enable faster design prototyping and material inspections. “It’s a brand new approach,” says Zhenze Yang, adding that the algorithm “completes the whole process without any domain knowledge of physics.”

The research appears today in the journal Science Advances. Yang is the paper’s lead author and a PhD student in the Department of Materials Science and Engineering. Co-authors include former MIT postdoc Chi-Hua Yu and Markus Buehler, the McAfee Professor of Engineering and the director of the Laboratory for Atomistic and Molecular Mechanics.

Engineers spend lots of time solving equations. They help reveal a material’s internal forces, like stress and strain, which can cause that material to deform or break. Such calculations might suggest how a proposed bridge would hold up amid heavy traffic loads or high winds. Unlike Sir Isaac, engineers today don’t need pen and paper for the task. “Many generations of mathematicians and engineers have written down these equations and then figured out how to solve them on computers,” says Buehler. “But it’s still a tough problem. It’s very expensive — it can take days, weeks, or even months to run some simulations. So, we thought: Let’s teach an AI to do this problem for you.”

The researchers turned to a machine learning technique called a Generative Adversarial Neural Network. They trained the network with thousands of paired images — one depicting a material’s internal microstructure subject to mechanical forces,  and the other depicting that same material’s color-coded stress and strain values. With these examples, the network uses principles of game theory to iteratively figure out the relationships between the geometry of a material and its resulting stresses.

“So, from a picture, the computer is able to predict all those forces: the deformations, the stresses, and so forth,” Buehler says. “That’s really the breakthrough — in the conventional way, you would need to code the equations and ask the computer to solve partial differential equations. We just go picture to picture.”

That image-based approach is especially advantageous for complex, composite materials. Forces on a material may operate differently at the atomic scale than at the macroscopic scale. “If you look at an airplane, you might have glue, a metal, and a polymer in between. So, you have all these different faces and different scales that determine the solution,” say Buehler. “If you go the hard way — the Newton way — you have to walk a huge detour to get to the answer.”

But the researcher’s network is adept at dealing with multiple scales. It processes information through a series of “convolutions,” which analyze the images at progressively larger scales. “That’s why these neural networks are a great fit for describing material properties,” says Buehler.

The fully trained network performed well in tests, successfully rendering stress and strain values given a series of close-up images of the microstructure of various soft composite materials. The network was even able to capture “singularities,” like cracks developing in a material. In these instances, forces and fields change rapidly across tiny distances. “As a material scientist, you would want to know if the model can recreate those singularities,” says Buehler. “And the answer is yes.”

The advance could “significantly reduce the iterations needed to design products,” according to Suvranu De, a mechanical engineer at Rensselaer Polytechnic Institute who was not involved in the research. “The end-to-end approach proposed in this paper will have a significant impact on a variety of engineering applications — from composites used in the automotive and aircraft industries to natural and engineered biomaterials. It will also have significant applications in the realm of pure scientific inquiry, as force plays a critical role in a surprisingly wide range of applications from micro/nanoelectronics to the migration and differentiation of cells.”

In addition to saving engineers time and money, the new technique could give nonexperts access to state-of-the-art materials calculations. Architects or product designers, for example, could test the viability of their ideas before passing the project along to an engineering team. “They can just draw their proposal and find out,” says Buehler. “That’s a big deal.”

Once trained, the network runs almost instantaneously on consumer-grade computer processors. That could enable mechanics and inspectors to diagnose potential problems with machinery simply by taking a picture.

In the new paper, the researchers worked primarily with composite materials that included both soft and brittle components in a variety of random geometrical arrangements. In future work, the team plans to use a wider range of material types. “I really think this method is going to have a huge impact,” says Buehler. “Empowering engineers with AI is really what we’re trying to do here.”

Funding for this research was provided, in part, by the Army Research Office and the Office of Naval Research. More

“MIT’s work to understand and improve human health spans decades and covers the Institute,” said W. Eric L. Grimson PhD ’80, at MIT Better World (Health), a virtual gathering in February. “More than a third of the faculty representing every department at MIT engage in research directly related to health science and innovation.” Grimson, who is MIT’s chancellor for academic advancement and the Bernard M. Gordon Professor of Medical Engineering, spoke of the many achievements of Institute scholars in the human health arena: “Serving as the hub of the densest innovation cluster in the world, MIT is nimble and inventive, particularly when it comes to the life sciences.”

MIT alumni and friends from around the globe were invited to attend the online event, which featured presentations from Institute leaders, faculty, and alumni about human health-related research at the Institute. With more than 1,000 participants from 27 countries, the evening began with video greetings from nearly a dozen alumni working in a range of health-care roles all over the world. Their graduation years spanned five decades, from 1967 to 2019.

Play video

Innovations in Human Health Main Session and Q&A

Grimson then turned the spotlight over to the presenting speakers: Daniel P. Huttenlocher SM ’84 PhD ’88, dean of the MIT Stephen A. Schwarzman College of Computing and Henry Ellis Warren (1894) Professor of Electrical Engineering and Computer Science; Mariana Arcaya MCP ’08, associate professor of urban planning and public health; and Steven Truong ’20, a Marshall Scholar studying computational biology at the University of Cambridge in England.

Huttenlocher spoke about the role of artificial intelligence in health research. Last year, he said, faculty at MIT’s Abdul Latif Jameel Clinic for Machine Learning in Health identified a new antibiotic candidate capable of killing drug-resistant bacteria. “In the search for new antibiotics, there are so many possibilities that it’s not practical to try even a small fraction of them,” he explained. “This is where machine learning comes in.”

He also discussed the Schwarzman College’s mission of educating “computing bilinguals” — “people [who] are equipped with knowledge about computing and AI in addition to their field of expertise” — and emphasized the need for experts in different disciplines to collaborate. “By truly integrating computing across MIT — that’s how we’ll make unparalleled leaps in making a better world.”

“The work we heard about tonight embodies the MIT commitment to curiosity and discovery in the pursuit of a better, healthier world.”

When the Covid-19 pandemic struck, according to Arcaya, “everyone could guess who would suffer first and most.” She explained that social epidemiologists have repeatedly demonstrated that socially vulnerable people face elevated disease risk. Through participatory action research in Massachusetts cities like Chelsea and Everett, Arcaya’s students learned that the high cost of Boston-area housing has forced many community members to live in overcrowded apartments or become transient, increasing their likelihood of exposure. Concluding that rapidly increasing home values in previously affordable neighborhoods also increased Covid-19 infection rates, Arcaya’s team made a compelling case for public policy that protects affordable housing. “Putting residents at the center of place-based research improves social science,” she said.

Truong offered a sobering statistic: People of Asian descent are three times more likely than their white counterparts to have undiagnosed diabetes, because they often lack the obesity commonly associated with the disease. “My dad was a perfect example of this,” he said. “Because he didn’t look like the ‘typical’ American with diabetes, the doctors didn’t test him for it. So he was diagnosed so late in his disease that his body had already been seriously damaged.” While his father’s death reinforced Truong’s determination to study the genetic basis of diabetes in Vietnamese people, he noted the limitations of large data resources such as the UK Biobank, which includes genetic information representative of the demographic breakdown of the UK as it currently is: 95 percent white. “I was able to kickstart something in Vietnam; hopefully, it not only sheds a little light onto these questions but also brings more awareness to this issue of representation in general,” he told the audience. “I hope you uplift those underrepresented in whatever fields you represent.”

“The work we heard about tonight,” remarked Grimson as the main program concluded, “embodies the MIT commitment to curiosity and discovery in the pursuit of a better, healthier world.” More

Strategic use of data is vital for progress in science, commerce, and even politics, but at the same time, citizens are demanding more responsible, respectful use of personal data. Internet users have never felt more helpless about how their data are being used: Surveys show that the vast majority of U.S. adults feel that they have little to no control over the data that the government and private companies collect about them. In response to these concerns, new privacy laws are being enacted in Europe, California, Virginia, and elsewhere around the world.

To conduct more-focused research and analysis of these issues, last week MIT launched a new initiative to bring state-of-the-art computer science research together with public policy expertise and engagement.

Launched on April 6, the MIT Future of Data, Trust, and Privacy initiative (FOD) will involve collaboration between experts specializing in five distinct technical areas:

database systems
applied cryptography
AI and machine learning
data portability and new information architectures; and
human-computer interaction.

In addition to technical research, FOD will provide forums for dialogue amongst MIT researchers, policymakers, and industry consortium members, with a structure similar to MIT’s 2019 AI Policy Congress, which included members of the Organization for Economic Cooperation and Development. 

Future of Data: Law, Technology and Policy

FOD is a collaboration between MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) and the MIT Internet Policy Research Initiative (IPRI). Co-director Daniel Weitzner is both a researcher at CSAIL and founding director of IPRI, and previously served as the White House deputy CTO under President Obama.

Weitzner says that one of the larger goals is to reduce the cycle time between the development of new policies and new software systems. He also hopes to work with industry to develop new privacy-preserving tools and to help steer conversations focused on “shaping the future of data governance.”

Founding member companies include American Family Insurance, Capital One, and MassMutual. Initiative Co-director Srini Devadas, a professor at MIT, says that the effort will draw on expertise across MIT in the fields of cryptography, machine learning, systems security, and public policy.

“The goal is to solve challenging problems of collaborative data analytics and machine learning where sharing data provides significant benefit to all participants, while also preserving strong privacy protections,” says Devadas.

At the launch event, CSAIL Director Daniela Rus cited MIT’s long history of work in the privacy space, from foundational work on cryptography, to IPRI and the Trust:Data Consortium, which has created tools and architectures that foster the development of a secure internet-based network of trusted data.

Member companies stressed the benefits they see in being part of this initiative as not only helping navigate a changing policy landscape but also developing technical tools to help manage the new policies, laws, and regulations more efficiently. Speaking at the launch were MassMutual’s Head of Data Adam Fox, Capital One’s Machine Learning Research Director Bayan Bruss, and American Family Insurance’s Enterprise Chief Data Officer Brad Burke.

Companies interested in participating in the new initiative can visit the CSAIL site for more information. More