3 Questions: What’s it like winning the MIT $100K Entrepreneurship Competition?

Posted on August 9, 2023 by - News

Solar power plays a major role in nearly every roadmap for global decarbonization. But solar panels are large, heavy, and expensive, which limits their deployment. But what if solar panels looked more like a yoga mat?

Such a technology could be transported in a roll, carried to the top of a building, and rolled out across the roof in a matter of minutes, slashing installation costs and dramatically expanding the places where rooftop solar makes sense.

That was the vision laid out by the MIT spinout Active Surfaces as part of the winning pitch at this year’s MIT $100K Entrepreneurship Competition, which took place May 15. The company is leveraging materials science and manufacturing innovations from labs across MIT to make ultra-thin, lightweight, and durable solar a reality.

The $100K is one of MIT’s most visible entrepreneurship competitions, and past winners say the prize money is only part of the benefit that winning brings to a burgeoning new company. MIT News sat down with Active Surface founders Shiv Bhakta, a graduate student in MIT’s Leaders for Global Operations dual-degree program within the MIT Sloan School of Management and Department of Civil and Environmental Engineering, and Richard Swartwout SM ’18 PhD ’21, an electrical engineering and computer science graduate and former Research Laboratory of Electronics postdoc and MIT.nano innovation fellow, to learn what the last couple of months have been like since they won.

Q: What is Active Surfaces’ solution, and what is its potential?

Bhakta: We’re commercializing an ultrathin film, flexible solar technology. Solar is one of the most broadly distributed resources in the world, but access is limited today. It’s heavy — it weighs 50 to 60 pounds a panel — it requires large teams to move around, and the form factor can only be deployed in specific environments.

Our approach is to develop a solar technology for the built environment. In a nutshell, we can create flexible solar panels that are as thin as paper, just as efficient as traditional panels, and at unprecedented cost floors, all while being applied to any surface. Same area, same power. That’s our motto.

When I came to MIT, my north star was to dive deeper in my climate journey and help make the world a better, greener place. Now, as we build Active Surfaces, I’m excited to see that dream taking shape. The prospect of transforming any surface into an energy source, thereby expanding solar accessibility globally, holds the promise of significantly reducing CO2 emissions at a gigaton scale. That’s what gets me out of bed in the morning.

Swartwout: Solar and a lot of other renewables tend to be pretty land-inefficient. Solar 1.0 is using low hanging fruit: cheap land next to easy interconnects and new buildings designed to handle the weight of current panels. But as we ramp up solar, those things will run out. We need to utilize spaces and assets better. That’s what I think solar 2.0 will be: urban PV deployments, solar that’s closer to demand, and integrated into the built environment. These next-generation use cases aren’t just a racking system in the middle of nowhere.

We’re going after commercial roofs, which would cover most [building] energy demand. Something like 80-90 percent of building electricity demands in the space can be met by rooftop solar.

The goal is to do the manufacturing in-house. We use roll-to-roll manufacturing, so we can buy tons of equipment off the shelf, but most roll-to-roll manufacturing is made for things like labeling and tape, and not a semiconductor, so our plan is to be the core of semiconductor roll-to-roll manufacturing. There’s never been roll-to-roll semiconductor manufacturing before.

Q: What have the last few months been like since you won the $100K competition?

Bhakta: After winning the $100K, we’ve gotten a lot of inbound contact from MIT alumni. I think that’s my favorite part about the MIT community — people stay connected. They’ve been congratulating us, asking to chat, looking to partner, deploy, and invest.

We’ve also gotten contacted by previous $100K competition winners and other startups that have spun out of MIT that are a year or two or three ahead of us in terms of development. There are a lot of startup scaling challenges that other startup founders are best equipped to answer, and it’s been huge to get guidance from them.

We’ve also gotten into top accelerators like Cleantech Open, Venture For Climatetech, and ACCEL at Greentown Labs. We also onboarded two rockstar MIT Sloan interns for the summer. Now we’re getting to the product-development phase, building relationships with potential pilot partners, and scaling up the area of our technology.      

Swartwout: Winning the $100K competition was a great point of validation for the company, because the judges themselves are well known in the venture capital community as well as people who have been in the startup ecosystem for a long time, so that has really propelled us forward. Ideally, we’ll be getting more MIT alumni to join us to fulfill this mission.

Q: What are your plans for the next year or so?

Swartwout: We’re planning on leveraging open-access facilities like those at MIT.nano and the University of Massachusetts Amherst. We’re pretty focused now on scaling size. Out of the lab, [the technology] is a 4-inch by 4-inch solar module, and the goal is to get up to something that’s relevant for the industry to offset electricity for building owners and generate electricity for the grid at a reasonable cost.

Bhakta: In the next year, through those open-access facilities, the goal is to go from 100-millimeter width to 300-millimeter width and a very long length using a roll-to-roll manufacturing process. That means getting through the engineering challenges of scaling technology and fine tuning the performance.

When we’re ready to deliver a pilotable product, it’s my job to have customers lined up ready to demonstrate this works on their buildings, sign longer term contracts to get early revenue, and have the support we need to demonstrate this at scale. That’s the goal.

AI model can help determine where a patient’s cancer arose

Posted on August 9, 2023 by - News

Decorative image

For a small percentage of cancer patients, doctors are unable to determine where their cancer originated. This makes it much more difficult to choose a treatment for those patients, because many cancer drugs are typically developed for specific cancer types.

A new approach developed by researchers at MIT and Dana-Farber Cancer Institute may make it easier to identify the sites of origin for those enigmatic cancers. Using machine learning, the researchers created a computational model that can analyze the sequence of about 400 genes and use that information to predict where a given tumor originated in the body.

Using this model, the researchers showed that they could accurately classify at least 40 percent of tumors of unknown origin with high confidence, in a dataset of about 900 patients. This approach enabled a 2.2-fold increase in the number of patients who could have been eligible for a genomically guided, targeted treatment, based on where their cancer originated.

“That was the most important finding in our paper, that this model could be potentially used to aid treatment decisions, guiding doctors toward personalized treatments for patients with cancers of unknown primary origin,” says Intae Moon, an MIT graduate student in electrical engineering and computer science who is the lead author of the new study.

Alexander Gusev, an associate professor of medicine at Harvard Medical School and Dana-Farber Cancer Institute, is the senior author of the paper, which appears today in Nature Medicine.

Mysterious origins

In 3 to 5 percent of cancer patients, particularly in cases where tumors have metastasized throughout the body, oncologists don’t have an easy way to determine where the cancer originated. These tumors are classified as cancers of unknown primary (CUP).

This lack of knowledge often prevents doctors from being able to give patients “precision” drugs, which are typically approved for specific cancer types where they are known to work. These targeted treatments tend to be more effective and have fewer side effects than treatments that are used for a broad spectrum of cancers, which are commonly prescribed to CUP patients.

“A sizeable number of individuals develop these cancers of unknown primary every year, and because most therapies are approved in a site-specific way, where you have to know the primary site to deploy them, they have very limited treatment options,” Gusev says.

Moon, an affiliate of the Computer Science and Artificial Intelligence Laboratory who is co-advised by Gusev, decided to analyze genetic data that is routinely collected at Dana-Farber to see if it could be used to predict cancer type. The data consist of genetic sequences for about 400 genes that are often mutated in cancer. The researchers trained a machine-learning model on data from nearly 30,000 patients who had been diagnosed with one of 22 known cancer types. That set of data included patients from Memorial Sloan Kettering Cancer Center and Vanderbilt-Ingram Cancer Center, as well as Dana-Farber.

The researchers then tested the resulting model on about 7,000 tumors that it hadn’t seen before, but whose site of origin was known. The model, which the researchers named OncoNPC, was able to predict their origins with about 80 percent accuracy. For tumors with high-confidence predictions, which constituted about 65 percent of the total, its accuracy rose to roughly 95 percent.

After those encouraging results, the researchers used the model to analyze a set of about 900 tumors from patients with CUP, which were all from Dana-Farber. They found that for 40 percent of these tumors, the model was able to make high-confidence predictions.

The researchers then compared the model’s predictions with an analysis of the germline, or inherited, mutations in a subset of tumors with available data, which can reveal whether the patients have a genetic predisposition to develop a particular type of cancer. The researchers found that the model’s predictions were much more likely to match the type of cancer most strongly predicted by the germline mutations than any other type of cancer.

Guiding drug decisions

To further validate the model’s predictions, the researchers compared data on the CUP patients’ survival time with the typical prognosis for the type of cancer that the model predicted. They found that CUP patients who were predicted to have cancer with a poor prognosis, such as pancreatic cancer, showed correspondingly shorter survival times. Meanwhile, CUP patients who were predicted to have cancers that typically have better prognoses, such as neuroendocrine tumors, had longer survival times.

Another indication that the model’s predictions could be useful came from looking at the types of treatments that CUP patients analyzed in the study had received. About 10 percent of these patients had received a targeted treatment, based on their oncologists’ best guess about where their cancer had originated. Among those patients, those who received a treatment consistent with the type of cancer that the model predicted for them fared better than patients who received a treatment typically given for a different type of cancer than what the model predicted for them.

Using this model, the researchers also identified an additional 15 percent of patients (2.2-fold increase) who could have received an existing targeted treatment, if their cancer type had been known. Instead, those patients ended up receiving more general chemotherapy drugs.

“That potentially makes these findings more clinically actionable because we’re not requiring a new drug to be approved. What we’re saying is that this population can now be eligible for precision treatments that already exist,” Gusev says.

The researchers now hope to expand their model to include other types of data, such as pathology images and radiology images, to provide a more comprehensive prediction using multiple data modalities. This would also provide the model with a comprehensive perspective of tumors, enabling it to predict not just the type of tumor and patient outcome, but potentially even the optimal treatment.

The research was funded by the National Institutes of Health, the Louis B. Mayer Foundation, the Doris Duke Charitable Foundation, the Phi Beta Psi Sorority, and the Emerson Collective.

Sanjoy Mitter, interdisciplinary explorer, dies at 89.

Posted on August 3, 2023 by - News

Professor of Electrical Engineering (Post-Tenure) Sanjoy Mitter died June 26th, 2023. He was 89. An expert in the theoretical foundations of systems, communication and control, Mitter contributed to significant engineering applications, most notably in the control of interconnected power systems and pattern recognition. 

Sanjoy Mitter was born in 1933 in Calcutta, India to a prominent family with a distinguished line of jurists. His paternal grandfather, Sir Binod Mitter, was a member of the Judicial Committee of Britain’s Privy Council. His paternal great grandfather was Sir Ramesh Mitter, the first Indian Chief Justice of the Calcutta High court in the nineteenth century. His maternal grandfather, Sir C. C. Ghose, was a justice of the Calcutta High Court and was several times acting Chief Justice.

He was born to father Subodh Mitter and mother Protiva Mitter née Ghose. Subodh Mitter broke with the judicial lineage of his family and went on to become an electrical engineer and industrialist. Sanjoy followed suit, receiving his BS in Mathematics in Calcutta in 1954, his B.Sc in Electrical Engineering from the Imperial College of Science and Technology, London in 1957, and his Ph.D. from the Imperial College of Science and Technology in 1965.

His career began with a 4-year stint as a Development Engineer with Brown Boveri and Co., Ltd., in Baden, Switzerland, before moving on to the Battelle Memorial Institute in Geneva, followed by Imperial College (where he served as Fellow in the Central Electricity Research Board from 1962 to 1965), and then Case Western Reserve University (where he taught from 1965 to 1969 before joining MIT). A true citizen of the world, Mitter was fluent in both French and German as well as English; his first wife, Adriana Mitter Fachini, hailed from Milan, Italy; their marriage lasted until her death in 2008 and sprawled between Cambridge and Florence, where they maintained two homes. Irvin Schick, co-editor with Theodore E. Djaferis of the seminal System Theory, Modeling, and Control: a tribute to Sanjoy Mitter (Springer Science+Business Media, 2000), wrote in his introduction to the text that “[Sanjoy and Adriana’s] whereabouts at any given time, much like the precise location of subatomic particles, can only be stated probabilistically.” 

Although MIT remained his home base, Mitter’s travels extended to multiple sojourns at other universities: he was a Professor of Mathematics at the Scuola Normale, in Pisa, Italy from 1986 to 1996, and held visiting positions at Imperial College, London; University of Groningen, Holland; INRIA, France; Tata Institute of Fundamental Research, India; ETH, Zürich, Switzerland; and several other American universities and institutions, including the University of California, Berkeley and the Los Alamos National Laboratories. However, his most profound and lasting imprints were undoubtedly upon the hundreds, if not thousands, of students, mentees, co-workers and friends he accumulated during his long and notable career at MIT. 

The CICS “red booklet” outlined the interdisciplinary nature of the center, co-run by Brown University, Harvard University, and MIT. In the initial proposal for the Center, Mitter and his co-founders wrote that “we have put together a research program blending the analytical and theoretical insights of control and signal processing with the problematique of artificial intelligence.”

In 1981, Mitter became Director and then co-director of the MIT Laboratory for Information and Decision Systems (LIDS), which today is housed by the MIT Schwarzman College of Computing. During this same time, he was also director of the Center for Intelligent Control Systems (CICS). Mitter oversaw many foundational advances in the field now broadly referred to as information and decision science, including innovations in communication networks, numerical algorithms for control design and large-scale optimization, nonlinear estimation, statistical signal and image processing, and the newly emerging field of robust control. As co-director of LIDS alongside Robert Gallagher from 1986-1999, he saw the lab pioneer neuro-dynamic programming and contribute to significant advances in coding and information theory, large-scale statistical inference, and estimation. 

The animating principle which made CICS such a hotbed of innovation was almost certainly the interdisciplinary approach baked into the Center’s structure. “I believe that there are three key aspects of [Mitter’s] leadership of the Center,” wrote Peter Doerschuk, then an Associate Professor of ECE at Purdue (now at Cornell), in a letter supporting the work of CICS. “(1) The breadth of vision of what intellectual areas belong in the Center, (2) the willingness to take risk, and (3) the willingness to pursue curiosity-driven activities.”

In a similar letter, Pietro Perona, Professor of Electrical Engineering and of Computation and Neural Systems at the California Institute of Technology and Director of the NSF Engineering Research Center in Neuromorphic Systems Engineering compared his time at LIDS to a visit to the Renaissance: “The intellectual and social atmosphere that I met in Professor Mitter’s group at LIDS was unique and precious, wholly new to me. I liken it to the atmosphere of the renaissance court as one could have met in a small city-state such as Urbino and Mantova in the 1500s. At its center, the enlightened prince, Sanjoy Mitter, would provide funding, inspiration, emotional support and an aesthetic canon. The courtiers–the graduate students and post-docs, all of them absolutely first-rate–would be given the intellectual freedom, the encouragement, and the resources to follow their own curiosity, and encouraged to work on a diverse array of problems (speech, vision, probabilistic representations of geometrical objects, variational problems, handwriting) with the common requirement that that any investigation should be intellectually and analytically rigorous, and aim at the fundamental issues underlying every problem in computational perception…”

Mitter’s interdisciplinary approach to intellectual pursuits was almost certainly informed by his near-encyclopedic memory for papers, books, and other scholarly materials he’d encountered months or even years before, as his advisee Tom Baran, now Co-Founder and CEO of Fathom Optics, recalls: “Many wonderful conversations with Sanjoy started with him asking something like ‘So, how is progress on your PhD work going?’, and ended with a pad of paper in hand, absolutely filled with references to seek out, the exact authors and exact years of publication all recalled from his memory, going back to the 1970s, 1960s and earlier.” 

Mitter’s uncanny recall was tempered with great personal warmth and a gift for conversation, all of which made him a popular mentor. Paul-Peter Sotiriadis PhD ‘02, Electronics Lab Director and Professor of ECE at the National Technical University of Athens, Greece, paid tribute to Mitter thusly: “I will forever hold Professor Mitter in the highest regard, both as a remarkable scientist and an exceptional human being. His groundbreaking work in mathematical systems and control theory has left an enduring impact. Throughout my time at MIT, his unwavering support, guidance, and inspiring teaching were invaluable to me. I felt honored to have him on my Ph.D. committee.” And Frank Moss SM ‘72, PhD ‘77, co-founder of Bluefin Labs and former director of the MIT Media Lab, paid tribute to Mitter’s kind nature: “I remember him as the quintessential mannered gentleman, very warm, friendly and collegial, a great conversationalist with a deep intellect and sharp wit. A rarity at MIT at that time, and even now. I ran into him on campus several times when I was director of the Media Lab between 2006 and 2012, nearly 40 years after that. He would greet me with a warm hello as if no time at all had passed.” 

A pencil sketch of Sanjoy Mitter drawn by his student Dr. Aman Chawla (SM 2006), now a research associate within the Bharti School at IIT Delhi, who said of Mitter, “His constant inspiration and the example he set, of rigor and discipline, were a constant reminder of what one can accomplish simply using the intellect…. To me his most fascinating research contribution was the information theoretic view of the Kalman-Bucy filter. This triggered my SM thesis and influenced my view of–basically–everything.”

Michael Warren BS ’69, MS ’71, PhD ’74 recalled Mitter’s near-infinite capacity for curiosity: “He was inspirational and always a fine gentleman, as much at ease with discussing economics or politics as he was with discoursing on system theory.” Gordon Kaufman, Morris A. Adelman Professor of Management Emeritus and Professor of Statistics Emeritus within the MIT Sloan School of Management, agreed, referring to Mitter as “among the last surviving members of a generation of MIT intellectual princes. He was old-school gracious, knew how to listen and was beloved by his students. We were closest of friends for fifty years, sharing fine wine and gastronomic adventures in France and occasional mathematical moments here in the US. Sanjoy was a strategic thinker, one who was not afraid to address big intellectual problems.”

That intellectual fearlessness yielded a body of work which garnered Mitter international acclaim. Among his many awards and honors, Mitter was a Fellow of the IEEE and IFAC; was elected to the National Academy of Engineering (1988); won the IEEE Control Systems Award (2000); elected a Foreign Member of Istituto Veneto di Scienze, Lettere ed Arti (2003); awarded the AACC Richard E. Bellman Control Heritage Award (2007); received the IEEE Eric E. Sumner Award (2015); and was named a Foreign Fellow of the Indian National Academy of Engineering (2015). 

Following the death of his first wife Adriana, Mitter reconnected with, and married, childhood friend Rekha Ghosh, who survives him. Other members of Mitter’s extended and accomplished family include younger brother Pronob Mitter; niece Anjali Mitter Duva; and nephew Siddhartha Mitter. Additionally, Mitter is survived by a large network of friends, students, and mentees who considered him akin to family; as news spread of his passing, condolences and personal remembrances arrived from dozens of academic institutions around the globe. “He was Uncle Sanjoy to our children and grandchildren (save the latest, born in May),” recalled Peter Falb, Professor Emeritus of Applied Mathematics, Brown University and a LIDS Research Affiliate, of his 50+ year friendship with Mitter. “He spent many Thanksgivings, Christmases, birthdays and visits with our family. While he was, of course, a distinguished colleague, he was primarily a close and dear friend.” Professor Kaufman agreed, alluding to Mitter’s impact with devastating simplicity: “He made his mark. I will miss him.”

New method simplifies the construction process for complex materials

Posted on August 2, 2023 by - News

Engineers are constantly searching for materials with novel, desirable property combinations. For example, an ultra-strong, lightweight material could be used to make airplanes and cars more fuel-efficient, or a material that is porous and biomechanically friendly could be useful for bone implants.

Cellular metamaterials — artificial structures composed of units, or cells, that repeat in various patterns — can help achieve these goals. But it is difficult to know which cellular structure will lead to the desired properties. Even if one focuses on structures made of smaller building blocks like interconnected beams or thin plates, there are an infinite number of possible arrangements to consider. So, engineers can manually explore only a small fraction of all the cellular metamaterials that are hypothetically possible.

Researchers from MIT and the Institute of Science and Technology Austria have developed a computational technique that makes it easier for a user to quickly design a metamaterial cell from any of those smaller building blocks, and then evaluate the resulting metamaterial’s properties.

Their approach, like a specialized CAD (computer-aided design) system for metamaterials, allows an engineer to quickly model even very complex metamaterials and experiment with designs that may have otherwise taken days to develop. The user-friendly interface also enables the user to explore the entire space of potential metamaterial shapes, since all building blocks are at their disposal.

“We came up with a representation that can cover all of the different shapes engineers have traditionally shown interest in. Because you can build them all the same way, that means you can switch between them more fluidly,” says MIT electrical engineering and computer science graduate student Liane Makatura, co-lead author of a paper on this technique.

Makatura wrote the paper with co-lead author Bohan Wang, an MIT postdoc; Yi-Lu Chen, a graduate student at the Institute of Science and Technology Austria (ISTA); Bolei Deng, an MIT postdoc; Chris Wojtan and Bernd Bickel, professors at ISTA; and senior author Wojciech Matusik, a professor of electrical engineering and computer science at MIT who leads the Computational Design and Fabrication Group within the MIT Computer Science and Artificial Intelligence Laboratory. The research will be presented at SIGGRAPH.

A unified method

When a scientist develops a cellular metamaterial, she typically begins by choosing a representation that will be used to describe her potential designs. This choice determines the set of shapes that will be available for exploration.

For instance, she may choose a technique that represents metamaterials using many interconnecting beams. However, this prevents her from exploring metamaterials based on other elements, such as thin plates or 3D structures like spheres. Those shapes are given by different representations, but so far, there hasn’t been a unified way to describe all shapes in one method.

“By choosing a specific subspace ahead of time, you limit your exploration and introduce a bias based on your intuition. While this can be useful, intuition can be incorrect, and some of the other shapes may have also been worth exploring for your particular application,” says Makatura.

She and her collaborators took a step back and closely examined different metamaterials. They saw that the shapes that comprise the overall structure could be easily represented by lower-dimensional shapes — a beam could be reduced to a line or a thin-shell could be compressed to a flat surface.

They also noticed that cellular metamaterials often have symmetries, so only a small part of the structure needs to be represented. The rest can be built by rotating and mirroring that initial piece. “By combining those two observations, we arrived at this idea that cellular metamaterials could be well-represented as a graph structure,” she says. With their graph-based representation, a user builds a metamaterial skeleton using building blocks that are created by vertices and edges. For instance, to create a beam structure, one places a vertex at each end point of the beam and connects them with a line. Then the user employs a function over that line to specify the thickness of the beam, which can be varied so one part of the beam is thicker than another. The process for surfaces is similar — the user marks the most important features with vertices and then chooses a solver that infers the rest of the surface. These easy-to-use solvers even allow users to quickly construct a highly complex type of metamaterial, called a triply periodic minimal surface (TPMS). These structures are incredibly powerful, but the usual process to develop them is arduous and prone to failure. “With our representation, you can also start combining these shapes. Perhaps a unit cell containing both a TPMS structure and a beam structure could give you interesting properties. But so far, those combinations really haven’t been explored to any degree,” she says. At the end of the process, the system outputs the entire graph-based procedure, showing every operation the user took to reach the final structure — all the vertices, edges, solvers, transformations, and thickening operations. Within the user interface, designers can preview the current structure at any point in the building procedure and directly predict certain properties, such as its stiffness. Then, the user can iteratively tweak some parameters and evaluate it again until a suitable design is reached.

A user-friendly framework

The researchers used their system to recreate structures that spanned many unique classes of metamaterials. Once they had designed the skeletons, each metamaterial structure took only seconds to generate.

This GIF shows a rendered cellular metamaterial that the researchers designed using their system. The rotating 4x4x4 tiling of the unit cell is composed of beams, shells, and simple volumetric shapes.

This GIF shows a rendered cellular metamaterial that the researchers designed using their system. This rendering, a 4x4x4 tiling of the unit cell, is composed of beams, shells, and simple volumetric shapes. It would have been much more difficult to create this using another approach because of the different types of architectural elements involved. Image courtesy of the researchers.

They also created automated exploration algorithms, giving each a set of rules and then turning it loose in their system. In one test, an algorithm returned more than 1,000 potential truss-based structures in about an hour.

In addition, the researchers conducted a user-study with 10 individuals who had little prior experience modeling metamaterials. The users were able to successfully model all six structures they were given, and most agreed that the procedural graph representation made the process easier.

“Our representation makes all sorts of structures more accessible to people. We were especially pleased with users’ ability to generate TPMS. These complex structures are usually difficult even for experts to generate. Still, one TPMS in our study had the lowest average modeling time out of all six structures, which was surprising and exciting,” she says.

In the future, the researchers want to enhance their technique by incorporating more complex skeleton thickening procedures, so the system can model a wider variety of shapes. They also want to continue exploring the use of automatic generation algorithms.

And in the long term, they’d like to use this system for inverse design, where one would specify desired material properties and then use an algorithm to find the optimal metamaterial structure.

This research is funded, in part, by a National Science Foundation Graduate Research Fellowship, the MIT Morningside Academy Design Fellowship, the Defense Advanced Research Projects Agency (DARPA), an ERC Consolidator Grant, and the NewSat project.

2023 EECS Awards

Posted on August 1, 2023 by - EECS Celebrates Awards

It’s the summer of 2023, and we want to celebrate the accomplishments and contributions of our incredible EECS community by sharing some of the awards given by the department this year. Congratulations to all the winners!

The following is a list of all 2023 award winners.

Seth J. Teller Award for Excellence, Diversity and Inclusion

  • Taylor Baum
  • Marija Ilic
  • Martin Nisser

Frederick C. Hennie III Award for Teaching Excellence

  • Ajay R. Brahmakshatriva
  • Benjamin B. Kettle
  • Joseph Ravichandran

Burgess (1952) & Elizabeth Jamieson Award for Excellence in Teaching

  • Adam Chlipala
  • Fredo Durand

Ruth and Joel Spira Award for Excellence in Teaching

  • YuFeng Kevin Chen
  • Daniel Sanchez

Jerome Salzer Award for Excellence in Teaching Recitation Sections

  • Henry Corrigan-Gibbs

Louis D. Smullin (‘39) Award for Teaching Excellence

  • Jelena Notaros

EECS Outstanding Educator Award

  • Jonathan Ragan-Kelley
  • Tess Smidt

Kolokotrones Education Award

  • Joseph D. Steinmeyer

Carlton E. Tucker Award for teaching excellence

  • Hope K. Dargan
  • Amir H. Karamlou

Harold L. Hazen Award for teaching excellence

  • Gabrielle E. Ecanow
  • Miloslawa Piszczek
  • Peggy Yang

EECS Digital Innovation Award

  • Shen Shen

Department Head Special Recognition Award

  • Peter Franklin Satterthwaite

Richard J Caloggero Award

  • Dennis M. Freeman

Behring Undergraduate Prize

  • Thomas Bergamaschi

Ernst A. Guillemin SM Thesis Award in EECS

  • Sayed S. Afzal for “Battery-Free Subsea Internet-of-Things”. Thesis supervised by Fadel Adib.
  • Nikola Samardzic for “Making Computation on Encrypted Data Practical through Hardware Acceleration of Fully Homomorphic Encryption”. Supervised by Daniel Sanchez.

Ernst A. Guillemin SM Thesis Award in Artificial Intelligence and Decision-Making

  • Brice Huang for “Computational Hardness in Random Optimization Problems from the Overlap Gap Property”. Supervised by Guy Bresler.
  • Olivia Seow for “An Intuitive Tool For 3D Design Creation”. Supervised by Stefanie Mueller.

William A. Martin SM Thesis Award in Computer Science

  • Simon H. Langowski for “Fast, Metadata-private Anonymous Broadcast”. Supervised by Srini Devadas.
  • Rachel Zhang for “The Optimal Error Resilience of Interactive Communication Over Binary Channels”. Supervised by Vinod Vaikuntanathan and Yael Kalai.

Jin Au Kong PhD Thesis in EE

  • Murat Onen for “Devices and Algorithms for Analog Deep Learning”. Supervised by Jesus Del Alamo.
  • Christopher L. Panuski for “Resonant Spatial Light Modulation: Optical Programming and Sensing at the Fundamental Limit”. Supervised by Dirk Englund.
  • Youngkyu Sung for “High-fidelity Two-qubit Gates and Noise Spectroscopy with Superconducting Qubits”. Supervised by Will Oliver.

George M. Sprowls PhD Thesis Award in AI+D

  • Jason M. Altschuler for “Transport and Beyond: Efficient Optimization over Probability Distributions”. Thesis supervised by Pablo Parrilo.
  • Feras A. Saad for “Scalable Structure Learning, Inference, and Analysis with Probabilistic Programs”. Thesis supervised by Vikash Mansinghka.
  • Jingzhao Zhang for “Optimization Theory and Machine Learning Practice: Mind the Gap”. Thesis supervised by Suvrit Sra.

George M. Sprowls PhD Thesis Award in Computer Science

  • Tej Chajed for “Verifying a concurrent, crash-safe file system with sequential reasoning”. Thesis supervised by Nickolai Zeldovich and Frans Kaashoek.
  • Lijie Chen for “Better Hardness via Algorithms, and New Forms of Hardness versus Randomness”. Thesis supervised by Ryan Williams.
  • Alex Lombardi for “Provable Instantiations of Correlation Intractability and the Fiat-Shamir Heuristic”. Thesis supervised by Vinod Vaikuntanathan.

Robert M. Fano UROP Award

  • Purvaja Balaji for “EntailClass: A Classification Approach to EntailSum and Evaluation”. Supervised by Amar Gupta.
  • Helena A. Merker for “End-to-End Document Extraction, Identification, and Evaluation”.

Jeremy Gerstle UROP

  • Aparna Ajit Gupte for “Cryptographic Hardness of Sparse Linear Regression and Gaussian Mixture Learning”. Supervised by Vinod Vaikuntanathan.
  • Raiphy Jerez for “Electronic and mechanical design for mass manufacturing of modular robots”. Supervised by Stefanie Mueller.

Morais (1986) and Rosenblum (1986) Fund UROP Award

  • Rui Yao for “Efficient Truncated Linear Regression with Unknown Noise Variance”. Supervised by Costis Daskalakis.

Anna Pogosyants UROP Prize

  • Ahmed Katary, for “Functionality-Aware Styling: Fabricating Functional Objects with Selective”. Supervised by Stefanie Mueller.

Licklider UROP Award

  • Divya V. Nori for “De novo PROTAC Design Using Graph-based Deep Generative Models”. Supervised by Connor W. Coley.

2023 SuperUROP Outstanding Research Award

  • William Li for “Polygenic Dissection of Phenotypic Heterogeneity in Alzheimer’s Disease”. Supervised by Manolis Kellis and Yosuke Tanigawa.
  • Katherine Mohr for “Transforming Datacenter Applications to Microsecond Latency with Continuous Profile-Guided Optimization”. Supervised by Saman Amarasinghe.

George C. Newton Undergraduate Lab Prize

  • Stephen S. Kandeh
  • Julio A. Rodriguez

Northern Telecom/BNR Project Award

  • Thanadol (Pleng) Chomphoochan
  • Dillon M. DuPont

David Adler Memorial EE MEng Thesis Award

  • William Wu for “Neural Data Shaping and Evaluation via Mutual Information Estimation”. Supervised by Muriel Médard.
  • Jeanne L. Harabedian for “Modeling the Arterial System to Improve Ultrasound Measurements of Hemodynamic Parameters”. Supervised by Charles Sodini.

Charles & Jennifer Johnson Computer Science MEng thesis award

  • Alexandra Dima for “GSTACO: A Generalized Sparse Tensor Algebra Compiler”. Supervised by Saman Amarasinghe.
  • Amir Farhat for “Increasing DoS-Resilience for Cross-Protocol Proxies”. Supervised by Karen Sollins.

Charles & Jennifer Johnson Artificial Intelligence and Decision-Making MEng Thesis Award

  • Laura Dodds for “A Portable Handheld Fine-Grained RFID Localization System with Complex-Controlled Polarization”. Supervised by Fadel Adib.
  • Veerapatr Yotamornsunthorn for “Decoding Invisible 3D Printed Tags with Convolutional Neural Networks”. Supervised by Stefanie Mueller.

J. Francis Reintjes Excellence in 6-A Industrial Practice Award

  • Cooper Jones
  • Jialan (Karen) Wang

Using AI to protect against AI image manipulation

Posted on July 31, 2023 by - News

As we enter a new era where technologies powered by artificial intelligence can craft and manipulate images with a precision that blurs the line between reality and fabrication, the specter of misuse looms large. Recently, advanced generative models such as DALL-E and Midjourney, celebrated for their impressive precision and user-friendly interfaces, have made the production of hyper-realistic images relatively effortless. With the barriers of entry lowered, even inexperienced users can generate and manipulate high-quality images from simple text descriptions — ranging from innocent image alterations to malicious changes. Techniques like watermarking pose a promising solution, but misuse requires a preemptive (as opposed to only post hoc) measure. 

In the quest to create such a new measure, researchers from MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) developed “PhotoGuard,” a technique that uses perturbations — minuscule alterations in pixel values invisible to the human eye but detectable by computer models — that effectively disrupt the model’s ability to manipulate the image.

PhotoGuard uses two different “attack” methods to generate these perturbations. The more straightforward “encoder” attack targets the image’s latent representation in the AI model, causing the model to perceive the image as a random entity. The more sophisticated “diffusion” one defines a target image and optimizes the perturbations to make the final image resemble the target as closely as possible.

“Consider the possibility of fraudulent propagation of fake catastrophic events, like an explosion at a significant landmark. This deception can manipulate market trends and public sentiment, but the risks are not limited to the public sphere. Personal images can be inappropriately altered and used for blackmail, resulting in significant financial implications when executed on a large scale,” says Hadi Salman, an MIT graduate student in electrical engineering and computer science (EECS), affiliate of MIT CSAIL, and lead author of a new paper about PhotoGuard. 

“In more extreme scenarios, these models could simulate voices and images for staging false crimes, inflicting psychological distress and financial loss. The swift nature of these actions compounds the problem. Even when the deception is eventually uncovered, the damage — whether reputational, emotional, or financial — has often already happened. This is a reality for victims at all levels, from individuals bullied at school to society-wide manipulation.”

PhotoGuard in practice

AI models view an image differently from how humans do. It sees an image as a complex set of mathematical data points that describe every pixel’s color and position — this is the image’s latent representation. The encoder attack introduces minor adjustments into this mathematical representation, causing the AI model to perceive the image as a random entity. As a result, any attempt to manipulate the image using the model becomes nearly impossible. The changes introduced are so minute that they are invisible to the human eye, thus preserving the image’s visual integrity while ensuring its protection.

The second and decidedly more intricate “diffusion” attack strategically targets the entire diffusion model end-to-end. This involves determining a desired target image, and then initiating an optimization process with the intention of closely aligning the generated image with this preselected target.

In implementing, the team created perturbations within the input space of the original image. These perturbations are then used during the inference stage, and applied to the images, offering a robust defense against unauthorized manipulation.

“The progress in AI that we are witnessing is truly breathtaking, but it enables beneficial and malicious uses of AI alike,” says MIT professor of EECS and CSAIL principal investigator Aleksander Madry, who is also an author on the paper. “It is thus urgent that we work towards identifying and mitigating the latter. I view PhotoGuard as our small contribution to that important effort.”

The diffusion attack is more computationally intensive than its simpler sibling, and requires significant GPU memory. The team says that approximating the diffusion process with fewer steps mitigates the issue, thus making the technique more practical.

To better illustrate the attack, consider an art project, for example. The original image is a drawing, and the target image is another drawing that’s completely different. The diffusion attack is like making tiny, invisible changes to the first drawing so that, to an AI model, it begins to resemble the second drawing. However, to the human eye, the original drawing remains unchanged.

By doing this, any AI model attempting to modify the original image will now inadvertently make changes as if dealing with the target image, thereby protecting the original image from intended manipulation. The result is a picture that remains visually unaltered for human observers, but protects against unauthorized edits by AI models.

As far as a real example with PhotoGuard, consider an image with multiple faces. You could mask any faces you don’t want to modify, and then prompt with “two men attending a wedding.” Upon submission, the system will adjust the image accordingly, creating a plausible depiction of two men participating in a wedding ceremony.

Now, consider safeguarding the image from being edited; adding perturbations to the image before upload can immunize it against modifications. In this case, the final output will lack realism compared to the original, non-immunized image.

Interactive Demo: Raising the cost of malicious AI-powered image editing

All hands on deck

Key allies in the fight against image manipulation are the creators of the image-editing models, says the team. For PhotoGuard to be effective, an integrated response from all stakeholders is necessary. “Policymakers should consider implementing regulations that mandate companies to protect user data from such manipulations. Developers of these AI models could design APIs that automatically add perturbations to users’ images, providing an added layer of protection against unauthorized edits,” says Salman.

Despite PhotoGuard’s promise, it’s not a panacea. Once an image is online, individuals with malicious intent could attempt to reverse engineer the protective measures by applying noise, cropping, or rotating the image. However, there is plenty of previous work from the adversarial examples literature that can be utilized here to implement robust perturbations that resist common image manipulations.

“A collaborative approach involving model developers, social media platforms, and policymakers presents a robust defense against unauthorized image manipulation. Working on this pressing issue is of paramount importance today,” says Salman. “And while I am glad to contribute towards this solution, much work is needed to make this protection practical. Companies that develop these models need to invest in engineering robust immunizations against the possible threats posed by these AI tools. As we tread into this new era of generative models, let’s strive for potential and protection in equal measures.”

“The prospect of using attacks on machine learning to protect us from abusive uses of this technology is very compelling,” says Florian Tramèr, an assistant professor at ETH Zürich. “The paper has a nice insight that the developers of generative AI models have strong incentives to provide such immunization protections to their users, which could even be a legal requirement in the future. However, designing image protections that effectively resist circumvention attempts is a challenging problem: Once the generative AI company commits to an immunization mechanism and people start applying it to their online images, we need to ensure that this protection will work against motivated adversaries who might even use better generative AI models developed in the near future. Designing such robust protections is a hard open problem, and this paper makes a compelling case that generative AI companies should be working on solving it.”

Salman wrote the paper alongside fellow lead authors Alaa Khaddaj and Guillaume Leclerc MS ’18, as well as Andrew Ilyas ’18, MEng ’18; all three are EECS graduate students and MIT CSAIL affiliates. The team’s work was partially done on the MIT Supercloud compute cluster, supported by U.S. National Science Foundation grants and Open Philanthropy, and based upon work supported by the U.S. Defense Advanced Research Projects Agency. It was presented at the International Conference on Machine Learning this July.

Armando Solar-Lezama named inaugural Distinguished Professor of Computing

Posted on July 31, 2023 by - News

The MIT Stephen A. Schwarzman College of Computing named Armando Solar-Lezama as the inaugural Distinguished Professor of Computing, effective July 1. 

Solar-Lezama is the first person appointed to this position generously endowed by Professor Jae S. Lim of the Department of Electrical Engineering and Computer Science (EECS). Established in the MIT Schwarzman College of Computing, the chair is being awarded to Solar-Lezama for being an outstanding faculty member who is recognized as a leader and innovator.

“I’m pleased to make this appointment and recognize Armando for his remarkable contributions to MIT and the scientific community,” says Daniel Huttenlocher, dean of the MIT Schwarzman College of Computing and the Henry Ellis Warren Professor of Electrical Engineering and Computer Science. “I’m greatly appreciative of Professor Lim for his thoughtful gesture in creating this new chair in the college, providing us with the opportunity to acknowledge the accomplishments of our faculty.”

Solar-Lezama, a professor of electrical engineering and computer science, leads the Computer-Aided Programming Group in the Computer Science and Artificial Intelligence Laboratory (CSAIL) that focuses on program synthesis, an area of research that lies at the intersection of programming systems and artificial intelligence. The group’s research ranges from designing new analysis techniques and automated reasoning mechanisms to developing new programming models that automate challenging aspects of programming.

A member of the EECS faculty since 2008, Solar-Lezama, who also serves as the associate director and chief operating officer for CSAIL, is most interested in software synthesis and its applications to particular program domains such as high-performance computing. He first found this niche area of program synthesis as a graduate student at the University of California at Berkeley, for which his thesis project, a language called Sketch, treats program synthesis as a search problem in which the algorithms pare down the search space to make the search faster and more efficient. Since then, program synthesis research has greatly expanded into the active field it is today.

QS Ranks MIT the world’s No. 1 university for 2023-24

Posted on June 30, 2023 by - EECS Celebrates Awards, News

MIT has again been named the world’s top university by the QS World University Rankings, which were announced today. This is the 12th year in a row MIT has received this distinction.

The full 2024 edition of the rankings — published by Quacquarelli Symonds, an organization specializing in education and study abroad — can be found at TopUniversities.com. The QS rankings are based on factors including academic reputation, employer reputation, citations per faculty, student-to-faculty ratio, proportion of international faculty, and proportion of international students.

MIT was also ranked the world’s top university in 11 of the subject areas ranked by QS, as announced in March of this year.

The Institute received a No. 1 ranking in the following QS subject areas: Chemical Engineering; Civil and Structural Engineering; Computer Science and Information Systems; Data Science and Artificial Intelligence; Electrical and Electronic Engineering; Linguistics; Materials Science; Mechanical, Aeronautical, and Manufacturing Engineering; Mathematics; Physics and Astronomy; and Statistics and Operational Research.

MIT also placed second in five subject areas: Accounting and Finance; Architecture/Built Environment; Biological Sciences; Chemistry; and Economics and Econometrics.

Professor Emeritus Dick Thornton, adventurer, dies at 93

Posted on June 27, 2023 by - News

Professor Emeritus Richard “Dick” Thornton SM ’54, ScD ’57 passed away on May 16th, 2023. He was 93. 

An innovator, entrepreneur, adventurer, and outdoorsman, Thornton’s influence upon all the communities he touched was profound. During his 40+ year tenure within MIT’s Department of Electrical Engineering and Computer Science, he mentored and inspired countless individuals, consistently advocating for bold and fearless exploration in all areas of life. 

Born on September 24, 1929 in Scarsdale, NY, Thornton earned his undergraduate degree from Princeton University in Electrical Engineering, and his Masters and PhD from MIT before taking a position within what was then called the Department of Electrical Engineering, where he would stay for the remainder of his academic career. Thornton was, in many ways, the prototypical MIT professor: an inventive tinkerer who once burned down his own garage while repairing his home’s solar water heating system, he “was much more at home in the laboratory than in the classroom,” according to his then-student John Kassakian, now Professor of Electrical Engineering, Emeritus. “During the ‘62-‘63 school year he created a lab course where students designed and built an oscilloscope… He also invented what we students called the ‘Dickie Board’, a plastic board with an array of hollow pins that allowed one to build a circuit using toothpicks to fasten and connect the component leads in the pins.” 

Thornton’s gift for improvisation belied a stringent sense of academic and intellectual rigor. “My single most important and memorable moment with Dick was in the context of an EE lab that he taught and I took as a sophomore,” remembered Chuck Counselman, now MIT Professor of Planetary Science, Emeritus. “He wrote in my laboratory notebook, “Stop being a ham.” He was referring to my hobby, since childhood, of amateur ham radio.  As a “ham” I had designed and built electronic circuits by qualitative thinking and trial-and-error, a.k.a. “cut and try.”  I did not analyze circuits theoretically and quantitatively, by means of linear algebra, differential equations, and Fourier transforms. [In other words], I had not yet learned to think analytically. Dick’s four-word comment changed the course of my life as an engineer and a scientist.” 

Counselman’s experience of having his life’s course affected by Thornton is one shared by hundreds of MIT students–over the years, Thornton supervised over 100 bachelor’s level theses, 60 master’s and engineer’s level theses, and was a supervisor or reader for over two dozen doctoral theses. “He was one of relatively few faculty to teach our entire common undergraduate core at the time,” says Steve Leeb, now Emanuel E. Landsman (1958) Professor of EE and CS. “He was beyond generous with his time, insight, and resources, and a critical mentor to graduate students and junior faculty far outside of his own group.” One of Thornton’s SM thesis advisees, Jeff Lang, now Vitesse Professor of EE, remembers that “Dick was one of several faculty (Melcher, Haus and Staelin too) that got me interested in electromechanics, and E&M in general. He was a great advisor in the sense that he involved me not only in the technical part of my thesis, but the “business” side too, i.e. dealing with industrial sponsors before, during and after our project.” 

Thornton’s business sense was the product of extensive experience; during his time at MIT, he founded two separate companies. The first was Thornton Associates, Inc., which focused on improving methods of measuring and controlling water purity. Leeb explains: “Instruments that Dick Thornton designed and patented are now used by virtually all major semiconductor manufacturers to monitor the purity of water and to provide automatic control of fabrication operations. He designed and patented techniques and circuit topologies which have introduced fundamental, widely used techniques in power electronic drives, for, among other applications, the increasingly important variable reluctance machine.” 

The second company that Thornton founded, alongside some of his graduate students, was MagneMotion, later acquired by Rockwell Automation, which focused on the commercial application of linear synchronous motors in a variety of settings, including manufacturing lines, luggage handling, elevators (such as weapons elevators in US military ships), biological sample handling (including for COVID-testing and blood-testing equipment), and amusement park rides (notably at Disney).  “His Maglev work spanned a long duration up to his founding of MagneMotion, including seminal work in the early 1970s with Henry Kolm,” says Ford Foundation Professor of Engineering David Perreault, noting an article, “Electromagnetic Flight”, which Thornton and Kolm published in the October 1, 1973 issue of Scientific American with the tempting subheader: ‘The future of high-speed ground transportation may well lie not with wheeled trains but with vehicles that “fly” a foot or so above a guideway, lifted and propelled by electromagnetic forces’. Not content merely to spark scientific readers’ imaginations, Thornton and Kolm went on to develop a scale version of the “magplane” they proposed, and to author multiple papers on linear motor propulsion. Thornton’s work on maglev made him an in-demand lecturer, with frequent invitations from the United States Department of Transportation, the National Maglev Initiative, Draper Laboratories, and Bechtel; as his career at MIT progressed, he pioneered critical innovations and techniques for making maglev systems practical, implementable, and safe–including crucial work on fault tolerant power electronic drives; low cost guideways for maglev; and the development of safe, economical techniques to shield passengers and bystanders from the strong magnetic fields involved in these transportation systems. 

Additionally, Thornton was a significant contributor to the SEEC (Semiconductor Electronics Education Committee) book series, acting as lead author on three of the influential texts, collectively the first such series to address semiconductor devices and circuits. He also co-taught the first course in solid state electronics at MIT with co-authors Campbell Searle, Paul Gray, Richard Adler, and Arthur Smith. Kassakian, who worked with the co-authors during the summer of 1963, remembers that seminal course: “The lectures were lively since all the faculty not lecturing were in the room criticizing, commenting, and correcting, as all the material had never been taught before.” 

A great advocate for career flexibility, Thornton stressed the importance of staying curious and taking career risks. “Over my career, I observed that virtually all MIT faculty at some point had to make significant changes in their career,” he recalled in a 2022 interview with MIT’s Alumni Association. “It can be difficult for a faculty member to pivot because, ironically, if you’re an expert in the field, it’s easier to gain funding to be a better expert in that field. It’s much harder to transition.” To ease that transition for other MIT faculty members, Thornton established the Thornton Family Faculty Research Innovation Fund (FRIF), a grant designed to encourage mid-career faculty members to explore new and exciting directions in their research, with an emphasis on novel and untested ideas. 

In congruence with his risk-taking career persona, Thornton was personally an adventurous outdoorsman, whose hobbies included whitewater canoeing, electric car racing, skiing, hiking, and sailing (a sport he was first introduced to at the MIT Sailing Pavilion). He met his wife, Marian, when she signed on as an extra crew member for a 2-week cruise along the coast of Maine. The two enjoyed not only cruising but racing, according to Kassakian: “I sailed with him several times, including the race around Nantucket where we came in last because the spinnaker wrapped itself around the forestry. Poor Marian spent 20 minutes at the top of the mast, swinging around in a bosun’s chair, trying to untangle it.”

The Thornton family poses atop the car entered by MIT in a cross-country race of electric vehicles. From left to right: Richard “Dick” Thornton SM ’54, ScD ’57; son Richard, son Douglas, daughter Margo, and wife Marian Thornton. Photo courtesy of the Thornton family.

On other occasions, Kassakian remembers the couple being more characteristically ahead of the crowd. “They were always in the lead [in our annual bike ride from MIT to Provincetown.] During one trip, after a 65 mile leg to spend the night at my place on the Cape, we were missing some key ingredient for dinner. Dick said he had it at his home in Woods Hole, which was about 15 miles from my house. So he hopped back on his bike, rode to Woods Hole, got whatever it was, and returned so we could make dinner.” James Kirtley, Professor of Electrical Engineering (Post-Tenure), corroborated the couple’s remarkable speed: “Dick and Marian also met the LEES group on the Tour de Cape on bicycles, but were both much faster than the rest of us.” The couple remained married for 60 years, by all accounts decades filled with an enormous number of wilderness adventures, including canoeing through spring floods and regular hiking pilgrimages to the White Mountains, plus many peregrinations between Maine and New York by sail. 

Thornton was predeceased by Marian, as well as by a sister, Julie Wagoner. He is survived by daughter Margo Webber and husband Todd; son Dick and wife Toni; son Doug and wife Alison; seven grandchildren, including Alex Webber and wife Nicole, Kip Webber and wife Katie, Karen Thornton, Margo Thornton, Christopher Thornton, Arianna Thornton and Nicholas Thornton; great-grandson Riley Webber; sisters, Mary Carr and Nina Asgeirsson; dearest friend and companion, Kathleen Lang; and many nieces, nephews, extended family, and friends. 

Of Thornton’s extraordinary and personal impact on his community, both at MIT and beyond, Steve Leeb summarized, “He was consummately knowledgeable as an engineer, yet always humble and compassionate… a man of diamond clarity and sterling honesty, he defined the phrase ‘a gentleman and a scholar.’  He will be greatly missed.”

The author gratefully acknowledges the extensive contributions to this obituary of Steve Leeb, John Kassakian, James Kirtley, David Perreault, George Verghese, Jeffrey Lang, Chuck Counselman, and other friends and colleagues of Richard Thornton.

MIT PhD student enhances STEM education in underrepresented communities in Puerto Rico

Posted on June 26, 2023 by - News

Taylor Baum knows that access is everything. So the fourth-year MIT PhD candidate in the Department of Electrical Engineering and Computer Science has been working in recent years to enhance STEM education in underrepresented communities in Puerto Rico.

As the founder of social impact venture Sprouting, Baum has been leading programs to facilitate community between K-12, undergraduate, post-graduate students, and teachers of science, technology, engineering and math (STEM) disciplines, including two recent workshops, one in 2022 and the latest in May 2023.

Making an impact

In the summer of 2022, Baum organized Sprouting’s first large five-day hackathon for teachers and students in Ponce, Puerto Rico, with the goal of empowering educators to teach coding and computer science to their students. The hackathon began with virtual training for teachers, followed by students and teachers working on an applied project together, and ended with an in-person presentation of their work.

Forty teachers and 80 students from middle and high schools signed up for the hackathon, emphasizing the ever-present urge in Puerto Rico to explore computer science and STEM in general. Additionally, 10 volunteer teaching assistants from around the world assisted in the hackathon, with speakers attending from different nonprofits and organizations in the area. The event was supported by a grant from the MIT Center for Brains, Minds and Machines (CBMM), with Neptuno supporting internet services and a speaker, and Amazon Web Services (AWS) providing food, prizes, and another speaker.

“Our goal was to provide teachers with the skills and materials they need to teach coding and computer science in their own classrooms,” Baum says. “We believe that by providing teachers with the tools they need, we could ensure that all students have access to quality STEM education, regardless of their background.”

Baum decided to host the hackathon in Ponce, the island’s second-largest city on its southern coast, so that it could serve a Puerto Rican audience that doesn’t get as much attention with STEM events, which are often hosted in the northern capital city of San Juan. Sprouting organizers recruited participants through Facebook groups and communities of local public school teachers and students on the island.

With previous experience teaching in Puerto Rico, Baum understood that residents have faced a number of day-to-day challenges from government corruption to a decade-long financial crisis to natural disasters. Despite this, she saw the obvious talent and potential in every student and teacher she interacted with, and wanted to put on a high-energy, impactful event that would benefit the community without straining or burdening them. The hackathon was an opportunity for her to give back to the community and make a real impact on the future of education.

“We were thrilled to be a part of this initiative,” said Baum. “We believed that it would make a real difference in the lives of teachers and students in Puerto Rico, and the participants took full advantage of the event, showing clear potential and passion for progressing technological education on the island. We were grateful for the support of CBMM, Neptuno, and AWS, and we look forward to seeing continued results of this event in the future.”

This hackathon is an example of the continued desire that Baum has for enhancing coding education for underrepresented communities. She has also taught machine learning in Uruguay and has been working on her Spanish language skills to make it more comfortable for the folks she is working with by holding Sprouting events in their native language. She now gives lectures entirely in Spanish to reduce the engagement barrier. Like the students working to increase fluency in their coding languages, Baum shows that getting out of your comfort zone can lead to incredibly rewarding experiences. 

Taylor Baum (center) instructs participants during a 2023 workshop on control theory. Photo credit: Kris Brewer

Growing a community

With the success of the first workshop in 2022, Baum organized a second event that was held last month “Sprouting a STEM Community 2023” (SSC23), in Mayagüez, Puerto Rico.

SSC23 continues the journey toward a sustainable effort to support computer science education in underrepresented communities. The fully Spanish lessons that were provided to the teachers during the event are meant to be used in their classrooms after the event, thus the teachers were able to take back the knowledge and skills gained from the hackathon and continue teaching coding and computer science to their students.

This year’s event was sponsored by a grant from the CBMM and hosted by the Department of Biology at the University of Puerto Rico at Mayagüez. Neptuno provided internet services, ensuring each workshop ran smoothly. Each morning started with two to three lectures from faculty, postdocs, and graduate students with presentations pertaining to their field. Many faculty presentations included sections on how rewarding a career as a scientist can be and how it can make a significant impact on not only oneself, but also one’s family, one’s community, and the world at large.

After a social break, participants split into volunteer teaching assistant-led groups for interactive workshops consisting of a range of topics from control theory to the cardiovascular system. The format of the workshops emphasized active learning methods commonly employed at MIT, ranging from group discussions to guided tutorials to group exercises. One popular exercise teaching control theory consisted of participants taking turns guiding a “blind human robot” (a fellow participant with their eyes closed) through a maze of desks to retrieve a water bottle using only (verbally spoken) program commands such as “forward 2 meters, turn right 90 degrees.”

“SSC23 was entirely organized by volunteers,” says Sprouting Program Manager Paloma Sanchez-Jauregui. “What was most inspiring was seeing how volunteers and teaching assistants were willing to altruistically put in hours of work to make this event happen, even when they had final exams and full time jobs. During the workshops, we saw the participants guide others in solving the workshop contents that Taylor built. Seeing my own community become mentors with their peers made me really happy and proud of being part of a caring and passionate community”

A new addition to this year’s event included the creation of the Sprouting Ambassadors program. As the program website describes, “Sprouting Ambassadors will commit to unlock the existing potential present throughout the communities they grew up in, in turn growing as a leader and activist in education. Sprouting will fund a trip to MIT for the ambassadors to learn more about research in STEM and, more importantly, help them organize a Sprouting event so they can become mentors in their own community.” Thirteen applicants from this year’s Sprouting event were selected and will be coming to MIT’s campus this summer.

This year’s event was a success drawing about 150 participants, volunteers, and speakers from around the island continuing to emphasize the talent and passion for technological innovation in Puerto Rico. With the event being fully in-person this year, all lectures were recorded, live streamed, and posted to the Sprouting website video page, with captions in Spanish, for those who could not attend in person and allow for continuing use of the materials.

For her continuing work on Sprouting activities, along with her academic pursuits, Baum was recently honored as an MIT Woman of Excellence sponsored by the Office of Graduate Education, and awarded the 2023 Seth J. Teller Award for Excellence, Inclusion and Diversity from the Department of Electrical Engineering and Computer Science at MIT.