Teaching a robot new skills used to require coding expertise. But a new generation of robots could potentially learn from just about anyone.

Engineers are designing robotic helpers that can “learn from demonstration.” This more natural training strategy enables a person to lead a robot through a task, typically in one of three ways: via remote control, such as operating a joystick to remotely maneuver a robot; by physically moving the robot through the motions; or by performing the task themselves while the robot watches and mimics.

Learning-by-doing robots usually train in just one of these three demonstration approaches. But MIT engineers have now developed a three-in-one training interface that allows a robot to learn a task through any of the three training methods. The interface is in the form of a handheld, sensor-equipped tool that can attach to many common collaborative robotic arms. A person can use the attachment to teach a robot to carry out a task by remotely controlling the robot, physically manipulating it, or demonstrating the task themselves — whichever style they prefer or best suits the task at hand.

The MIT team tested the new tool, which they call a “versatile demonstration interface,” on a standard collaborative robotic arm. Volunteers with manufacturing expertise used the interface to perform two manual tasks that are commonly carried out on factory floors.

The researchers say the new interface offers increased training flexibility that could expand the type of users and “teachers” who interact with robots. It may also enable robots to learn a wider set of skills. For instance, a person could remotely train a robot to handle toxic substances, while further down the production line another person could physically move the robot through the motions of boxing up a product, and at the end of the line, someone else could use the attachment to draw a company logo as the robot watches and learns to do the same.

“We are trying to create highly intelligent and skilled teammates that can effectively work with humans to get complex work done,” says Mike Hagenow, a postdoc at MIT in the Department of Aeronautics and Astronautics. “We believe flexible demonstration tools can help far beyond the manufacturing floor, in other domains where we hope to see increased robot adoption, such as home or caregiving settings.”

Hagenow will present a paper detailing the new interface, at the IEEE Intelligent Robots and Systems (IROS) conference in October. The paper’s MIT co-authors are Dimosthenis Kontogiorgos, a postdoc at the MIT Computer Science and Artificial Intelligence Lab (CSAIL); Yanwei Wang PhD ’25, who recently earned a doctorate in electrical engineering and computer science; and Julie Shah, MIT professor and head of the Department of Aeronautics and Astronautics.

Training together

Shah’s group at MIT designs robots that can work alongside humans in the workplace, in hospitals, and at home. A main focus of her research is developing systems that enable people to teach robots new tasks or skills “on the job,” as it were. Such systems would, for instance, help a factory floor worker quickly and naturally adjust a robot’s maneuvers to improve its task in the moment, rather than pausing to reprogram the robot’s software from scratch — a skill that a worker may not necessarily have.

The team’s new work builds on an emerging strategy in robot learning called “learning from demonstration,” or LfD, in which robots are designed to be trained in more natural, intuitive ways. In looking through the LfD literature, Hagenow and Shah found LfD training methods developed so far fall generally into the three main categories of teleoperation, kinesthetic training, and natural teaching.

One training method may work better than the other two for a particular person or task. Shah and Hagenow wondered whether they could design a tool that combines all three methods to enable a robot to learn more tasks from more people.

“If we could bring together these three different ways someone might want to interact with a robot, it may bring benefits for different tasks and different people,” Hagenow says.

Tasks at hand

With that goal in mind, the team engineered a new versatile demonstration interface (VDI). The interface is a handheld attachment that can fit onto the arm of a typical collaborative robotic arm. The attachment is equipped with a camera and markers that track the tool’s position and movements over time, along with force sensors to measure the amount of pressure applied during a given task.

When the interface is attached to a robot, the entire robot can be controlled remotely, and the interface’s camera records the robot’s movements, which the robot can use as training data to learn the task on its own. Similarly, a person can physically move the robot through a task, with the interface attached. The VDI can also be detached and physically held by a person to perform the desired task. The camera records the VDI’s motions, which the robot can also use to mimic the task when the VBI is reattached.

To test the attachment’s usability, the team brought the interface, along with a collaborative robotic arm, to a local innovation center where manufacturing experts learn about and test technology that can improve factory-floor processes. The researchers set up an experiment where they asked volunteers at the center to use the robot and all three of the interface’s training methods to complete two common manufacturing tasks: press-fitting and molding. In press-fitting, the user trained the robot to press and fit pegs into holes, similar to many fastening tasks. For molding, a volunteer trained the robot to push and roll a rubbery, dough-like substance evenly around the surface of a center rod, similar to some thermomolding tasks.

For each of the two tasks, the volunteers were asked to use each of the three training methods, first teleoperating the robot using a joystick, then kinesthetically manipulating the robot, and finally, detaching the robot’s attachment and using it to “naturally” perform the task as the robot recorded the attachment’s force and movements.

The researchers found the volunteers generally preferred the natural method over teleoperation and kinesthetic training. The users, who were all experts in manufacturing, did offer scenarios in which each method might have advantages over the others. Teleoperation, for instance, may be preferable in training a robot to handle hazardous or toxic substances. Kinesthetic training could help workers adjust the positioning of a robot that is tasked with moving heavy packages. And natural teaching could be beneficial in demonstrating tasks that involve delicate and precise maneuvers.

“We imagine using our demonstration interface in flexible manufacturing environments where one robot might assist across a range of tasks that benefit from specific types of demonstrations,” says Hagenow, who plans to refine the attachment’s design based on user feedback and will use the new design to test robot learning. “We view this study as demonstrating how greater flexibility in collaborative robots can be achieved through interfaces that expand the ways that end-users interact with robots during teaching.”

This work was supported, in part, by the MIT Postdoctoral Fellowship Program for Engineering Excellence and the Wallenberg Foundation Postdoctoral Research Fellowship.

Large language models (LLMs) excel at using textual reasoning to understand the context of a document and provide a logical answer about its contents. But these same LLMs often struggle to correctly answer even the simplest math problems.

Textual reasoning is usually a less-than-ideal way to deliberate over computational or algorithmic tasks. While some LLMs can generate code like Python to handle symbolic queries, the models don’t always know when to use code, or what kind of code would work best.

LLMs, it seems, may need a coach to steer them toward the best technique.

Enter CodeSteer, a smart assistant developed by MIT researchers that guides an LLM to switch between code and text generation until it correctly answers a query.

CodeSteer, itself a smaller LLM, automatically generates a series of prompts to iteratively steer a larger LLM. It reviews the model’s current and previous answers after each round and provides guidance for how it can fix or refine that solution until it deems the answer is correct.

The researchers found that augmenting a larger LLM with CodeSteer boosted its accuracy on symbolic tasks, like multiplying numbers, playing Sudoku, and stacking blocks, by more than 30 percent. It also enabled less sophisticated models to outperform more advanced models with enhanced reasoning skills.

This advance could improve the problem-solving capabilities of LLMs for complex tasks that are especially difficult to solve with textual reasoning alone, such as generating paths for robots in uncertain environments or scheduling shipments in an international supply chain.

“There is a race to develop better and better models that are capable of doing everything, but we’ve taken a complementary approach. Researchers have spent years developing effective technologies and tools to tackle problems in many domains. We want to enable LLMs to select the right tools and methods, and make use of others’ expertise to enhance their own capabilities,” says Chuchu Fan, an associate professor of aeronautics and astronautics (AeroAstro) and principal investigator in the MIT Laboratory for Information and Decision Systems (LIDS).

Fan, the senior author of the study, is joined on a paper about the work by LIDS graduate student Yongchao Chen; AeroAstro graduate student Yilun Hao; University of Illinois at Urbana-Champaign graduate student Yueying Liu; and MIT-IBM Watson AI Lab Research Scientist Yang Zhang. The research will be presented at the International Conference on Machine Learning.

An LLM “trainer”  

Ask an LLM which number is bigger, 9.11 or 9.9, and it will often give the wrong answer by using textual reasoning. But ask it to use code to answer the same question, and it can generate and execute a Python script to compare the two numbers, easily solving the problem.

Initially trained to understand and predict human language, LLMs are more likely to answer queries using text, even when code would be more effective. And while they have learned to generate code through fine-tuning, these models often generate an incorrect or less efficient version of the code.

Rather than trying to retrain a powerful LLM like GPT-4 or Claude to improve these capabilities, the MIT researchers fine-tune a smaller, lightweight LLM to guide a larger model between text and code. Fine-tuning a smaller model doesn’t change the larger LLM, so there is no risk it would undermine the larger model’s other abilities.

“We were also inspired by humans. In sports, a trainer may not be better than the star athlete on the team, but the trainer can still give helpful suggestions to guide the athlete. This steering method works for LLMs, too,” Chen says.

This trainer, CodeSteer, works in conjunction with the larger LLM. It first reviews a query and determines whether text or code is suitable for this problem, and which sort of code would be best.

Then it generates a prompt for the larger LLM, telling it to use a coding method or textual reasoning to answer the query. The larger model follows this prompt to answer the query and sends the result back to CodeSteer, which reviews it.

If the answer is not correct, CodeSteer will continue prompting the LLM to try different things that might fix the problem, such as incorporating a search algorithm or constraint into its Python code, until the answer is correct.

“We found that oftentimes, the larger LLM will try to be lazy and use a shorter, less efficient code that will not carry the correct symbolic calculation. We’ve designed CodeSteer to avoid this phenomenon,” Chen says.

A symbolic checker evaluates the code’s complexity and sends a signal to CodeSteer if it is too simple or inefficient. The researchers also incorporate a self-answer checker into CodeSteer, which prompts the LLM to generate code that calculates the answer to verify it is correct.

Tackling complex tasks

As the researchers designed CodeSteer, they couldn’t find suitable symbolic datasets to fine-tune and test the model, since many existing benchmarks don’t point out whether a certain query could be best solved with text or code.

So, they gathered a corpus of 37 complex symbolic tasks, including spatial reasoning, mathematics, order reasoning, and optimization, and built their own dataset, called SymBench. They implemented a fine-tuning approach that leverages SymBench to maximize the performance of CodeSteer.

In their experiments, CodeSteer outperformed all nine baseline methods they evaluated and boosted average accuracy from 53.3 percent to 86.4 percent. It maintains similar performance even on unseen tasks, and on a variety of LLMs.

In addition, a general-purpose model augmented with CodeSteer can achieve higher accuracy than state-of-the-art models designed to focus on complex reasoning and planning, while requiring much less computation.

“Our method uses an LLM’s own capabilities. By augmenting an LLM with the ability to smartly use coding, we can take a model that is already very strong and improve its performance even more,” Chen says.

In the future, the researchers want to streamline CodeSteer to speed up its iterative prompting process. In addition, they are studying how to effectively fine-tune a unified model with the ability to switch between textual reasoning and code generation, rather than relying on a separate assistant.

“The authors present an elegant solution to the critical challenge of tool utilization in LLMs. This simple yet impactful method enables state-of-the-art LLMs to achieve significant performance improvements without requiring direct fine-tuning,” says Jinsung Yoon, a staff research scientist at Google Cloud AI, who was not involved with this work. “This research represents a substantial contribution that promises to significantly enhance the application of LLMs to a diverse range of tasks with which they currently struggle.”

“Their success in training a smaller, specialized model to strategically guide larger, advanced models is particularly impactful,” adds Chi Wang, a senior staff scientist at Google DeepMind who was not involved with this work. “This intelligent collaboration among diverse AI ‘agents’ paves the way for more robust and versatile applications in complex real-world scenarios.”

This research is supported, in part, by the U.S. Office of Naval Research and the MIT-IBM Watson AI Lab.

Imagine a future where artificial intelligence quietly shoulders the drudgery of software development: refactoring tangled code, migrating legacy systems, and hunting down race conditions, so that human engineers can devote themselves to architecture, design, and the genuinely novel problems still beyond a machine’s reach. Recent advances appear to have nudged that future tantalizingly close, but a new paper by researchers at MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) and several collaborating institutions argues that this potential future reality demands a hard look at present-day challenges. 

Titled “Challenges and Paths Towards AI for Software Engineering,” the work maps the many software-engineering tasks beyond code generation, identifies current bottlenecks, and highlights research directions to overcome them, aiming to let humans focus on high-level design while routine work is automated. 

“Everyone is talking about how we don’t need programmers anymore, and there’s all this automation now available,” says Armando Solar‑Lezama, MIT professor of electrical engineering and computer science, CSAIL principal investigator, and senior author of the study. “On the one hand, the field has made tremendous progress. We have tools that are way more powerful than any we’ve seen before. But there’s also a long way to go toward really getting the full promise of automation that we would expect.”

Solar-Lezama argues that popular narratives often shrink software engineering to “the undergrad programming part: someone hands you a spec for a little function and you implement it, or solving LeetCode-style programming interviews.” Real practice is far broader. It includes everyday refactors that polish design, plus sweeping migrations that move millions of lines from COBOL to Java and reshape entire businesses. It requires nonstop testing and analysis — fuzzing, property-based testing, and other methods — to catch concurrency bugs, or patch zero-day flaws. And it involves the maintenance grind: documenting decade-old code, summarizing change histories for new teammates, and reviewing pull requests for style, performance, and security.

Industry-scale code optimization — think re-tuning GPU kernels or the relentless, multi-layered refinements behind Chrome’s V8 engine — remains stubbornly hard to evaluate. Today’s headline metrics were designed for short, self-contained problems, and while multiple-choice tests still dominate natural-language research, they were never the norm in AI-for-code. The field’s de facto yardstick, SWE-Bench, simply asks a model to patch a GitHub issue: useful, but still akin to the “undergrad programming exercise” paradigm. It touches only a few hundred lines of code, risks data leakage from public repositories, and ignores other real-world contexts — AI-assisted refactors, human–AI pair programming, or performance-critical rewrites that span millions of lines. Until benchmarks expand to capture those higher-stakes scenarios, measuring progress — and thus accelerating it — will remain an open challenge.

If measurement is one obstacle, human‑machine communication is another. First author Alex  Gu, an MIT graduate student in electrical engineering and computer science, sees today’s interaction as “a thin line of communication.” When he asks a system to generate code, he often receives a large, unstructured file and even a set of unit tests, yet those tests tend to be superficial. This gap extends to the AI’s ability to effectively use the wider suite of software engineering tools, from debuggers to static analyzers, that humans rely on for precise control and deeper understanding. “I don’t really have much control over what the model writes,” he says. “Without a channel for the AI to expose its own confidence — ‘this part’s correct … this part, maybe double‑check’ — developers risk blindly trusting hallucinated logic that compiles, but collapses in production. Another critical aspect is having the AI know when to defer to the user for clarification.” 

Scale compounds these difficulties. Current AI models struggle profoundly with large code bases, often spanning millions of lines. Foundation models learn from public GitHub, but “every company’s code base is kind of different and unique,” Gu says, making proprietary coding conventions and specification requirements fundamentally out of distribution. The result is code that looks plausible yet calls non‑existent functions, violates internal style rules, or fails continuous‑integration pipelines. This often leads to AI-generated code that “hallucinates,” meaning it creates content that looks plausible but doesn’t align with the specific internal conventions, helper functions, or architectural patterns of a given company. 

Models will also often retrieve incorrectly, because it retrieves code with a similar name (syntax) rather than functionality and logic, which is what a model might need to know how to write the function. “Standard retrieval techniques are very easily fooled by pieces of code that are doing the same thing but look different,” says Solar‑Lezama. 

The authors mention that since there is no silver bullet to these issues, they’re calling instead for community‑scale efforts: richer, having data that captures the process of developers writing code (for example, which code developers keep versus throw away, how code gets refactored over time, etc.), shared evaluation suites that measure progress on refactor quality, bug‑fix longevity, and migration correctness; and transparent tooling that lets models expose uncertainty and invite human steering rather than passive acceptance. Gu frames the agenda as a “call to action” for larger open‑source collaborations that no single lab could muster alone. Solar‑Lezama imagines incremental advances—“research results taking bites out of each one of these challenges separately”—that feed back into commercial tools and gradually move AI from autocomplete sidekick toward genuine engineering partner.

“Why does any of this matter? Software already underpins finance, transportation, health care, and the minutiae of daily life, and the human effort required to build and maintain it safely is becoming a bottleneck. An AI that can shoulder the grunt work — and do so without introducing hidden failures — would free developers to focus on creativity, strategy, and ethics” says Gu. “But that future depends on acknowledging that code completion is the easy part; the hard part is everything else. Our goal isn’t to replace programmers. It’s to amplify them. When AI can tackle the tedious and the terrifying, human engineers can finally spend their time on what only humans can do.”

“With so many new works emerging in AI for coding, and the community often chasing the latest trends, it can be hard to step back and reflect on which problems are most important to tackle,” says Baptiste Rozière, an AI scientist at Mistral AI, who wasn’t involved in the paper. “I enjoyed reading this paper because it offers a clear overview of the key tasks and challenges in AI for software engineering. It also outlines promising directions for future research in the field.”

Gu and Solar-Lezama wrote the paper with University of California at Berkeley Professor Koushik Sen and PhD students Naman Jain and Manish Shetty, Cornell University Assistant Professor Kevin Ellis and PhD student Wen-Ding Li, Stanford University Assistant Professor Diyi Yang and PhD student Yijia Shao, and incoming Johns Hopkins University assistant professor Ziyang Li. Their work was supported, in part, by the National Science Foundation (NSF), SKY Lab industrial sponsors and affiliates, Intel Corp. through an NSF grant, and the Office of Naval Research.

The researchers are presenting their work at the International Conference on Machine Learning (ICML). 

MIT equips students with the tools to advance science and engineering — but a new class aims to ensure they also develop their own values and learn how to navigate conflicting viewpoints.

Offered as a pilot this past spring, the multidisciplinary class 21.01 (Compass Course: Love, Death, and Taxes: How to Think — and Talk to Others — About Being Human), invites students to wrestle with difficult questions like:

• What do we value (and why)?
• What do we know (and how do we know it)?
• What do we owe to each other (and what should we do about it)?

The class is part of the Compass Initiative, which is led by faculty from across the MIT School of Humanities, Arts, and Social Sciences (SHASS). 

Lily L. Tsai, Ford Professor of Political Science and lead faculty for Compass, says the new course is meant to help students use the humanities and social sciences as their guide to thinking about the kind of humans they want to be and what kind of society they want to help create.

"At MIT, we're some of the people who are creating the technologies that are accelerating change and leading to more unpredictability in the world. We have a special responsibility to envision and reimagine a moral and civic education that enables people to navigate it," says Tsai.

The course is the result of a multi-year collaboration involving over 30 faculty from 19 departments, ranging from Philosophy and Literature to Brain and Cognitive Sciences and Electrical Engineering and Computer Science, all led by a core team of 14 faculty from SHASS and a student advisory board.

During its initial run in the spring, Compass followed an arc that began with students investigating questions of value. Early in the semester, students explored what makes a genius, using Beethoven's "Symphony No. 9" as a case study, accompanied by lectures from Emily Richmond Pollock, associate professor of music, and a podcast conversation with Larry Guth, professor of mathematics, and David Kaiser, professor of physics and science, technology, and society. 

Students then grappled with the concept of a merit-based society by digging into the example of the imperial Chinese civil service exam, guided by professor of history Tristan Brown. Next, they questioned what humans really know to be true by examining the universality of language through lectures by professor of linguistics Adam Albright, and the philosophy of truth and knowledge through lectures by professor of philosophy Alex Byrne.

The semester ended with challenging debates about what humans owe one another, including a class designed by Nobel laureate and professor of economics Esther Duflo on taxation and climate burdens. 

More than anything, Tsai says, she hopes that Compass prepares students to navigate dorm hallways, the family Thanksgiving table, or future labs or boardroom tables, and learn how to express opinions and actively listen to others with whom they may disagree — all without canceling one another. 

The class takes a "flipped classroom" approach: Students watch recorded lectures at home and come to class prepared for discussion and debate. Each section is co-taught by two faculty members, combining disciplines and perspectives.

Second-year mechanical engineering major Kayode Dada signed up because it fulfilled a communications-intensive requirement and offered cross-departmental exposure. But Compass ultimately became more than that to him. "College isn't just about learning science stuff — it's also about how we grow as people," he says. Dada was assigned to a section co-taught by Tsai and professor of literature Arthur Bahr. 

Forming a social contract

In the first week, students draft a Rousseau-inspired social compact and learn firsthand how to build a classroom community. "We knew these were deep topics," Dada says. "To get the most out of the class, we had to open up, respect each other, and keep conversations confidential."

One early exercise was especially impactful. After watching lectures by Ford Professor of Philosophy and Women’s and Gender Studies Sally Haslanger on value, students were asked to draw a map representing their values, with arrows pointing from ones that were more instrumental to ones that were fundamental.

At first, Dada felt stuck. Growing up in Kentucky, the son of a Nigerian immigrant who had dreamed of attending MIT himself, Dada had focused for years on gaining admission to the Institute. "I thought getting into MIT would make me feel fulfilled," he admits. "But once I got here, I realized the work alone wasn't enough."

The values exercise helped him reorient. He identified practicing Christianity, hard work, helping others, and contributing to society as central to his belief system. The exercise influenced Dada, leading him to choose to volunteer at a robotics camp for kids in Louisville to share his MIT education with others.

Who governs science? 

Later in the semester, Dada was animatedly representing a figure whose views contradicted his own: James D. Watson, the Nobel Prize winner who co-discovered DNA's structure — and is also a controversial figure. 

That week, each student had been assigned a persona from a 1976 Cambridge City Council hearing debating recombinant DNA research. The class, designed by Associate Professor Robin Scheffler, was investigating the question: Who governs science — scientists, the government, those who fund research, or the public?

They revisited a real-life debate around recombinant DNA research and the dangers for biological weapons development and other threats to the public that citizens of that time believed it posed when carried out in MIT and Harvard University labs. Pioneered in the 1970s, the technique involved the splicing of genes related to the E. coli bacterium. In the Compass classroom, students argued different sides from their personas: banning the research, moving labs outside city limits, or proceeding without government interference.

Dada notes how faculty intentionally seeded conflicting viewpoints. "It taught me how to negotiate with someone who has different values and come to a resolution that respects everyone involved," he says. "That's something I want to keep exploring."

When Dada closed his presentation with frantically-Googled sentimental music piped unexpectedly from his phone, his classmates laughed in appreciation. The atmosphere was more intimate than academic — an ethos Tsai hoped to cultivate. "They really built intellectual relationships based on trust," she says. "There was a lot of laughter. They took joy in disagreeing and debating."

Changing opinions 

First-year student-athlete Shannon Cordle, who is majoring in mechanical engineering, didn't know what to expect from Compass. Since it was new, there were no student reviews. What stood out to her was the grading system: 15 percent of the final grade is based on a rubric each student created for themselves.

Cordle's goal was to become more comfortable expressing an opinion — even before she's fully formed it. "It's easy to stay quiet when you're unsure," she says. "Compass helped me practice speaking up and being willing to be wrong, because that's how you learn."

One week, the class debated whether a meritocracy creates a just society — an especially relevant topic at MIT, given its famously selective admissions process. 

Students were able to pick their stance beforehand, and then invited to change it as they gained more perspectives during the debate.

"This helps students grasp not only the flaws in another viewpoint, but also how to strengthen their arguments," Tsai says.

Cordle, who hopes to go into prosthetics, views her future field as representing the perfect balance between creativity and ethics. "The humanities challenge how we view our fields as scientists and engineers," she says.

A compass helps travelers find their way — but it's most useful when they need to reorient and change direction. In that spirit, Compass prepares students not just to ask big questions, but to keep asking — and keep adapting — as their lives and careers evolve.

“Bringing these unexpected class elements together with students and faculty generated magical alchemy — a kind of transformation that we didn't even know we could create,” Tsai says.

In addition to the class, the MIT Compass Podcast engages in these fundamental questions with guests from across the MIT schools of Science and Engineering. There are also plans to adapt the residential version of this class for online learners on MITx.

In addition to philanthropic support from MIT Corporation life member emeritus Ray Stata '57, the initiative is supported by the Office of the Vice Chancellor and the MIT Human Insight Collaborative's SHASS Education Innovation Fund, which promotes new, transformative educational approaches in SHASS fields.

MIT researchers have developed a new theoretical framework for studying the mechanisms of treatment interactions. Their approach allows scientists to efficiently estimate how combinations of treatments will affect a group of units, such as cells, enabling a researcher to perform fewer costly experiments while gathering more accurate data.

As an example, to study how interconnected genes affect cancer cell growth, a biologist might need to use a combination of treatments to target multiple genes at once. But because there could be billions of potential combinations for each round of the experiment, choosing a subset of combinations to test might bias the data their experiment generates. 

In contrast, the new framework considers the scenario where the user can efficiently design an unbiased experiment by assigning all treatments in parallel, and can control the outcome by adjusting the rate of each treatment.

The MIT researchers theoretically proved a near-optimal strategy in this framework and performed a series of simulations to test it in a multiround experiment. Their method minimized the error rate in each instance.

This technique could someday help scientists better understand disease mechanisms and develop new medicines to treat cancer or genetic disorders.

“We’ve introduced a concept people can think more about as they study the optimal way to select combinatorial treatments at each round of an experiment. Our hope is this can someday be used to solve biologically relevant questions,” says graduate student Jiaqi Zhang, an Eric and Wendy Schmidt Center Fellow and co-lead author of a paper on this experimental design framework.

She is joined on the paper by co-lead author Divya Shyamal, an MIT undergraduate; and senior author Caroline Uhler, the Andrew and Erna Viterbi Professor of Engineering in EECS and the MIT Institute for Data, Systems, and Society (IDSS), who is also director of the Eric and Wendy Schmidt Center and a researcher at MIT’s Laboratory for Information and Decision Systems (LIDS). The research was recently presented at the International Conference on Machine Learning.

Simultaneous treatments

Treatments can interact with each other in complex ways. For instance, a scientist trying to determine whether a certain gene contributes to a particular disease symptom may have to target several genes simultaneously to study the effects.

To do this, scientists use what are known as combinatorial perturbations, where they apply multiple treatments at once to the same group of cells.

“Combinatorial perturbations will give you a high-level network of how different genes interact, which provides an understanding of how a cell functions,” Zhang explains.

Since genetic experiments are costly and time-consuming, the scientist aims to select the best subset of treatment combinations to test, which is a steep challenge due to the huge number of possibilities.

Picking a suboptimal subset can generate biased results by focusing only on combinations the user selected in advance.

The MIT researchers approached this problem differently by looking at a probabilistic framework. Instead of focusing on a selected subset, each unit randomly takes up combinations of treatments based on user-specified dosage levels for each treatment.

The user sets dosage levels based on the goal of their experiment — perhaps this scientist wants to study the effects of four different drugs on cell growth. The probabilistic approach generates less biased data because it does not restrict the experiment to a predetermined subset of treatments.

The dosage levels are like probabilities, and each cell receives a random combination of treatments. If the user sets a high dosage, it is more likely most of the cells will take up that treatment. A smaller subset of cells will take up that treatment if the dosage is low.

“From there, the question is how do we design the dosages so that we can estimate the outcomes as accurately as possible? This is where our theory comes in,” Shyamal adds.

Their theoretical framework shows the best way to design these dosages so one can learn the most about the characteristic or trait they are studying.

After each round of the experiment, the user collects the results and feeds those back into the experimental framework. It will output the ideal dosage strategy for the next round, and so on, actively adapting the strategy over multiple rounds.

Optimizing dosages, minimizing error

The researchers proved their theoretical approach generates optimal dosages, even when the dosage levels are affected by a limited supply of treatments or when noise in the experimental outcomes varies at each round.

In simulations, this new approach had the lowest error rate when comparing estimated and actual outcomes of multiround experiments, outperforming two baseline methods.

In the future, the researchers want to enhance their experimental framework to consider interference between units and the fact that certain treatments can lead to selection bias. They would also like to apply this technique in a real experimental setting.

“This is a new approach to a very interesting problem that is hard to solve. Now, with this new framework in hand, we can think more about the best way to design experiments for many different applications,” Zhang says.

This research is funded, in part, by the Advanced Undergraduate Research Opportunities Program at MIT, Apple, the National Institutes of Health, the Office of Naval Research, the Department of Energy, the Eric and Wendy Schmidt Center at the Broad Institute, and a Simons Investigator Award.

Scientists have long known that the brain’s visual system isn’t fully hardwired from the start — it becomes refined by what babies see — but the authors of a new MIT study still weren’t prepared for the degree of rewiring they observed when they took a first-ever look at the process in mice as it happened in real-time.

As the researchers in The Picower Institute for Learning and Memory tracked hundreds of “spine” structures housing individual network connections, or “synapses,” on the dendrite branches of neurons in the visual cortex over 10 days, they saw that only 40 percent of the ones that started the process survived. Refining binocular vision (integrating input from both eyes) required numerous additions and removals of spines along the dendrites to establish an eventual set of connections.

Former graduate student Katya Tsimring led the study, published this month in Nature Communications, which the team says is the first in which scientists tracked the same connections all the way through the “critical period,” when binocular vision becomes refined.

“What Katya was able to do is to image the same dendrites on the same neurons repeatedly over 10 days in the same live mouse through a critical period of development, to ask, what happens to the synapses or spines on them?,” says senior author Mriganka Sur, the Paul and Lilah Newton Professor in the Picower Institute and MIT’s Department of Brain and Cognitive Sciences. “We were surprised by how much change there is.”

Extensive turnover

In the experiments, young mice watched as black-and-white gratings with lines of specific orientations and directions of movement drifted across their field of view. At the same time, the scientists observed both the structure and activity of the neurons’ main body (or “soma”) and of the spines along their dendrites. By tracking the structure of 793 dendritic spines on 14 neurons at roughly Day 1, Day 5 and Day 10 of the critical period, they could quantify the addition and loss of the spines, and therefore the synaptic connections they housed. And by tracking their activity at the same time, they could quantify the visual information the neurons received at each synaptic connection. For example, a spine might respond to one specific orientation or direction of grating, several orientations, or might not respond at all. Finally, by relating a spine’s structural changes across the critical period to its activity, they sought to uncover the process by which synaptic turnover refined binocular vision.

Structurally, the researchers saw that 32 percent of the spines evident on Day 1 were gone by Day 5, and that 24 percent of the spines apparent on Day 5 had been added since Day 1. The period between Day 5 and Day 10 showed similar turnover: 27 percent were eliminated, but 24 percent were added. Overall, only 40 percent of the spines seen on Day 1 were still there on Day 10.

Meanwhile, only four of the 13 neurons they were tracking that responded to visual stimuli still responded on Day 10. The scientists don’t know for sure why the other nine stopped responding, at least to the stimuli they once responded to, but it’s likely they now served a different function.

What are the rules?

Having beheld this extensive wiring and rewiring, the scientists then asked what entitled some spines to survive over the 10-day critical period.

Previous studies have shown that the first inputs to reach binocular visual cortex neurons are from the “contralateral” eye on the opposite side of the head (so in the left hemisphere, the right eye’s inputs get there first), Sur says. These inputs drive a neuron’s soma to respond to specific visual properties such as the orientation of a line — for instance, a 45-degree diagonal. By the time the critical period starts, inputs from the “ipsilateral” eye on the same side of the head begin joining the race to visual cortex neurons, enabling some to become binocular.

It’s no accident that many visual cortex neurons are tuned to lines of different directions in the field of view, Sur says.

“The world is made up of oriented line segments,” Sur notes. “They may be long line segments; they may be short line segments. But the world is not just amorphous globs with hazy boundaries. Objects in the world — trees, the ground, horizons, blades of grass, tables, chairs — are bounded by little line segments.”

Because the researchers were tracking activity at the spines, they could see how often they were active and what orientation triggered that activity. As the data accumulated, they saw that spines were more likely to endure if (a) they were more active, and (b) they responded to the same orientation as the one the soma preferred. Notably, spines that responded to both eyes were more active than spines that responded to just one, meaning binocular spines were more likely to survive than non-binocular ones.

“This observation provides compelling evidence for the ‘use it or lose it’ hypothesis,” says Tsimring. “The more active a spine was, the more likely it was to be retained during development.”

The researchers also noticed another trend. Across the 10 days, clusters emerged along the dendrites in which neighboring spines were increasingly likely to be active at the same time. Other studies have shown that by clustering together, spines are able to combine their activity to be greater than they would be in isolation.

By these rules, over the course of the critical period, neurons apparently refined their role in binocular vision by selectively retaining inputs that reinforced their budding orientation preferences, both via their volume of activity (a synaptic property called “Hebbian plasticity”) and their correlation with their neighbors (a property called “heterosynaptic plasticity”). To confirm that these rules were enough to produce the outcomes they were seeing under the microscope, they built a computer model of a neuron, and indeed the model recapitulated the same trends as what they saw in the mice.

“Both mechanisms are necessary during the critical period to drive the turnover of spines that are misaligned to the soma and to neighboring spine pairs,” the researchers wrote, “which ultimately leads to refinement of [binocular] responses such as orientation matching between the two eyes.”

In addition to Tsimring and Sur, the paper’s other authors are Kyle Jenks, Claudia Cusseddu, Greggory Heller, Jacque Pak Kan Ip, and Julijana Gjorgjieva. Funding sources for the research came from the National Institutes of Health, The Picower Institute for Learning and Memory, and the Freedom Together Foundation.

Daniel Kleppner, the Lester Wolfe Professor Emeritus of Physics at MIT whose work in experimental atomic physics made an immense mark on the field, died on June 16 at the age of 92, in Palo Alto, California.

Kleppner’s varied research examined the interactions of atoms with static electric and magnetic fields and radiation. His work included creating precision measurements with hydrogen masers, including the co-invention of the hydrogen maser atomic clock; his research into the physics of Rydberg atoms and cavity quantum electrodynamics; and his pioneering work in Bose-Einstein condensation (BEC).

Kleppner, who retired in 2003 after 37 years at MIT, was a highly literate and articulate scientist whose exacting research and communication skills helped set the direction of modern atomic, molecular, and optical (AMO) physics. From 1987 to 2000, he was associate director of the MIT Research Laboratory of Electronics (RLE), and served as interim director in 2001. He also co-founded the MIT-Harvard Center for Ultracold Atoms (CUA) in 2000, where he was co-director until 2006.

While he was never awarded a Nobel Prize, Kleppner's impact on the field of atomic physics and quantum optics, and his generous mentorship, enabled the Nobel achievements of many others. His patient and exacting pursuit of discovery led to basic research insights that led to major achievements. His extensive research into the tiny atom provided the fundamental knowledge necessary for the huge: the eventual development of groundbreaking technologies such as the global positioning system (GPS), magnetic resonance imaging (MRI), and quantum computing.

“He was a leader in the department, and a leader in the American Physical Society,” says Wolfgang Ketterle, the John D. MacArthur Professor of Physics at MIT and a 2001 Nobel laureate. “He was a statesman of science. He was this eloquent person, this master of words who could express things in memorable ways, and at the same time he has this sense of humility.”

“Dan Kleppner was a giant in the area of AMO physics, and in science more broadly,” says John Doyle PhD ’91, Harvard Quantum Initiative co-director and Kleppner advisee who helped Kleppner create the Bose-Einstein condensate from atomic hydrogen. “Perhaps his most impactful legacy is leading a culture of respect and supportive community actions that all scientists in the area of AMO physics enjoy today. Not only did his science lay the path for current research directions, his kindness, erudition, and commitment to community — and community service — are now ever-expanding waves that guide AMO physics. He was a mentor and friend to me."

Kleppner’s daughter Sofie Kleppner notes: “People who worked on early lasers never imagined we would be scanning groceries at the checkout counter. When they developed the hydrogen maser, they were a bunch of nerdy people who really wanted to understand Einstein’s theory of relativity. This was the basis for GPS, this is how our flights run on time. Our dad was convinced that basic research today could lead to all sorts of valuable things down the road.”

Early life and career

Born in Manhattan on Dec. 16, 1932, Kleppner was the son of Vienna native and advertising agency founder Otto Kleppner, who wrote the best-selling book “Advertising Procedure.” His mother, Beatrice (Taub) Kleppner, grew up in New Jersey and was a graduate of Barnard College. She helped with Otto’s manuscripts. Daniel Kleppner was the second of three siblings; his brother, the late Adam Kleppner, was a professor of mathematics at the University of Maryland, and his sister, Susan Folkman, was a research psychologist at the University of California at Berkeley.

“As a teenager, I just liked building things,” Kleppner once said. “And that turned out to be very useful when I went on to become an experimental physicist. I had a crystal radio, so I could listen to the radio over earphones. And the thought that the signals were just coming out of the atmosphere, I remember thinking: totally remarkable. And actually, I still do. In fact, the idea of the electromagnetic field, although it’s very well understood in physics, always seems like a miracle to me.”

In high school, he was inspired by his physics teacher, Arthur Hussey, who allowed Kleppner to work all hours in the labs. “There was one time when the whole school was having a pep rally, and I wasn’t that interested in cheering football, so I stayed up and worked in the lab, and the high school principal noticed that I was in there and called me in and gave me a dressing down for lack of school spirit.”

He didn’t care. Hussey talked with Kleppner about quantum mechanics, and “that sort of put a bee in my bonnet on that,” and taught him a little calculus. “In those years, physics was extremely fashionable. These were the post-war years, and physicists were considered heroes for having brought the war to conclusion with the atom bomb, and … the development of radar.”

He knew by then that he was “destined to spend a life in physics,” he said in a video interview for InfiniteMIT. “It was an easy era to become delighted by physics, and I was.”

Studying physics at Williams College, he was drawn to Albert Einstein’s theory of general relativity. He built a programmable machine that he called a forerunner of cybernetics. Williams also instilled in him a lifelong love of literature, and he almost became an English major. However, he didn’t appreciate what he called the school fraternities’ “playboy” and “anti-intellectual” atmosphere, and worked to graduate quickly within three years, in 1953.

He deferred his acceptance to Harvard University with a Fulbright Fellowship to Cambridge University, where he met the young physicist Kenneth Smith, whose research was with atomic beam resonance. Smith introduced him to the book “Nuclear Moments,” by Harvard professor Norman Ramsey, and presented a proposal by Ramsey’s advisor I.I. Rabi, who invented a technique that could make an atomic clock so precise “that you could see the effect of gravity on time that Einstein predicted,” said Kleppner.

“I found that utterly astonishing,” Kleppner noted. “The thought that gravity affects time: I had a hard time just visualizing that.”

When Kleppner wandered Harvard’s halls in 1955, he was excited to see a door with Ramsey’s name on it. He was interested in Ramsey’s research on molecular beam magnetic resonance, atomic clocks, and precision measurements. “Fortunately, I came along at a time when he had an opening in his research group,” Kleppner recalled.

A new atomic clock

As Kleppner’s advisor, Ramsey encouraged him to create a new type of atomic clock, believing that cesium and ammonia masers, a technology of amplified microwaves, were not precise enough to measure the effect of gravity on time.

Kleppner’s thesis was on using the concepts behind an ammonia maser to advance toward a hydrogen maser, which uses the natural microwave frequency of hydrogen atoms and amplifies it through stimulated emission of radiation. Kleppner discovered that coherent cesium atoms can bounce from properly prepared surfaces without losing their coherence.

After his 1959 PhD, Kleppner stayed on at Harvard, becoming an assistant professor in 1962.

Kleppner’s research on hydrogen led to a method to keep hydrogen atoms locked in a glass container for study over a longer period of time. The result, featuring hydrogen atoms bouncing within a microwave cavity, is used to stabilize the frequency of a clock to a precision better than one microsecond in a year.

In 1960, he and Ramsey successfully created a new atomic clock whose significant stability could confirm the minute effects of gravity on time, as predicted by Einstein’s theory of general relativity.

The current generation of optical clocks “are good enough to see the gravitational red shift for a few centimeters in height, so that’s quite extraordinary, and it’s had an extraordinary result,” said Kleppner. “We got to rethink just what we mean by time.”

While the hydrogen maser did verify Einstein’s conjecture about time and gravity, it took more than a decade before being widely used, at first by radio astronomers. Today, atomic clocks such as the hydrogen maser are used in applications requiring high short-term stability, such as the synchronization of ground-based timing systems that track global positioning satellites, for timekeeping and communication by naval observatories to maintain a precise and stable time reference known as UTC (USNO); very long-baseline microwave interferometry (VLBI) that enables astronomers to achieve very high resolution and study distant radio sources, including black holes; and, indirectly, in magnetic resonance imaging.

“When we first set out to make these atomic clocks, our goals were about the least practical you can think of,” Kleppner said in an interview with the MIT Physics Department. “From being a rather abstract idea that you’d like to somehow witness, it becomes a very urgent thing for the conduct of human affairs.”

Ramsey went on to win the Nobel Prize in Physics in 1989 for his work on the separated oscillatory fields method and its application in the hydrogen maser and atomic clocks.

MIT, ultracold gases, and BEC advancements

Kleppner figured he wouldn’t get tenure at Harvard, “because no matter how generous and good-spirited Norman was, he casts a long shadow, and it was good for me to be at just the right distance. When I came to MIT, I had a pallet of experiments that I wanted to pursue, and some ideas about teaching that I wanted to pursue, and the transition was very simple.”

Kleppner joined the Institute in 1966, and his Harvard PhD student (and current MIT professor post-tenure) David Pritchard followed him, to work on scattering experiments: Kleppner worked with pulsed lasers, and Pritchard with continuous-wave (CW) lasers.

“He was young, he was verbal, and he seemed to have new ideas about what to do,” says Pritchard. “We foresaw how important lasers would become. For a long time, it was just Dan and myself. That was actually the era in which lasers took over. Dan and I started off, we both got into lasers, and he did Rydberg atoms, and I did collisions and spectroscopy of weakly bound molecules and two-photon spectroscopy.”

Kleppner led the tiny MIT Atomic Physics Group to eventually become the US News and World Report’s No. 1 nationally ranked atomic physics group in 2012. “Dan was the leader on this,” recalled Pritchard. “To start from non-tenure and build it into the number-one ranked department in your subfield, that’s a lifetime achievement.”

The group became what Pritchard called “the supergroup” of laser developers that included Charles Townes, who won the Nobel for his work; Ali Javan, who established a major laser research center at MIT; and Dolly Shibles. Pritchard joined the faculty in 1970, and Ketterle joined in 1990 as his postdoc. “We were pioneers, and the result was of course that our total group had a bigger impact.”

“He’s not just the father figure of the field, he is my scientific father,” says Pritchard. “When I’m writing something and it’s not going very well, I would sort of think to myself, ‘What would Dan say? What would he advise you?”

With MIT low-temperature physicist Tom Greytak ’63, PhD ’67, Kleppner developed two revolutionary techniques — magnetic trapping and evaporative cooling. When the scientific community combined these techniques with laser cooling, atomic physics went into a major new direction.

In 1995, a group of researchers, led by Kleppner's former students Eric Cornell PhD ’90 and Carl Weiman ’73, made a BEC using rubidium atoms, and Ketterle succeeded with sodium atoms. For this achievement, they received the 2001 Nobel Prize in Physics. Kleppner called BEC “the most exciting advance in atomic physics for decades.” 

At a conference on BEC in 1996, Ketterle recalls Kleppner describing his own contributions: “'I feel like Moses, who showed his people the Holy Land, but he never reached it himself.' This was exactly what Dan did. He showed us the Holy Land of Bose-Einstein condensation. He showed us what is possible … He was the godfather of Bose-Einstein condensation.”

But he did reach the Holy Land. In 1998, when only a few groups had been able to create BECs, Kleppner and Greytak realized a hydrogen BEC. When he presented their work at the summer school in Varenna soon afterward, he received a long-lasting standing ovation — after 20 years of hard work, he had reached his goal.

“It is an irony that when Dan started this work, hydrogen was the only choice to reach the low temperatures for BEC,” says Ketterle. But in the end, it turned out that hydrogen has special properties that made it much harder to reach BEC than with other atoms. 

Rydberg atoms

In 1976, Kleppner pioneered the field of Rydberg atoms, a highly excited atom that shares the simple properties that characterize hydrogen. Kleppner showed that these states could be excited by a tunable laser and easily detected with field ionization. He then mapped out their response in high electric and magnetic fields, which he used to provide new physical insights into the connections between quantum mechanics and classical chaos.

In 1989, his research into atomic energy levels, under conditions where the corresponding classical motion is chaotic, mapped out the positions of thousands of quantum levels as a function of laser frequency and applied field using high-resolution laser spectroscopy. His observations gave new physical insight into the implications of classical chaos on quantum systems.

“I see Dan as being the inventor of Rydberg atoms,” says Dan’s former student William Phillips PhD ’76, physicist at the Institute of Standards and Technology (NIST). “Of course, Rydberg atoms is something that nature gives you, but Dan was the one who really understood this was something that you could use to do really new and wonderful things.”

Such atoms have proved to be useful for studying the transition between quantum mechanics and classical chaos. Kleppner’s 1976 paper on Rydberg atoms’ strong interactions, long lifetimes, and sensitivity to external fields has led to current scientific research and multimillion-dollar startups interested in developing the promising Rydberg quantum computer; highly accurate measurements of electric and magnetic fields; and in quantum optics experiments.

“Largely due to Dan’s seminal roadmap, Rydberg atoms have become atomic physics’ E. coli for investigating the interaction of radiation with matter,” wrote Ketterle in his nomination for Kleppner’s 2017 APS Medal for Exceptional Achievement in Research. “They are being used by others in quests for experimental systems to realize Schrödinger’s cat, as well as for making a quantum computer.”

In 1981, Kleppner suggested in a theoretical paper the possibility of suppressing spontaneous emission with a cavity: excited atoms cannot decay when the cavity lacks the oscillatory modes to receive their emissions. This was followed by his demonstration of this effect, and launched the field of cavity quantum electrodynamics (cQED), the study of how light confined within a reflective cavity interacts with atoms or other particles. This field has led to the creation of new lasers and photonic devices.

“This work fundamentally changed the way physicists regard the process of spontaneous emission by showing that it is not a fixed property of a quantum state, but can be modified and controlled,” said Ketterle. “Current applications of these principles, which Dan terms ‘wrecking the vacuum,’ include thresholdless lasers and the construction of photonic bandgap materials in which light propagation is forbidden at certain frequencies.”

MIT-Harvard Center for Ultracold Atoms

In 2000, Kleppner secured National Science Foundation funding to co-found the Center for Ultracold Atoms (CUA), an MIT-Harvard collaboration that linked RLE with the Harvard Department of Physics to explore the physics of ultracold atoms and quantum gases. Kleppner served as its first director until 2006, and was a member of a group that included MIT professors Ketterle, Pritchard, Vladan Vuletic, Martin W. Zwierlein, Paola Cappellaro PhD ’06, and Isaac Chuang ’90.

“Many centers disappear after 10 to 20 years; sometimes their mission is fulfilled,” says Ketterle, the CUA director from 2006 to 2023. “But given the excitement and the rapid evolution in atomic physics, the CUA is a super-active center brimming with excitement, and we just recently got renewed. That’s partially due to the efforts of Dan. He created the tradition of atomic physics at MIT. We are one of the best atomic physics groups in the world. And we are really a family.”

Boost-phase intercept report

Kleppner co-authored a highly influential 2003 report that examined the technical feasibility of boost-phase intercept, a concept central to President George H.W. Bush’s proposed controversial Strategic Defense Initiative (SDI), nicknamed "Star Wars,” which purportedly would render nuclear weapons obsolete. The focus of the APS Study on Boost-Phase Intercept for National Missile Defense, published as a special supplement to Reviews of Modern Physics, was on the physics and engineering challenges of intercepting a missile during its boost phase.

“This was a subject on which I had no technical background at all,” Kleppner recalled, so he expressed gratitude for the skills of co-chair Fred Lamb of the University of Illinois. “But the APS [American Physical Society] felt that it was important to have information for the public … and no one knew anything about it. It was the point in my life where I could do that. And I feel that you have an obligation when the need arises and you can do it, to do that.”

The result? “Technically, it really would not succeed, except in very limited circumstances,” Kleppner said. Added Pritchard, “It vastly changed the path of the nation.”

“He was the perfect person to chair the committee,” says Ketterle. “He excelled in being neutral and unbiased, and to create a no-nonsense report. I think the APS was very proud of this report. It shows how physicists analyze something which was at that moment of immense political and societal importance. This report helped to understand what laser weapons cannot do and what they can do. The fact that (SDI) eventually, slowly, disappeared, the report may have contributed to that.”

Dedicated educator

Kleppner trained generations of physicists, including as advisor to 23 PhD students who have gone on to attain positions in major universities and achieve major scientific awards.

He was awarded the Oersted Medal of the American Association of Physics Teachers in 1997, and earned the Institute’s prestigious 1995-1996 James R. Killian, Jr. Faculty Achievement Award for his service to MIT and society on behalf of atomic physics. “He has given generously of his time and effort to the formation of national science policy, and he has served the Institute with distinction as teacher, administrator and counselor,” the Killian committee wrote.

Kleppner and Ramsey wrote the widely used text “Quick Calculus” in 1972 — the third edition of the book was updated in 2022 edition with MIT Department of Physics’ Peter Dourmashkin. With Robert J. Kolenkow, Kleppner also wrote “An Introduction to Mechanics” in 1973, and its second edition in 2013. Physics department head Deepto Chakrabarty ’88 called it “a masterpiece:” “It has formed the foundation of our freshman 8.012 course for potential physics majors for over 50 years and has provided a deep, elegant, and mathematically sophisticated introduction to classical mechanics of physics majors across the U.S. It was my own introduction to serious physics as an MIT freshman in 1984.”

Recently, while Kleppner was being wheeled into surgery, one of the medical personnel noticed that his patient was the author of that book and blurted out, “Oh my God, I still am wondering about one of those problems that I found so difficult,” recalls his wife, Bea, laughing.

Kleppner called his method of teaching “an engagement with the students and with the subject.” He said that his role model for teaching was his wife, who taught psychology at Beaver Country Day High School. “Fortunately, at MIT, the students are so great. There’s nothing tough about teaching here, except trying to stay ahead of the students.”

He leaves a legacy of grateful physicists impacted by his generous teaching style.

“I’ve always felt that I’ve just been incredibly lucky to be part of Dan’s group,” says Phillips, who was at Princeton when his research into magnetic resonance caught Kleppner’s attention, and invited him to MIT. “Dan extended this idea to putting this hydrogen maser in a much higher magnetic field. Not that many people are trained by somebody like Dan Kleppner in the art of precision measurement.”

Kleppner also gifted Phillips an apparatus he built for his thesis, which shaved years off the laser cooling experiments that led to Phillips’ Nobel.

Ketterle credited Kleppner’s mentorship for his career at MIT. “He was an older, experienced person who believed in me. He had more trust in me than I had initially myself. I felt whenever I was at a crossroads, I could go to Dan and ask him for advice. When I gave him a paper to edit … there was red ink all over it, but he was absolutely right on almost everything.’”

In 2003, Kleppner was dismayed at the statistic that over 60 percent of middle and high school teachers teaching physics have no background in the subject. He started the CUA’s Teaching Opportunities in Physical Science summer program with his then-postdoc Ted Ducas to train physics majors to prepare and teach physics material to middle and high school students. In its 14-year run, they worked with 112 students.

According to Ducas, one survey “indicates over 60 percent of our undergraduates have gone into, or plan to go into, pre-college teaching — a higher percentage than expected, because physics majors have so many other career opportunities often paying significantly more. The potential positive impact of that number of highly qualified and motivated teachers is dramatic.”

Kleppner also partnered with Japanese mathematician Heisuke Hironaka on the mentoring program Japanese Association for Mathematical Sciences (JAMS), which connected American college science students with their Japanese counterparts. “His interest in ensuring that future generations also see the value of international communities was reflected in JAMS,” says Sofie Kleppner.

Recognitions and public service

Kleppner was promoted to professor in 1974 and headed the physics department’s Division of Atomic, Plasma and Condensed Matter Physics from 1976 to 1979. He was named the Lester Wolfe Professor of Physics in 1985.

Active in the interface between physics and public policy, he served on more than 30 committees. For the APS, he was on the Panel on Public Affairs (POPA), chaired the Physics Planning Committee and the Division of Atomic, Molecular and Optical Physics, and contributed to a study on the growth and mentorship of young physics professors. He chaired a report for the National Academy of Sciences on atomic physics that he presented on various congressional committees, served on the National Research Council's Physics Survey Committee, and was chair of the International Union of Pure and Applied Physics’ Commission on Atomic and Molecular Physics. At MIT, he was also an ombuds of the Physics Department.

Kleppner was a fellow of the American Academy of Arts and Sciences, American Association for the Advancement of Science, OSA (now Optica), French Academy of Sciences, and the American Philosophical Society; a member of the National Academy of Sciences; and a Phi Beta Kappa lecturer.

His interest in literature at Williams bloomed into a secondary career as a writer, including decades of writing witty and insightful, yet accessible, pieces for Physics Today, including his “Reference Frame” columns on physics history and policy.

Kleppner was a recipient of many awards, including the prestigious Wolf Prize in 2005 “for groundbreaking work in atomic physics of hydrogenic systems, including research on the hydrogen maser, Rydberg atoms, and Bose-Einstein condensation.” Other accolades include a 2014 Benjamin Franklin Medal and a 2006 National Medal of Science, presented by U.S. President George W. Bush. He also received the Frederic Ives Medal (2007), the William F. Meggers Award (1991), the Lilienfeld Prize (1991), and the Davisson-Germer Prize (1986).

His articles, congressional testimony, and advocating on behalf of physicists around the world at one point inspired his Physics Planning Committee colleagues to present him with a Little League trophy of a golden baseball player, with the inscription “Dan Kleppner — Who Went to Bat for Atomic Physics.”

Kleppner said that he was inspired by his mentor, Ramsey, to get involved in the scientific community. “It’s a privilege to be a scientist in this country,” said Kleppner. “And I think that one has some obligation to pay for the privilege, when you can.”

He wrote, “Any scenario for a decent future of our nation and the world must include a reasonable component of science that is devoted to the search for new knowledge. We cannot afford to abandon this vision under a barrage of criticism, no matter how eloquent or powerful the critics.”

Family and retired life

Kleppner met his future wife, Beatrice Spencer, in 1954 on the USS United States, when both were England-bound and in their second year of studying at Cambridge. They began as friends, and eventually married in 1958, in Ipswich, Massachusetts. They raised their three children, Sofie, Paul, and Andrew, at their home in Belmont, Massachusetts, and their vacation home in Vermont.

Kleppner’s family described him as an optimist who didn’t believe in lying, worrying, or unethical behavior. He and Bea generously invited into their home anyone in need. “When we were growing up, we had the international community in our house,” recalls Sofie. “He was just a tremendously generous person. At my father’s 80th birthday celebration at MIT, there were three hours of five-minute reminiscences. It was really moving to hear the number of people who felt that just having the open door at my parents’ house meant the difference to them as they went through difficult times.”

In his retirement, Kleppner continued with his woodworking projects, including building beds, lamps, cabinets, a beautiful spiral staircase, a cradle curved like the hull of a boat, and bookcases featuring super ellipses, a closed curve that blends elements of an ellipse and a rectangle.

“I enjoy designing,” he said in one video. “It’s the same instinct for making things work in experimental physics. It’s lovely to make a piece of apparatus that starts functioning, and even if the experiment doesn’t do what you want it to do. There’s always a lot of jubilation when the apparatus is first turned on and first works.”

His last article for Physics Today was in 2020. In his later years, he kept in touch with his colleagues, swapping book ideas with Ketterle’s wife, Michele Plott, and, since the Covid-19 pandemic, maintained regular Zoom meetings with a group of his former students, hosted by Mike Kash; and another, what they called “The Famous Physicists,” that included Phillips and  their Brazilian colleague Vanderlei Bagnato.

“In recent years, I would still go to Dan for advice about difficult questions,” says Phillips, “sometimes about physics, sometimes just about life and public policy, because maybe I always felt that if there was anything you wanted done in which physics or science was part of the question that Dan would be the best person to do it.”

His family says that Kleppner suddenly fell ill at a Father’s Day dinner. According to his wife, his last words before being rushed to the hospital were a toast to his grandson, who recently graduated high school: “To Darwin and all youth who have new and exciting ideas.”

Says Bea, “​​He always said that you have to be optimistic to be a scientist, because you have to be patient. Things don’t work out and they’re fiddly, and there are lots of things that go wrong. His last words were ones that make you feel there’s hope for the future.”

The National Academy of Sciences (NAS) has elected 120 members and 30 international members, including five MIT faculty members and 13 MIT alumni. Professors Rodney Brooks, Parag Pathak, Scott Sheffield, Benjamin Weiss, and Yukiko Yamashita were elected in recognition of their “distinguished and continuing achievements in original research.” Membership to the National Academy of Sciences is one of the highest honors a scientist can receive in their career.

Elected MIT alumni include: David Altshuler ’86, Rafael Camerini-Otero ’66, Kathleen Collins PhD ’92, George Daley PhD ’89, Scott Doney PhD ’91, John Doyle PhD ’91, Jonathan Ellman ’84, Shanhui Fan PhD ’97, Julia Greer ’97, Greg Lemke ’78, Stanley Perlman PhD ’72, David Reichman PhD ’97, and Risa Wechsler ’96. 

Those elected this year bring the total number of active members to 2,662, with 556 international members. The NAS is a private, nonprofit institution that was established under a congressional charter signed by President Abraham Lincoln in 1863. It recognizes achievement in science by election to membership, and — with the National Academy of Engineering and the National Academy of Medicine — provides science, engineering, and health policy advice to the federal government and other organizations.

Rodney Brooks

Rodney A. Brooks is the Panasonic Professor of Robotics Emeritus at MIT and the chief technical officer and co-founder of Robust AI. Previously, he was founder, chair, and CTO of Rethink Robotics and founder and CTO of iRobot Corp. He is also the former director of the MIT Artificial Intelligence Laboratory and the MIT Computer Science and Artificial Intelligence Laboratory. Brooks received degrees in pure mathematics from the Flinders University of South Australia and a PhD in computer science from Stanford University in 1981. He held research positions at Carnegie Mellon University and MIT, and a faculty position at Stanford before joining the faculty of MIT in 1984.

Brooks’ research is concerned with both the engineering of intelligent robots to operate in unstructured environments, and with understanding human intelligence through building humanoid robots. He has published papers and books in model-based computer vision, path planning, uncertainty analysis, robot assembly, active vision, autonomous robots, micro-robots, micro-actuators, planetary exploration, representation, artificial life, humanoid robots, and compiler design.

Brooks is a member of the National Academy of Engineering, a founding fellow of the Association for the Advancement of Artificial Intelligence, a fellow of the American Academy of Arts and Sciences, the American Association for the Advancement of Science, the Association for Computing Machinery, a foreign fellow of The Australian Academy of Technological Sciences and Engineering, and a corresponding member of the Australian Academy of Science. He won the Computers and Thought Award at the 1991 International Joint Conference on Artificial Intelligence, and the IEEE Founders Medal in 2023.

Parag Pathak

Parag Pathak is the Class of 1922 Professor of Economics and a founder and director of MIT’s Blueprint Labs. He joined the MIT faculty in 2008 after completing his PhD in business economics and his master’s and bachelor’s degrees in applied mathematics, all at Harvard University.

Pathak is best known for his work on market design and education. His research has informed student placement and school choice mechanisms across the United States, including in Boston, New York City, Chicago, and Washington, and his recent work applies ideas from market design to the rationing of vital medical resources. Pathak has also authored leading studies on school quality, charter schools, and affirmative action. In urban economics, he has measured the effects of foreclosures on house prices and how the housing market reacted to the end of rent control in Cambridge, Massachusetts.

Pathak’s research on market design was recognized with the 2018 John Bates Clark Medal, given by the American Economic Association to the economist under 40 whose work is judged to have made the most significant contribution to the field. He is a fellow of the American Academy of Arts and Sciences, the Econometric Society, and the Society for the Advancement of Economic Theory. Pathak is also the founding co-director of the market design working group at the National Bureau of Economic Research, and a co-founder of Avela Education.

Scott Sheffield

Scott Sheffield, Leighton Family Professor of Mathematics, joined the MIT faculty in 2008 after a faculty appointment at the Courant Institute at New York University. He received a PhD in mathematics from Stanford University in 2003 under the supervision of Amir Dembo, and completed BA and MA degrees in mathematics from Harvard University in 1998.

Sheffield is a probability theorist, working on geometrical questions that arise in such areas as statistical physics, game theory, and metric spaces, as well as long-standing problems in percolation theory and the theory of random surfaces.

In 2017, Sheffield received the Clay Research Award with Jason Miller, “in recognition of their groundbreaking and conceptually novel work on the geometry of Gaussian free field and its application to the solution of open problems in the theory of two-dimensional random structures.” In 2023, he received the Leonard Eisenbud Prize with Jason Miller “for works on random two-dimensional geometries, and in particular on Liouville Quantum Gravity.” Later in 2023, Sheffield received the Frontiers of Science Award with Jason Miller for the paper “Liouville quantum gravity and the Brownian map I: the QLE(8/3,0) metric.” Sheffield is a fellow of the American Academy of Arts and Science.

Benjamin Weiss

Benjamin Weiss is the Robert R. Schrock Professor of Earth and Planetary Sciences. He studied physics at Amherst College as an undergraduate and went on to study planetary science and geology at Caltech, where he earned a master’s degree in 2001 and PhD in 2003. Weiss’ doctoral dissertation on Martian meteorite ALH 84001 revealed records of the ancient Martian climate and magnetic field, and provided evidence some meteorites could transfer materials from Mars to Earth without heat-sterilization. Weiss became a member of the Department of Earth, Atmospheric and Planetary Sciences faculty in 2004 and is currently chair of the Program in Planetary Science.

A specialist in magnetometry, Weiss seeks to understand the formation and evolution of the Earth, terrestrial planets, and small solar system bodies through laboratory analysis, spacecraft observations, and fieldwork. He is known for key insights into the history of our solar system, including discoveries about the early nebular magnetic field, the moon’s long-lived core dynamo, and asteroids that generated core dynamos in the past. In addition to leadership roles on current, active NASA missions — as deputy principal investigator for Psyche, and co-investigator for Mars Perseverance and Europa Clipper — Weiss has also been part of science teams for the SpaceIL Beresheet, JAXA Hayabusa 2, and ESA Rosetta spacecraft.

As principal investigator of the MIT Planetary Magnetism Laboratory, Weiss works to develop high-sensitivity, high-resolution techniques in magnetic microscopy to image the magnetic fields embedded in rock samples collected from meteorites, the lunar surface, and sites around the Earth. Studying these magnetic signatures can help answer questions about the conditions of the early solar system, past climates on Earth and Mars, and factors that promote habitability.

Yukiko Yamashita

Yukiko Yamashita is a professor of biology at MIT, a core member of the Whitehead Institute for Biomedical Research, and an investigator at the Howard Hughes Medical Institute (HHMI). Yamashita earned her BS in biology in 1994 and her PhD in biophysics in 1999 from Kyoto University. From 2001 to 2006, she did postdoctoral research at Stanford University. She was appointed to the University of Michigan faculty in 2007 and was named an HHMI Investigator in 2014. She became a member of the Whitehead Institute and a professor of biology at MIT in 2020.

Yukiko Yamashita studies two fundamental aspects of multicellular organisms: how cell fates are diversified via asymmetric cell division, and how genetic information is transmitted through generations via the germline.

Two remarkable feats of multicellular organisms are generation of many distinct cell types via asymmetric cell division and transmission of the germline genome to the next generation, essentially in eternity. Studying these processes using the Drosophila male germline as a model system has led us to venture into new areas of study, such as functions of satellite DNA, “genomic junk,” and how they might be involved in speciation.

Yamashita is a member of the American Academy of Arts and Sciences, a fellow of the American Society for Cell Biology, and the winner of the Tsuneko and Reiji Okazaki Award in 2016. She was named a MacArthur Fellow in 2011.

Researchers at MIT and other institutions have identified compounds that can fight off viral infection by activating a defense pathway inside host cells. These compounds, they believe, could be used as antiviral drugs that work against not just one but any kind of virus.

The researchers identified these compounds, which activate a host cell defense system known as the integrated stress response pathway, in a screen of nearly 400,000 molecules. In tests in human cells, the researchers showed that the compounds help cells fend off infection from RSV, herpes virus, and Zika virus. They also proved effective in combating herpes infection in a mouse model.

The research team now plans to test the compounds against additional viruses, in hopes of developing them for eventual clinical trials.

“We’re very excited about this work, which allows us to harness the stress response of the host cells to arrive at a means to identify and develop broad-spectrum antivirals,” says James Collins, the Termeer Professor of Medical Engineering and Science in MIT’s Institute for Medical Engineering and Science (IMES) and Department of Biological Engineering.

Collins and Maxwell Wilson, an associate professor of molecular biology at the University of California, Santa Barbara and chief scientific officer of Integrated Biosciences, are the senior authors of the new study, which appears in Cell. Felix Wong, a former MIT postdoc and chief executive officer of Integrated Biosciences, is the lead author of the paper. In addition to MIT, UCSB, and Integrated Biosciences, the research team also includes scientists from Illumina Ventures and Princeton University.

Boosting cell defense

In human cells, the integrated stress response pathway is turned on in response to viral infection as well as other types of stress such as starvation. During viral infection, the pathway is triggered by double-stranded RNA, a molecule produced during the replication cycle of viruses. When that RNA is detected, the cell shuts down protein synthesis, which blocks the virus from producing the proteins it needs to replicate.

Compounds that boost this pathway, the researchers believe, could be good candidates for new antiviral drugs that could combat any type of virus.

“Typically, how antivirals are developed is that you develop one antiviral for one specific virus,” Wong says. “In this case, we hypothesized that being able to modulate the host cell stress response might give us a new class of broad-spectrum antivirals — compounds that directly act on the host cells to alter something fundamental about how all viruses replicate.”

To help them identify compounds that would enhance the activity of this pathway during viral infection, the researchers invented a novel optogenetic screen. Optogenetics is a bioengineering technique that allows researchers to insert light-sensitive proteins into the genome of a cell. In this case, the researchers engineered modifications to a protein called PKR, which turns on the stress pathway, so that they could turn it on with light.

Using this technique, the researchers screened a library of nearly 400,000 commercially available and proprietary chemical compounds. Each of these compounds was applied to human cells as the cells were also exposed to blue light, which simulated viral infection by activating PKR.

By measuring the cells’ survival rates, the researchers could determine which compounds boosted activation of the pathway and amplified the cells’ ability to shut down viral reproduction. This screen yielded about 3,500 compounds with potential antiviral activity, which were evaluated further.

“If the pathway were turned on in response to viral infection, what our compounds do is they turn it on full blast,” Wong says. “Even in the presence of a small amount of virus, if the pathway is triggered, then the antiviral response is also maximized.”

Fighting infection

The researchers then selected eight of the most promising compounds and screened them for their ability to kill viruses while avoiding harmful effects in human cells. Based on these tests, the researchers chose three top candidates, which they called IBX-200, IBX-202, and IBX-204.

In cells that were infected with either Zika virus, herpes virus, or RSV, treatment with these compounds significantly reduced the amount of virus in the cells. The researchers then tested one of the compounds, IBX-200, in mice infected with herpes virus, and found that it was able to reduce the viral load and improve symptoms.

Experiments showed that these compounds appear to turn on an enzyme that is involved in detecting stress. This activates the stress response pathway and primes the cells to become more responsive to viral infection. When applied to cells that are not already infected, the compounds have no effect.

The researchers now plan to evaluate their lead candidates against a broader range of viruses. They also aim to identify additional compounds that activate the integrated stress response, as well as other cellular stress pathways with the potential to clear viral or bacterial infections.

The research was funded by the Defense Threat Reduction Agency, the National Science Foundation, the U.S. Army Research Office, and Integrated Biosciences.

When ChatGPT or Gemini give what seems to be an expert response to your burning questions, you may not realize how much information it relies on to give that reply. Like other popular generative artificial intelligence (AI) models, these chatbots rely on backbone systems called foundation models that train on billions, or even trillions, of data points.

In a similar vein, engineers are hoping to build foundation models that train a range of robots on new skills like picking up, moving, and putting down objects in places like homes and factories. The problem is that it’s difficult to collect and transfer instructional data across robotic systems. You could teach your system by teleoperating the hardware step-by-step using technology like virtual reality (VR), but that can be time-consuming. Training on videos from the internet is less instructive, since the clips don’t provide a step-by-step, specialized task walk-through for particular robots.

A simulation-driven approach called “PhysicsGen” from MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) and the Robotics and AI Institute customizes robot training data to help robots find the most efficient movements for a task. The system can multiply a few dozen VR demonstrations into nearly 3,000 simulations per machine. These high-quality instructions are then mapped to the precise configurations of mechanical companions like robotic arms and hands. 

PhysicsGen creates data that generalize to specific robots and condition via a three-step process. First, a VR headset tracks how humans manipulate objects like blocks using their hands. These interactions are mapped in a 3D physics simulator at the same time, visualizing the key points of our hands as small spheres that mirror our gestures. For example, if you flipped a toy over, you’d see 3D shapes representing different parts of your hands rotating a virtual version of that object.

The pipeline then remaps these points to a 3D model of the setup of a specific machine (like a robotic arm), moving them to the precise “joints” where a system twists and turns. Finally, PhysicsGen uses trajectory optimization — essentially simulating the most efficient motions to complete a task — so the robot knows the best ways to do things like repositioning a box.

Each simulation is a detailed training data point that walks a robot through potential ways to handle objects. When implemented into a policy (or the action plan that the robot follows), the machine has a variety of ways to approach a task, and can try out different motions if one doesn’t work.

“We’re creating robot-specific data without needing humans to re-record specialized demonstrations for each machine,” says Lujie Yang, an MIT PhD student in electrical engineering and computer science and CSAIL affiliate who is the lead author of a new paper introducing the project. “We’re scaling up the data in an autonomous and efficient way, making task instructions useful to a wider range of machines.”

Generating so many instructional trajectories for robots could eventually help engineers build a massive dataset to guide machines like robotic arms and dexterous hands. For example, the pipeline might help two robotic arms collaborate on picking up warehouse items and placing them in the right boxes for deliveries. The system may also guide two robots to work together in a household on tasks like putting away cups.

PhysicsGen’s potential also extends to converting data designed for older robots or different environments into useful instructions for new machines. “Despite being collected for a specific type of robot, we can revive these prior datasets to make them more generally useful,” adds Yang.

Addition by multiplication

PhysicsGen turned just 24 human demonstrations into thousands of simulated ones, helping both digital and real-world robots reorient objects.

Yang and her colleagues first tested their pipeline in a virtual experiment where a floating robotic hand needed to rotate a block into a target position. The digital robot executed the task at a rate of 81 percent accuracy by training on PhysicGen’s massive dataset, a 60 percent improvement from a baseline that only learned from human demonstrations.

The researchers also found that PhysicsGen could improve how virtual robotic arms collaborate to manipulate objects. Their system created extra training data that helped two pairs of robots successfully accomplish tasks as much as 30 percent more often than a purely human-taught baseline.

In an experiment with a pair of real-world robotic arms, the researchers observed similar improvements as the machines teamed up to flip a large box into its designated position. When the robots deviated from the intended trajectory or mishandled the object, they were able to recover mid-task by referencing alternative trajectories from their library of instructional data.

Senior author Russ Tedrake, who is the Toyota Professor of Electrical Engineering and Computer Science, Aeronautics and Astronautics, and Mechanical Engineering at MIT, adds that this imitation-guided data generation technique combines the strengths of human demonstration with the power of robot motion planning algorithms.

“Even a single demonstration from a human can make the motion planning problem much easier,” says Tedrake, who is also a senior vice president of large behavior models at the Toyota Research Institute and CSAIL principal investigator. “In the future, perhaps the foundation models will be able to provide this information, and this type of data generation technique will provide a type of post-training recipe for that model.”

The future of PhysicsGen

Soon, PhysicsGen may be extended to a new frontier: diversifying the tasks a machine can execute.

“We’d like to use PhysicsGen to teach a robot to pour water when it’s only been trained to put away dishes, for example,” says Yang. “Our pipeline doesn’t just generate dynamically feasible motions for familiar tasks; it also has the potential of creating a diverse library of physical interactions that we believe can serve as building blocks for accomplishing entirely new tasks a human hasn’t demonstrated.”

Creating lots of widely applicable training data may eventually help build a foundation model for robots, though MIT researchers caution that this is a somewhat distant goal. The CSAIL-led team is investigating how PhysicsGen can harness vast, unstructured resources — like internet videos — as seeds for simulation. The goal: transform everyday visual content into rich, robot-ready data that could teach machines to perform tasks no one explicitly showed them.

Yang and her colleagues also aim to make PhysicsGen even more useful for robots with diverse shapes and configurations in the future. To make that happen, they plan to leverage datasets with demonstrations of real robots, capturing how robotic joints move instead of human ones.

The researchers also plan to incorporate reinforcement learning, where an AI system learns by trial and error, to make PhysicsGen expand its dataset beyond human-provided examples. They may augment their pipeline with advanced perception techniques to help a robot perceive and interpret their environment visually, allowing the machine to analyze and adapt to the complexities of the physical world.

For now, PhysicsGen shows how AI can help us teach different robots to manipulate objects within the same category, particularly rigid ones. The pipeline may soon help robots find the best ways to handle soft items (like fruits) and deformable ones (like clay), but those interactions aren’t easy to simulate yet.

Yang and Tedrake wrote the paper with two CSAIL colleagues: co-lead author and MIT PhD student Hyung Ju “Terry” Suh SM ’22 and MIT PhD student Bernhard Paus Græsdal. Robotics and AI Institute researchers Tong Zhao ’22, MEng ’23, Tarik Kelestemur, Jiuguang Wang, and Tao Pang PhD ’23 are also authors. Their work was supported by the Robotics and AI Institute and Amazon.

The researchers recently presented their work at the Robotics: Science and Systems conference.

In order to produce effective targeted therapies for cancer, scientists need to isolate the genetic and phenotypic characteristics of cancer cells, both within and across different tumors, because those differences impact how tumors respond to treatment.

Part of this work requires a deep understanding of the RNA or protein molecules each cancer cell expresses, where it is located in the tumor, and what it looks like under a microscope.

Traditionally, scientists have looked at one or more of these aspects separately, but now a new deep learning AI tool, CellLENS (Cell Local Environment and Neighborhood Scan), fuses all three domains together, using a combination of convolutional neural networks and graph neural networks to build a comprehensive digital profile for every single cell. This allows the system to group cells with similar biology — effectively separating even those that appear very similar in isolation, but behave differently depending on their surroundings.

The study, published recently in Nature Immunology, details the results of a collaboration between researchers from MIT, Harvard Medical School, Yale University, Stanford University, and University of Pennsylvania — an effort led by Bokai Zhu, an MIT postdoc and member of the Broad Institute of MIT and Harvard and the Ragon Institute of MGH, MIT, and Harvard.

Zhu explains the impact of this new tool: “Initially we would say, oh, I found a cell. This is called a T cell. Using the same dataset, by applying CellLENS, now I can say this is a T cell, and it is currently attacking a specific tumor boundary in a patient.

“I can use existing information to better define what a cell is, what is the subpopulation of that cell, what that cell is doing, and what is the potential functional readout of that cell. This method may be used to identify a new biomarker, which provides specific and detailed information about diseased cells, allowing for more targeted therapy development.”

This is a critical advance because current methodologies often miss critical molecular or contextual information — for example, immunotherapies may target cells that only exist at the boundary of a tumor, limiting efficacy. By using deep learning, the researchers can detect many different layers of information with CellLENS, including morphology and where the cell is spatially in a tissue.

When applied to samples from healthy tissue and several types of cancer, including lymphoma and liver cancer, CellLENS uncovered rare immune cell subtypes and revealed how their activity and location relate to disease processes — such as tumor infiltration or immune suppression.

These discoveries could help scientists better understand how the immune system interacts with tumors and pave the way for more precise cancer diagnostics and immunotherapies.

“I’m extremely excited by the potential of new AI tools, like CellLENS, to help us more holistically understand aberrant cellular behaviors within tissues,” says co-author Alex K. Shalek, the director of the Institute for Medical Engineering and Science (IMES), the J. W. Kieckhefer Professor in IMES and Chemistry, and an extramural member of the Koch Institute for Integrative Cancer Research at MIT, as well as an Institute member of the Broad Institute and a member of the Ragon Institute. “We can now measure a tremendous amount of information about individual cells and their tissue contexts with cutting-edge, multi-omic assays. Effectively leveraging that data to nominate new therapeutic leads is a critical step in developing improved interventions. When coupled with the right input data and careful downsteam validations, such tools promise to accelerate our ability to positively impact human health and wellness.”

For many years, health experts have been concerned about “food deserts,” places where residents lack good nutritional options. Now, an MIT-led study of three major global cities uses a new, granular method to examine the issue, and concludes that having fewer and less nutritional eating options nearby correlates with obesity and other health outcomes.

Rather than just mapping geographic areas, the researchers examined the dietary value of millions of food items on roughly 30,000 restaurant menus and derived a more precise assessment of the connection between neighborhoods and nutrition.

“We show that what is sold in a restaurant has a direct correlation to people’s health,” says MIT researcher Fabio Duarte, co-author of a newly published paper outlining the study’s results. “The food landscape matters.”

The open-access paper, “Data-driven nutritional assessment of urban food landscapes: insights from Boston, London, Dubai,” was published this week in Nature: Scientific Reports.

The co-authors are Michael Tufano, a PhD student at Wageningen University, in the Netherlands; Duarte, associate director of MIT’s Senseable City Lab, which uses data to study cities as dynamic systems; Martina Mazzarello, a postdoc at the Senseable City Lab; Javad Eshtiyagh, a research fellow at the Senseable City Lab; Carlo Ratti, professor of the practice and director of the Senseable City Lab; and Guido Camps, a senior researcher at Wageningen University.

Scanning the menu

To conduct the study, the researchers examined menus from Boston, Dubai, and London, in the summer of 2023, compiling a database of millions of items available through popular food-delivery platforms. The team then evaluated the food items as rated by the USDA’s FoodData Central database, an information bank with 375,000 kinds of food products listed. The study deployed two main metrics, the Meal Balance Index, and the Nutrient-Rich Foods Index.

The researchers examined about 222,000 menu items from over 2,000 restaurants in Boston, about 1.6 million menu items from roughly 9,000 restaurants in Dubai, and about 3.1 million menu items from about 18,000 restaurants in London. In Boston, about 71 percent of the items were in the USDA database; in Dubai and London, that figure was 42 percent and 56 percent, respectively.

The team then rated the nutritional value of the items appearing on menus, and correlated the food data with health-outcome data from Boston and London. In London, they found a clear correlation between neighborhood menu offerings and obesity, or the lack thereof; with a slightly less firm correlation in Boston. Areas with food options that include a lot of dietary fibers, sometimes along with fruits and vegetables, tend to have better health data.

In Dubai, the researchers did not have the same types of health data available but did observe a strong correlation between rental prices and the nutritional value of neighborhood-level food, suggesting that wealthier residents have better nourishment options.

“At the item level, when we have less nutritional food, we see more cases of obsesity,” Tufano says. “It’s true that not only do we have more fast food in poor neighborhoods, but the nutritional value is not the same.”

Re-mapping the food landscape

By conducting the study in this fashion, the scholars added a layer of analysis to past studies of food deserts. While past work has broken ground by identifying neighborhoods and areas lacking good food access, this research makes a more comprehensive assessment of what people consume. The research moves toward evaluating the complex mix of food available in any given area, which can be true even of areas with more limited options.

“We were not satisfied with this idea that if you only have fast food, it’s a food desert, but if you have a Whole Foods, it’s not,” Duarte says. “It’s not necessarily like that.”

For the Senseable City Lab researchers, the study is a new technique further enabling them to understand city dynamics and the effects of the urban environment on health. Past lab studies have often focused on issues such as urban mobility, while extending to matters such as mobility and air pollution, among other topics.

Being able to study food and health at the neighborhood level, though, is still another example of the ways that data-rich spheres of life can be studied in close detail.

“When we started working on cities and data, the data resolution was so low,” Ratti says. “Today the amount of data is so immense we see this great opportunity to look at cities and see the influence of the urban environment as a big determinant of health. We see this as one of the new frontiers of our lab. It’s amazing how we can now look at this very precisely in cities.”

From the very beginning, MIT Professor Mark Bear’s philosophy for the textbook “Neuroscience: Exploring the Brain” was to provide an accessible and exciting introduction to the field while still giving undergraduates a rigorous scientific foundation. In the 30 years since its first print printing in 1995, the treasured 975-page tome has gone on to become the leading introductory neuroscience textbook, reaching hundreds of thousands of students at hundreds of universities around the world.

“We strive to present the hard science without making the science hard,” says Bear, the Picower Professor in The Picower Institute for Learning and Memory and the Department of Brain and Cognitive Sciences at MIT. The fifth edition of the textbook is out today from the publisher Jones & Bartlett Learning.

Bear says the book is conceived, written, and illustrated to instill students with the state of knowledge in the field without assuming prior sophistication in science. When he first started writing it in the late 1980s — in an effort soon joined by his co-authors and former Brown University colleagues Barry Connors and Michael Paradiso — there simply were no undergraduate neuroscience textbooks. Up until then, first as a graduate teaching assistant and then as a young professor, Bear taught Brown’s pioneering introductory neuroscience class with a spiral-bound stack of photocopied studies and other scrounged readings.

Don’t overwhelm

Because universities were only beginning to launch neuroscience classes and majors at the time, Bear recalls that it was hard to find a publisher. The demand was just too uncertain. With an unsure market, Bear says, the original publisher, Williams & Wilkins, wanted to keep costs down by printing only in black and white. But Bear and his co-authors insisted on color. Consistent with their philosophy for the book, they wanted students, even before they began reading, to be able to learn from attractive, high-quality illustrations.

“Rather than those that speak a thousand words, we wanted to create illustrations that each make a single point.” Bear says. “We don’t want to overwhelm students with a bunch of detail. If people want to know what’s in our book, just look at the pictures.”

Indeed, if the book had struck students as impenetrable and dull, Bear says, he and his co-authors would have squandered the advantage they had in presenting their subject: the inherently fascinating and exciting brain.

“Most good scientists are extremely enthusiastic about the science. It exciting. It’s fun. It turns them on,” Bear says. “We try to communicate the joy. We’re so lucky because the object of our affection is the brain.”

To help bring that joy and excitement across, another signature of the book throughout its 30-year-history has been the way it presents the process of discovery alongside the discoveries themselves, Bear says. While it’s instructive to provide students with the experimental evidence that supports the concepts they are learning, it would bog down the text to delineate the details of every experiment. Instead, Bear, Connors, and Paradiso have chosen to highlight the process of discovery via one-page guest essays by prominent neuroscientists who share their discovery stories personally. Each edition has featured about 25 such “Path of Discovery” essays, so more than 100 scientists have participated, including several Nobel Prize winners, such as the Picower Institute’s founding director, Susumu Tonegawa.

The new edition includes Path of Discovery essays by current Picower Institute Director Li-Huei Tsai and Picower Institute colleague Emery N. Brown. Tsai recounts her discovery that sensory stimulation of 40Hz rhythms in the brain can trigger a health-promoting response among many different cell types. Brown writes about how various biological cycles and rhythms in the brain and body, such as the circadian rhythms and brain waves, help organize our daily lives.

Immense impact

Jones & Bartlett reports that more than 470 colleges and universities in 48 U.S. states and the District of Columbia have used the fourth edition of the book. Various editions have also been translated into seven other languages, including Chinese, French, Portuguese, and Spanish. There are hundreds of reviews on Amazon.com with an average around 4.6 stars. One reviewer wrote about the fourth edition: “I never knew it was possible to love a textbook before!”

The reviews sometimes go beyond mere internet postings. Once, after Bear received an award in Brazil, he found himself swarmed at the podium by scores of students eager for him to sign their copies of the book. And earlier this year, when Bear needed surgery, the anesthesiologist was excited to meet him.

“The anesthesiologist was like, ‘Are you the Mark Bear who wrote the textbook?,’ and she was so excited, because she said, ‘This book changed my life,’” Bear recalls. “After I recovered, she showed up in the ICU for me to sign it. All of us authors have had this experience that there are people whose lives we’ve touched.”

While Bear is proud that so many students have benefited from the book, he also notes that teaching and textbook writing have benefited him as a scientist. They have helped him present his research more clearly, he says, and have given him a broad perspective on what’s truly important in the field.

“Experience teaching will influence the impact of your own science by making you more able to effectively communicate it.” Bear says. “And the teacher has a difficult job of surveying a field and saying, ‘I’ve got to capture the important advances and set aside the less-important stuff.’ It gives you a perspective that helps you to discriminate between more-important and less-important problems in your own research.”

Over the course of 30 years via their carefully crafted book, Bear, Connors, and Paradiso have lent that perspective to generations of students. And the next generation will start with today’s publishing of the new edition. 

The MIT Institute for Data, Systems, and Society (IDSS) announced the creation of a new endowed chair made possible by the generosity of IDSS professor post-tenure and “MIT lifer” Richard “Dick” Larson. Effective July 1, the fund provides a full professorship for senior IDSS faculty: the Distinguished Professorship in Data, Systems, and Society.

“As a faculty member, MIT has not only accepted but embraced my several mid-career changes of direction,” says Larson. “I have called five different academic departments my home, starting with Electrical Engineering (that is what it was called in the 1960s) and now finalized with the interdepartmental, interdisciplinary IDSS — Institute for Data, Systems and Society. Those beautiful three words  — data, systems, society — they represent my energy and commitment over the second half of my career. My gifted chair is an effort to keep alive those three words, with others following me doing research, teaching and mentoring centered around data, systems, society.”

Larson’s career has focused his operations research and systems expertise on a wide variety of problems, in both public and private sectors. His contributions span the fields of urban service systems (especially emergency response systems), disaster planning, pandemics, queueing, logistics, technology-enabled education, smart-energy houses, and workforce planning. His latest book, “Model Thinking for Everyday Life,” draws on decades of experience as a champion of STEM education at MIT and beyond, such as his leadership of MIT BLOSSOMS.

“Dick Larson has been making an impact at MIT for over half a century,” says IDSS Director Fotini Christia, the Ford International Professor in Political Science. “This gift extends his already considerable legacy and ensures his impact will continue to be felt for many years to come.”

Christia is pleased that IDSS and brain and cognitive science professor Alexander “Sasha” Rakhlin is the inaugural holder of the new professorship. The selection recognizes Rakhlin’s distinguished scholarly record, dedicated service to IDSS, excellence in teaching, and contributions to research in statistics and computation.

“Sasha’s analysis of neural network complexity, and his work developing tools for online prediction, are perfect examples of research which builds bridges across disciplines, and also connects different departments and units at MIT,” says Michale Fee, the Glen V. and Phyllis F. Dorflinger Professor of Neuroscience, and head of the Department of Brain and Cognitive Sciences. “It’s wonderful to see Sasha’s contributions recognized in this way, and I’m grateful to Dick Larson for supporting this vision.”

Rakhlin’s research is in machine learning, with an emphasis on statistics and computation. He is interested in formalizing the process of learning, in analyzing learning models, and in deriving and implementing emerging learning methods. A significant thrust of his research is in developing theoretical and algorithmic tools for online prediction, a learning framework where data arrives in a sequential fashion.

“I am honored to be the inaugural holder of the Distinguished Professorship in Data, Systems, and Society,” says Rakhlin. “Professor Larson’s commitment to education and service to MIT both serve as models to follow.”

A new security screener that people can simply walk past may soon be coming to an airport near you. Last year, U.S. airports nationwide began adopting HEXWAVE — a commercialized walkthrough security screening system based on microwave imaging technology developed at MIT Lincoln Laboratory — to satisfy a new Transportation Security Administration (TSA) mandate for enhanced employee screening to detect metallic and nonmetallic threats. The TSA is now in the process of evaluating HEXWAVE as a potential replacement of metal detectors to screen PreCheck passengers.

Typically, when you arrive at an airport security checkpoint line, you place your carry-on items on the conveyer belt, remove your shoes and any metallic items, and enter a body scanner. As you hold still for a few seconds with your feet spread apart and your arms extended over your head, the scanner creates a generic, featureless 3D body outline revealing any metallic or nonmetallic concealed weapons or other prohibited items.

Requiring individuals to stop, remove clothing and belongings, and pose for scans impedes traffic flow in airports and other highly populated venues, such as stadiums, shopping malls, mass transit stations, and schools. To enable more efficient screening of unstructured crowds and ensure public safety, the Department of Homeland Security (DHS) Science and Technology Directorate (S&T) sponsored Lincoln Laboratory to prototype a high-resolution imaging system capable of scanning people and their belongings as they walk by. This R&D effort was conducted as part of S&T's Surface Transportation Explosive Threat Detection Program, which aims to provide the surface-transportation end user-community (e.g., mass transit) with a layered and integrated capability to detect threat items at the speed of the traveling public.

The laboratory's prototype microwave imager, which consists of a set of antennas installed on flat panels, operates under the same fundamental principle as existing body scanners: low-energy radio waves (less powerful than those transmitted by a cellphone) are transmitted from antennas toward a person's body and reflect off skin and any hidden objects; the reflected waves return to the antennas and are processed by a computer to create an image, which security personnel then review to identify any potential concealed threats.

The novelty of the laboratory's invention lies in its ability to discreetly handle a constant stream of subjects in motion, measuring each subject very quickly (within tens of milliseconds) and reconstructing 3D microwave images of each subject at a video rate. To meet these challenging requirements, the laboratory team developed a cost-effective antenna array and efficient image-reconstruction algorithms. Compared to existing systems, the laboratory's 3D microwave imager runs 100 times faster using the same computing hardware. In 2017, the team demonstrated the prototype's ability to detect various simulated threat items at varying distances on a rail platform at the Massachusetts Bay Transit Authority (MBTA) Emergency Training Center in Boston.

"The goal of our work is to provide security staff with more effective tools to protect public spaces. To that end, microwave imaging technology can quickly and unobtrusively provide visibility of items carried into a venue," says William Moulder, who led the technology's development at Lincoln Laboratory.

In 2018, the security company Liberty Defense licensed the imaging technology and entered into a cooperative research and development agreement (CRADA) with Lincoln Laboratory. Transitioning technology to industry for commercialization is part of the laboratory's role as a federally funded research and development center, and CRADAs provide a mechanism for such transition to happen. Through the CRADA, Liberty Defense maintained Lincoln Laboratory's core image-reconstruction intellectual property and made the technology enhancements required for commercialization, including an entirely new hardware architecture, radio frequency (RF) antenna modules, and a transceiver system that meets Federal Communications Commission waveform and RF performance requirements for indoor and outdoor operation. The co-organizational team facilitating the transition of the technology was recognized by the Federal Laboratory Consortium for Technology Transfer with a 2019 Excellence in Technology Transfer Award for the Northeast region.

By 2021, Liberty Defense had prototyped a walk-through security screening system, HEXWAVE. That same year, through the TSA's On-Person Screening Capability Program, Liberty Defense received a contract award to demonstrate HEXWAVE's enhanced threat-detection and high-throughput capabilities for screening aviation workers. Following successful testing of HEXWAVE at sports complexes, entertainment arenas, and shopping centers, both nationally and internationally, Liberty Defense began offering the product for sale.

"HEXWAVE is a great example of how federally funded R&D can be successfully transitioned to industry to meet real-world security needs," says Asha Rajagopal, the laboratory's chief technology transfer officer. "By working with Liberty Defense, we helped accelerate the delivery of a critical capability into the hands of those protecting public spaces."

In 2023, TSA began testing HEXWAVE as a potential replacement of metal detectors used to screen passengers in TSA PreCheck lanes. Airports across the United States started deploying HEXWAVE in 2024 to meet the TSA's employee screening mandate by the April 2026 deadline. Liberty Defense notes various other markets for HEXWAVE; the first units for commercial applications were delivered to Los Alamos National Laboratory in 2023, and the technology has since been deployed at other national labs, correctional facilities, government buildings, and courthouses.

"Liberty was extremely fortunate to license the technology from MIT Lincoln Laboratory," says Bill Frain, CEO of Liberty Defense. "From the outset, they've been a true partner — bringing not only deep innovation and technical expertise, but also a clear vision for commercial deployment. Together, we've successfully brought next-generation technology to market to help protect people in public spaces."

In addition to the typical rigors of MIT classes, Terrascope Subject 2.00C/1.016/EC.746 (Design for Complex Environmental Issues) poses some unusual hurdles for students to navigate: collaborating across time zones, bridging different cultural and institutional experiences, and trying to do hands-on work over Zoom. That’s because the class includes students from not only MIT, but also Diné College in Tsaile, Arizona, within the Navajo Nation, and the University of Puerto Rico-Ponce (UPRP).

Despite being thousands of miles apart, students work in teams to tackle a real-world problem for a client, based on the Terrascope theme for the year. “Understanding how to collaborate over long distances with people who are not like themselves will be an important item in many of these students’ toolbelts going forward, in some cases just as much as — or more than — any particular design technique,” says Ari Epstein, Terrascope associate director and senior lecturer. Over the past several years, Epstein has taught the class along with Joel Grimm of MIT Beaver Works and Libby Hsu of MIT D-Lab, as well instructors from the two collaborating institutions. Undergraduate teaching fellows from all three schools are also key members of the instructional staff.

Since the partnership began three years ago (initially with Diné College, with the addition of UPRP two years ago), the class themes have included food security and sustainable agriculture in Navajo Nation; access to reliable electrical power in Puerto Rico; and this year, increasing museum visitors’ engagement with artworks depicting mining and landscape alteration in Nevada.

Each team — which includes students from all three colleges — meets with clients online early in the term to understand their needs; then, through an iterative process, teams work on designing prototypes. During MIT’s spring break, teams travel to meet with the clients onsite to get feedback and continue to refine their prototypes. At the end of the term, students present their final products to the clients, an expert panel, and their communities at a hybrid showcase event held simultaneously on all three campuses.

Free-range design engineering

“I really loved the class,” says Graciela Leon, a second-year mechanical engineering major who took the subject in 2024. “It was not at all what I was expecting,” she adds. While the learning objectives on the syllabus are fairly traditional — using an iterative engineering design process, developing teamwork skills, and deepening communication skills, to name a few — the approach is not. “Terrascope is just kind of like throwing you into a real-world problem … it feels a lot more like you are being trusted with this actual challenge,” Leon says.

The 2024 challenge was to find a way to help the clients, Puerto Rican senior citizens, turn on gasoline-powered generators when the electrical power grid fails; some of them struggle with the pull cords necessary to start the generators. The students were tasked with designing solutions to make starting the generators easier.

Terrascope instructors teach fundamental skills such as iterative design spirals and scrum workflow frameworks, but they also give students ample freedom to follow their ideas. Leon admits she was a bit frustrated at first, because she wasn’t sure what she was supposed to be doing. “I wanted to be building things and thought, ‘Wow, I have to do all these other things, I have to write some kind of client profile and understand my client’s needs.’ I was just like, ‘Hand me a drill! I want to design something!’”

When he took the class last year, Uziel Rodriguez-Andujar was also thrown off initially by the independence teams had. Now a second-year UPRP student in mechanical engineering, he’s accustomed to lecture-based classes. “What I found so interesting is the way [they] teach the class, which is, ‘You make your own project, and we need you to find a solution to this. How it will look, and when you have it — that’s up to you,’” he says.

Clearing hurdles

Teaching the course on three different campuses introduces a number of challenges for students and instructors to overcome — among them, operating in three different time zones, overcoming language barriers, navigating different cultural and institutional norms, communicating effectively, and designing and building prototypes over Zoom.

“The culture span is huge,” explains Epstein. “There are different ways of speaking, different ways of listening, and each organization has different resources.”

First-year MIT student EJ Rodriguez found that one of the biggest obstacles was trying to convey ideas to teammates clearly. He took the class this year, when the theme revolved around the environmental impacts of lithium mining. The client, the Nevada Museum of Art, wanted to find ways to engage visitors with its artwork collection related to mining-related landscape changes.

Rodriguez and his team designed a pendulum with a light affixed to it that illuminates a painting by a Native American artist. When the pendulum swings, it changes how the visitor experiences the artwork. The team built parts for the pendulum on different campuses, and they reached a point where they realized their pieces were incompatible. “We had different visions of what we wanted for the project, and different vocabulary we were using to describe our ideas. Sometimes there would be a misunderstanding … It required a lot of honesty from each campus to be like, ‘OK, I thought we were doing exactly this,’ and obviously in a really respectful way.”

It’s not uncommon for students at Diné College and UPRP to experience an initial hurdle that their MIT peers do not. Epstein notes, “There’s a tendency for some folks outside MIT to see MIT students as these brilliant people that they don’t belong in the same room with.” But the other students soon realize not only that they can hold their own intellectually, but also that their backgrounds and experiences are incredibly valuable. “Their life experiences actually put them way ahead of many MIT students in some ways, when you think about design and fabrication, like repairing farm equipment or rebuilding transmissions,” he adds.

That’s how Cauy Bia felt when he took the class in 2024. Currently a first-year graduate student in biology at Diné College, Bia questioned whether he’d be on par with the MIT students. “I’ve grown up on a farm, and we do a lot of building, a lot of calculations, a lot of hands-on stuff. But going into this, I was sweating it so hard [wondering], ‘Am I smart enough to work with these students?’ And then, at the end of the day, that was never an issue,” he says.

The value of reflection

Every two weeks, Terrascope students write personal reflections about their experiences in the class, which helps them appreciate their academic and personal development. “I really felt that I had undergone a process that made me grow as an engineer,” says Leon. “I understood the importance of people and engineering more, including teamwork, working with clients, and de-centering the project away from what I wanted to build and design.”

When Bia began the semester, he says, he was more of a “make-or-break-type person” and tended to see things in black and white. “But working with all three campuses, it kind of opened up my thought process so I can assess more ideas, more voices and opinions. And I can get broader perspectives and get bigger ideas from that point,” he says. It was also a powerful experience culturally for him, particularly “drawing parallels between Navajo history, Navajo culture, and seeing the similarities between that and Puerto Rican culture, seeing how close we are as two nations.”

Rodriguez-Andujar gained an appreciation for the “constant struggle between simplicity and complexity” in engineering. “You have all these engineers trying to over-engineer everything,” he says. “And after you get your client feedback [halfway through the semester], it turns out, ‘Oh, that doesn’t work for me. I’m sorry — you have to scale it down like a hundred times and make it a lot simpler.’”

For instructors, the students’ reflections are invaluable as they strive to make improvements every year. In many ways, you might say the class is an iterative design spiral, too. “The past three years have themselves been prototypes,” Epstein says, “and all of the instructional staff are looking forward to continuing these exciting partnerships.”

MIT researchers have developed a new bionic knee that can help people with above-the-knee amputations walk faster, climb stairs, and avoid obstacles more easily than they could with a traditional prosthesis.

Unlike prostheses in which the residual limb sits within a socket, the new system is directly integrated with the user’s muscle and bone tissue. This enables greater stability and gives the user much more control over the movement of the prosthesis.

Participants in a small clinical study also reported that the limb felt more like a part of their own body, compared to people who had more traditional above-the-knee amputations.

“A prosthesis that's tissue-integrated — anchored to the bone and directly controlled by the nervous system — is not merely a lifeless, separate device, but rather a system that is carefully integrated into human physiology, offering a greater level of prosthetic embodiment. It’s not simply a tool that the human employs, but rather an integral part of self,” says Hugh Herr, a professor of media arts and sciences, co-director of the K. Lisa Yang Center for Bionics at MIT, an associate member of MIT’s McGovern Institute for Brain Research, and the senior author of the new study.

Tony Shu PhD ’24 is the lead author of the paper, which appears today in Science.

Better control

Over the past several years, Herr’s lab has been working on new prostheses that can extract neural information from muscles left behind after an amputation and use that information to help guide a prosthetic limb.

During a traditional amputation, pairs of muscles that take turns stretching and contracting are usually severed, disrupting the normal agonist-antagonist relationship of the muscles. This disruption makes it very difficult for the nervous system to sense the position of a muscle and how fast it’s contracting.

Using the new surgical approach developed by Herr and his colleagues, known as agonist-antagonist myoneuronal interface (AMI), muscle pairs are reconnected during surgery so that they still dynamically communicate with each other within the residual limb. This sensory feedback helps the wearer of the prosthesis to decide how to move the limb, and also generates electrical signals that can be used to control the prosthetic limb.

In a 2024 study, the researchers showed that people with amputations below the knee who received the AMI surgery were able to walk faster and navigate around obstacles much more naturally than people with traditional below-the-knee amputations.

In the new study, the researchers extended the approach to better serve people with amputations above the knee. They wanted to create a system that could not only read out signals from the muscles using AMI but also be integrated into the bone, offering more stability and better sensory feedback.

To achieve that, the researchers developed a procedure to insert a titanium rod into the residual femur bone at the amputation site. This implant allows for better mechanical control and load bearing than a traditional prosthesis. Additionally, the implant contains 16 wires that collect information from electrodes located on the AMI muscles inside the body, which enables more accurate transduction of the signals coming from the muscles.

This bone-integrated system, known as e-OPRA, transmits AMI signals to a new robotic controller developed specifically for this study. The controller uses this information to calculate the torque necessary to move the prosthesis the way that the user wants it to move.

“All parts work together to better get information into and out of the body and better interface mechanically with the device,” Shu says. “We’re directly loading the skeleton, which is the part of the body that’s supposed to be loaded, as opposed to using sockets, which is uncomfortable and can lead to frequent skin infections.”

In this study, two subjects received the combined AMI and e-OPRA system, known as an osseointegrated mechanoneural prosthesis (OMP). These users were compared with eight who had the AMI surgery but not the e-OPRA implant, and seven users who had neither AMI nor e-OPRA. All subjects took a turn at using an experimental powered knee prosthesis developed by the lab.

The researchers measured the participants’ ability to perform several types of tasks, including bending the knee to a specified angle, climbing stairs, and stepping over obstacles. In most of these tasks, users with the OMP system performed better than the subjects who had the AMI surgery but not the e-OPRA implant, and much better than users of traditional prostheses.

“This paper represents the fulfillment of a vision that the scientific community has had for a long time — the implementation and demonstration of a fully physiologically integrated, volitionally controlled robotic leg,” says Michael Goldfarb, a professor of mechanical engineering and director of the Center for Intelligent Mechatronics at Vanderbilt University, who was not involved in the research. “This is really difficult work, and the authors deserve tremendous credit for their efforts in realizing such a challenging goal.”

A sense of embodiment

In addition to testing gait and other movements, the researchers also asked questions designed to evaluate participants’ sense of embodiment — that is, to what extent their prosthetic limb felt like a part of their own body.

Questions included whether the patients felt as if they had two legs, if they felt as if the prosthesis was part of their body, and if they felt in control of the prosthesis. Each question was designed to evaluate the participants’ feelings of agency, ownership of device, and body representation.

The researchers found that as the study went on, the two participants with the OMP showed much greater increases in their feelings of agency and ownership than the other subjects.

“Another reason this paper is significant is that it looks into these embodiment questions and it shows large improvements in that sensation of embodiment,” Herr says. “No matter how sophisticated you make the AI systems of a robotic prosthesis, it’s still going to feel like a tool to the user, like an external device. But with this tissue-integrated approach, when you ask the human user what is their body, the more it’s integrated, the more they’re going to say the prosthesis is actually part of self.”

The AMI procedure is now done routinely on patients with below-the-knee amputations at Brigham and Women’s Hospital, and Herr expects it will soon become the standard for above-the-knee amputations as well. The combined OMP system will need larger clinical trials to receive FDA approval for commercial use, which Herr expects may take about five years.

The research was funded by the Yang Tan Collective and DARPA.

As spaceflight becomes more affordable and accessible, the story of human life in space is just beginning. Aurelia Institute wants to make sure that future benefits all of humanity — whether in space or here on Earth.

Founded by Ariel Ekblaw SM ’17, PhD ’20; Danielle DeLatte ’11; and former MIT research scientist Sana Sharma, the nonprofit institute serves as a research lab for space technology and architecture, a center for education and outreach, and a policy hub dedicated to inspiring more people to work in the space industry.

At the heart of the Aurelia Institute’s mission is a commitment to making space accessible to all people. A big part of that work involves annual microgravity flights that Ekblaw says are equal part research missions, workforce training, and inspiration for the next generation of space enthusiasts.

“We’ve done that every year,” Ekblaw says of the flights. “We now have multiple cohorts of students that connect across years. It brings together people from very different backgrounds. We’ve had artists, designers, architects, ethicists, teachers, and others fly with us. In our R&D, we are interested in space infrastructure for the public good. That’s why we’re directing our technology portfolios toward near-term, massive infrastructure projects in low-Earth orbit that benefit life on Earth.”

From the annual flights to the Institute’s self-assembling space architecture technology known as TESSERAE, much of Aurelia’s work is an extension of projects Ekblaw started as a graduate student at MIT.

“My life trajectory changed when I came to MIT,” says Ekblaw, who is still a visiting researcher at MIT. “I am incredibly grateful for the education I got in the Media Lab and the Department of Aeronautics and Astronautics. MIT is what gave me the skill, the technology, and the community to be able to spin out Aurelia and do something important in the space industry at scale.”

“MIT changes lives”

Ekblaw has always been passionate about space. As an undergraduate at Yale University, she took part in a NASA microgravity flight as part of a research project. In the first year of her PhD program at MIT, she led the launch of the Space Exploration Initiative, a cross-Institute effort to drive innovation at the frontiers of space exploration. The ongoing initiative started as a research group but soon raised enough money to conduct microgravity flights and, more recently, conduct missions to the International Space Station and the moon.

“The Media Lab was like magic in the years I was there,” Ekblaw says. “It had this sense of what we used to call ‘anti-disciplinary permission-lessness.’ You could get funding to explore really different and provocative ideas. Our mission was to democratize access to space.”

In 2016, while taking a class taught by Neri Oxman, then a professor in the Media Lab, Ekblaw got the idea for the TESSERAE Project, in which tiles autonomously self-assemble into spherical space structures.

“I was thinking about the future of human flight, and the class was a seeding moment for me,” Ekblaw says. “I realized self-assembly works OK on Earth, it works particularly well at small scales like in biology, but it generally struggles with the force of gravity once you get to larger objects. But microgravity in space was a perfect application for self-assembly.”

That semester, Ekblaw was also taking Professor Neil Gershenfeld’s class MAS.863 (How to Make (Almost) Anything), where she began building prototypes. Over the ensuing years of her PhD, subsequent versions of the TESSERAE system were tested on microgravity flights run by the Space Exploration Initiative, in a suborbital mission with the space company Blue Origin, and as part of a 30-day mission aboard the International Space Station.

“MIT changes lives,” Ekblaw says. “It completely changed my life by giving me access to real spaceflight opportunities. The capstone data for my PhD was from an International Space Station mission.”

After earning her PhD in 2020, Ekblaw decided to ask two researchers from the MIT community and the Space Exploration Initiative, Danielle DeLatte and Sana Sharma, to partner with her to further develop research projects, along with conducting space education and policy efforts. That collaboration turned into Aurelia.

“I wanted to scale the work I was doing with the Space Exploration Initiative, where we bring in students, introduce them to zero-g flights, and then some graduate to sub-orbital, and eventually flights to the International Space Station,” Ekblaw says. “What would it look like to bring that out of MIT and bring that opportunity to other students and mid-career people from all walks of life?”

Every year, Aurelia charters a microgravity flight, bringing about 25 people along to conduct 10 to 15 experiments. To date, nearly 200 people have participated in the flights across the Space Exploration Initiative and Aurelia, and more than 70 percent of those fliers have continued to pursue activities in the space industry post-flight.

Aurelia also offers open-source classes on designing research projects for microgravity environments and contributes to several education and community-building activities across academia, industry, and the arts.

In addition to those education efforts, Aurelia has continued testing and improving the TESSERAE system. In 2022, TESSERAE was brought on the first private mission to the International Space Station, where astronauts conducted tests around the system’s autonomous self-assembly, disassembly, and stability. Aurelia will return to the International Space Station in early 2026 for further testing as part of a recent grant from NASA.

The work led Aurelia to recently spin off the TESSERAE project into a separate, for-profit company. Ekblaw expects there to be more spinoffs out of Aurelia in coming years.

Designing for space, and Earth

The self-assembly work is only one project in Aurelia’s portfolio. Others are focused on designing human-scale pavilions and other habitats, including a space garden and a massive, 20-foot dome depicting the interior of space architectures in the future. This space habitat pavilion was recently deployed as part of a six-month exhibit at the Seattle Museum of Flight.

“The architectural work is asking, ‘How are we going to outfit these systems and actually make the habitats part of a life worth living?’” Ekblaw explains.

With all of its work, Aurelia’s team looks at space as a testbed to bring new technologies and ideas back to our own planet.

“When you design something for the rigors of space, you often hit on really robust technologies for Earth,” she says.

Generative artificial intelligence is transforming the ways humans write, read, speak, think, empathize, and act within and across languages and cultures. In health care, gaps in communication between patients and practitioners can worsen patient outcomes and prevent improvements in practice and care. The Language/AI Incubator, made possible through funding from the MIT Human Insight Collaborative (MITHIC), offers a potential response to these challenges. 

The project envisions a research community rooted in the humanities that will foster interdisciplinary collaboration across MIT to deepen understanding of generative AI’s impact on cross-linguistic and cross-cultural communication. The project’s focus on health care and communication seeks to build bridges across socioeconomic, cultural, and linguistic strata.

The incubator is co-led by Leo Celi, a physician and the research director and senior research scientist with the Institute for Medical Engineering and Science (IMES), and Per Urlaub, professor of the practice in German and second language studies and director of MIT’s Global Languages program. 

“The basis of health care delivery is the knowledge of health and disease,” Celi says. “We’re seeing poor outcomes despite massive investments because our knowledge system is broken.”

A chance collaboration

Urlaub and Celi met during a MITHIC launch event. Conversations during the event reception revealed a shared interest in exploring improvements in medical communication and practice with AI.

“We’re trying to incorporate data science into health-care delivery,” Celi says. “We’ve been recruiting social scientists [at IMES] to help advance our work, because the science we create isn’t neutral.”

Language is a non-neutral mediator in health care delivery, the team believes, and can be a boon or barrier to effective treatment. “Later, after we met, I joined one of his working groups whose focus was metaphors for pain: the language we use to describe it and its measurement,” Urlaub continues. “One of the questions we considered was how effective communication can occur between doctors and patients.”

Technology, they argue, impacts casual communication, and its impact depends on both users and creators. As AI and large language models (LLMs) gain power and prominence, their use is broadening to include fields like health care and wellness. 

Rodrigo Gameiro, a physician and researcher with MIT’s Laboratory for Computational Physiology, is another program participant. He notes that work at the laboratory centers responsible AI development and implementation. Designing systems that leverage AI effectively, particularly when considering challenges related to communicating across linguistic and cultural divides that can occur in health care, demands a nuanced approach. 

“When we build AI systems that interact with human language, we’re not just teaching machines how to process words; we’re teaching them to navigate the complex web of meaning embedded in language,” Gameiro says.

Language’s complexities can impact treatment and patient care. “Pain can only be communicated through metaphor,” Urlaub continues, “but metaphors don’t always match, linguistically and culturally.” Smiley faces and one-to-10 scales — pain measurement tools English-speaking medical professionals may use to assess their patients — may not travel well across racial, ethnic, cultural, and language boundaries.

“Science has to have a heart” 

LLMs can potentially help scientists improve health care, although there are some systemic and pedagogical challenges to consider. Science can focus on outcomes to the exclusion of the people it’s meant to help, Celi argues. “Science has to have a heart,” he says. “Measuring students’ effectiveness by counting the number of papers they publish or patents they produce misses the point.”

The point, Urlaub says, is to investigate carefully while simultaneously acknowledging what we don’t know, citing what philosophers call Epistemic Humility. Knowledge, the investigators argue, is provisional, and always incomplete. Deeply held beliefs may require revision in light of new evidence. 

“No one’s mental view of the world is complete,” Celi says. “You need to create an environment in which people are comfortable acknowledging their biases.”

“How do we share concerns between language educators and others interested in AI?” Urlaub asks. “How do we identify and investigate the relationship between medical professionals and language educators interested in AI’s potential to aid in the elimination of gaps in communication between doctors and patients?” 

Language, in Gameiro’s estimation, is more than just a tool for communication. “It reflects culture, identity, and power dynamics,” he says. In situations where a patient might not be comfortable describing pain or discomfort because of the physician’s position as an authority, or because their culture demands yielding to those perceived as authority figures, misunderstandings can be dangerous. 

Changing the conversation

AI’s facility with language can help medical professionals navigate these areas more carefully, providing digital frameworks offering valuable cultural and linguistic contexts in which patient and practitioner can rely on data-driven, research-supported tools to improve dialogue. Institutions need to reconsider how they educate medical professionals and invite the communities they serve into the conversation, the team says. 

‘We need to ask ourselves what we truly want,” Celi says. “Why are we measuring what we’re measuring?” The biases we bring with us to these interactions — doctors, patients, their families, and their communities — remain barriers to improved care, Urlaub and Gameiro say.

“We want to connect people who think differently, and make AI work for everyone,” Gameiro continues. “Technology without purpose is just exclusion at scale.”

“Collaborations like these can allow for deep processing and better ideas,” Urlaub says.

Creating spaces where ideas about AI and health care can potentially become actions is a key element of the project. The Language/AI Incubator hosted its first colloquium at MIT in May, which was led by Mena Ramos, a physician and the co-founder and CEO of the Global Ultrasound Institute. 

The colloquium also featured presentations from Celi, as well as Alfred Spector, a visiting scholar in MIT’s Department of Electrical Engineering and Computer Science, and Douglas Jones, a senior staff member in the MIT Lincoln Laboratory’s Human Language Technology Group. A second Language/AI Incubator colloquium is planned for August.

Greater integration between the social and hard sciences can potentially increase the likelihood of developing viable solutions and reducing biases. Allowing for shifts in the ways patients and doctors view the relationship, while offering each shared ownership of the interaction, can help improve outcomes. Facilitating these conversations with AI may speed the integration of these perspectives. 

“Community advocates have a voice and should be included in these conversations,” Celi says. “AI and statistical modeling can’t collect all the data needed to treat all the people who need it.”

Community needs and improved educational opportunities and practices should be coupled with cross-disciplinary approaches to knowledge acquisition and transfer. The ways people see things are limited by their perceptions and other factors. “Whose language are we modeling?” Gameiro asks about building LLMs. “Which varieties of speech are being included or excluded?” Since meaning and intent can shift across those contexts, it’s important to remember these when designing AI tools. 

“AI is our chance to rewrite the rules”

While there’s lots of potential in the collaboration, there are serious challenges to overcome, including establishing and scaling the technological means to improve patient-provider communication with AI, extending opportunities for collaboration to marginalized and underserved communities, and reconsidering and revamping patient care. 

But the team isn’t daunted.

Celi believes there are opportunities to address the widening gap between people and practitioners while addressing gaps in health care. “Our intent is to reattach the string that’s been cut between society and science,” he says. “We can empower scientists and the public to investigate the world together while also acknowledging the limitations engendered in overcoming their biases.”

Gameiro is a passionate advocate for AI’s ability to change everything we know about medicine. “I’m a medical doctor, and I don’t think I’m being hyperbolic when I say I believe AI is our chance to rewrite the rules of what medicine can do and who we can reach,” he says.

“Education changes humans from objects to subjects,” Urlaub argues, describing the difference between disinterested observers and active and engaged participants in the new care model he hopes to build. “We need to better understand technology’s impact on the lines between these states of being.”

Celi, Gameiro, and Urlaub each advocate for MITHIC-like spaces across health care, places where innovation and collaboration are allowed to occur without the kinds of arbitrary benchmarks institutions have previously used to mark success.

“AI will transform all these sectors,” Urlaub believes. “MITHIC is a generous framework that allows us to embrace uncertainty with flexibility.”

“We want to employ our power to build community among disparate audiences while admitting we don’t have all the answers,” Celi says. “If we fail, it’s because we failed to dream big enough about how a reimagined world could look.”

Marine scientists have long marveled at how animals like fish and seals swim so efficiently despite having different shapes. Their bodies are optimized for efficient, hydrodynamic aquatic navigation so they can exert minimal energy when traveling long distances.

Autonomous vehicles can drift through the ocean in a similar way, collecting data about vast underwater environments. However, the shapes of these gliding machines are less diverse than what we find in marine life — go-to designs often resemble tubes or torpedoes, since they’re fairly hydrodynamic as well. Plus, testing new builds requires lots of real-world trial-and-error.

Researchers from MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) and the University of Wisconsin at Madison propose that AI could help us explore uncharted glider designs more conveniently. Their method uses machine learning to test different 3D designs in a physics simulator, then molds them into more hydrodynamic shapes. The resulting model can be fabricated via a 3D printer using significantly less energy than hand-made ones.

The MIT scientists say that this design pipeline could create new, more efficient machines that help oceanographers measure water temperature and salt levels, gather more detailed insights about currents, and monitor the impacts of climate change. The team demonstrated this potential by producing two gliders roughly the size of a boogie board: a two-winged machine resembling an airplane, and a unique, four-winged object resembling a flat fish with four fins.

Peter Yichen Chen, MIT CSAIL postdoc and co-lead researcher on the project, notes that these designs are just a few of the novel shapes his team’s approach can generate. “We’ve developed a semi-automated process that can help us test unconventional designs that would be very taxing for humans to design,” he says. “This level of shape diversity hasn’t been explored previously, so most of these designs haven’t been tested in the real world.”

But how did AI come up with these ideas in the first place? First, the researchers found 3D models of over 20 conventional sea exploration shapes, such as submarines, whales, manta rays, and sharks. Then, they enclosed these models in “deformation cages” that map out different articulation points that the researchers pulled around to create new shapes.

The CSAIL-led team built a dataset of conventional and deformed shapes before simulating how they would perform at different “angles-of-attack” — the direction a vessel will tilt as it glides through the water. For example, a swimmer may want to dive at a -30 degree angle to retrieve an item from a pool.

These diverse shapes and angles of attack were then used as inputs for a neural network that essentially anticipates how efficiently a glider shape will perform at particular angles and optimizes it as needed.

Giving gliding robots a lift

The team’s neural network simulates how a particular glider would react to underwater physics, aiming to capture how it moves forward and the force that drags against it. The goal: find the best lift-to-drag ratio, representing how much the glider is being held up compared to how much it’s being held back. The higher the ratio, the more efficiently the vehicle travels; the lower it is, the more the glider will slow down during its voyage.

Lift-to-drag ratios are key for flying planes: At takeoff, you want to maximize lift to ensure it can glide well against wind currents, and when landing, you need sufficient force to drag it to a full stop.

Niklas Hagemann, an MIT graduate student in architecture and CSAIL affiliate, notes that this ratio is just as useful if you want a similar gliding motion in the ocean.

“Our pipeline modifies glider shapes to find the best lift-to-drag ratio, optimizing its performance underwater,” says Hagemann, who is also a co-lead author on a paper that was presented at the International Conference on Robotics and Automation in June. “You can then export the top-performing designs so they can be 3D-printed.”

Going for a quick glide

While their AI pipeline seemed realistic, the researchers needed to ensure its predictions about glider performance were accurate by experimenting in more lifelike environments.

They first fabricated their two-wing design as a scaled-down vehicle resembling a paper airplane. This glider was taken to MIT’s Wright Brothers Wind Tunnel, an indoor space with fans that simulate wind flow. Placed at different angles, the glider’s predicted lift-to-drag ratio was only about 5 percent higher on average than the ones recorded in the wind experiments — a small difference between simulation and reality.

A digital evaluation involving a visual, more complex physics simulator also supported the notion that the AI pipeline made fairly accurate predictions about how the gliders would move. It visualized how these machines would descend in 3D.

To truly evaluate these gliders in the real world, though, the team needed to see how their devices would fare underwater. They printed two designs that performed the best at specific points-of-attack for this test: a jet-like device at 9 degrees and the four-wing vehicle at 30 degrees.

Both shapes were fabricated in a 3D printer as hollow shells with small holes that flood when fully submerged. This lightweight design makes the vehicle easier to handle outside of the water and requires less material to be fabricated. The researchers placed a tube-like device inside these shell coverings, which housed a range of hardware, including a pump to change the glider’s buoyancy, a mass shifter (a device that controls the machine’s angle-of-attack), and electronic components.

Each design outperformed a handmade torpedo-shaped glider by moving more efficiently across a pool. With higher lift-to-drag ratios than their counterpart, both AI-driven machines exerted less energy, similar to the effortless ways marine animals navigate the oceans.

As much as the project is an encouraging step forward for glider design, the researchers are looking to narrow the gap between simulation and real-world performance. They are also hoping to develop machines that can react to sudden changes in currents, making the gliders more adaptable to seas and oceans.

Chen adds that the team is looking to explore new types of shapes, particularly thinner glider designs. They intend to make their framework faster, perhaps bolstering it with new features that enable more customization, maneuverability, or even the creation of miniature vehicles.

Chen and Hagemann co-led research on this project with OpenAI researcher Pingchuan Ma SM ’23, PhD ’25. They authored the paper with Wei Wang, a University of Wisconsin at Madison assistant professor and recent CSAIL postdoc; John Romanishin ’12, SM ’18, PhD ’23; and two MIT professors and CSAIL members: lab director Daniela Rus and senior author Wojciech Matusik. Their work was supported, in part, by a Defense Advanced Research Projects Agency (DARPA) grant and the MIT-GIST Program.