Tissue engineers are finding ways to grow living organs and tissues from cells, with the aim of replacing diseased and damaged counterparts in the body. Scientists have successfully grown artificial muscles, livers, kidneys, skin, and other tissues. But there’s been no reliable way to engineer precisely patterned networks of blood vessels, some of which can be finer than a human hair. 

Without a vascular network to deliver nutrients, any artificial tissues, no matter how life-like, can’t function. 

Now MIT engineers have found they can engineer and control the growth of blood vessels by mechanically stretching them. 

The team has built a human “blood vessel on a chip,” composed of a central artery made from human endothelial cells, that is embedded in a gel that also contains a small magnet. The researchers studied how the main artery responded as they jostled the gel back and forth using an external magnet to move the magnet embedded within the gel.

They found that the simple mechanical action of repeatedly jostling the artery stimulated the artery to sprout other, smaller capillaries. By changing the direction in which the artery is jostled or stretched, the researchers could redirect the growing new vessels. And stretching the artery by various degrees influenced how many more new vessels sprouted. 

Their results, reported in the Proceedings of the National Academy of Sciences, offer scientists a new way to engineer artificial blood vessels and program the patterns in which they grow. 

“Healthy tissues depend on organized blood vessel networks, but state-of-the-art protocols don't enable fabricating such networks within engineered tissues,” says Ritu Raman, associate professor of mechanical engineering at MIT and the study’s co-lead author. “The ability to program blood vessel growth with physical cues may enable reproducible and scalable fabrication of engineered tissues that can be implanted in the body to restore function after debilitating disease or injury.”

The study’s MIT co-authors include Sina Kheiri, Jessica Shah, Shashaank Venkatesh, and Roger Kamm, along with Peiyuan Chai and Ryan Flynn at Harvard University. 

“Moving is good”

Blood vessels are tricky to grow and control using conventional fabrication techniques. While 3D printers can produce vessels at the scale of major arteries and veins, the technology is not precise enough to print intricate networks of much finer, thread-like capillaries. Scientists have had some success with growing blood vessels from individual cells, by cultivating them in Petri dishes filled with nutrients and growth factors. But controlling how and where they grow remains a challenge. 

“You can try to pattern chemical cues, like growth factors, to direct where vessels grow, but you can’t do this very precisely,” Raman says. “We thus need other types of patternable cues that can help us build tissues with organized vessels.”

She and her students wondered whether they could grow and control new blood vessels using a protocol they previously developed to grow artificial muscles and nerves. In their previous works, the team engineered a small chip filled with a gel that they infused with nutrients and growth factors. They embedded a small magnet within the gel, and then carpeted the surface of the gel with live muscle or neuron cells. They then manipulated an external magnet to pull the embedded magnet, and the cell-covered gel, back and forth. This work revealed that mechanical “exercise,” pulling the cells back and forth, directly influenced how the cells grew. 

In their new work, the team used a similar setup to see if they could grow and control new blood vessels. 

The researchers built a “blood-vessel-on-a-chip,” smaller than a postage stamp, and filled it with a similar nutrient-rich gel containing a small magnet. They poked a thin tube lengthwise through the gel to create a hollow channel, and coated the channel with live endothelial cells, which naturally grow and fuse to form blood vessels in the body. Once the cells took on the channel’s shape, they started sprouting new, capillary-like vessels in the gel. 

Placing the device under a motorized stage fitted with small, suspended magnets, the researchers moved the magnets back and forth in different directions, and by various degrees, and observed whether and how blood vessels sprouted from the central artery in response. 

“The main takeaway is: Stretching the blood vessel back and forth seems to enhance the number of new capillaries that grow,” Raman says. 

If the main artery were simply left alone in the gel, it would grow some new vessels in random locations along its length. But when the artery was jostled, significantly more vessels sprouted. When the team used the magnets to stretch the gel back and forth, by 5 percent of the gel’s total width, many new vessels grew out from the main artery. When they stretched by 15 percent, fewer vessels sprouted, but those that did grew longer. And when the team changed the direction of stretching, the new vessels followed in response, taking turns and following the pattern of the team’s mechanical stimulation.

“We’re finding that moving is good, which is always the takeaway of everything we do in our lab,” Raman says. “Mechanical forces play an important role in our bodies. That means that if you want to grow more or less vessels, or shorter or longer vessels, or vessels in certain directions, we now know how to do that.”

A gatekeeping gene

The researchers went a step further to investigate why blood vessels grow in response to mechanical forces. To do so, they looked to gene editing, and the role of one particular gene: Piezo1. 

Raman had recently attended a talk by molecular biologist Ardem Patapoutian. In 2021, Patapoutian received the Nobel Prize in Physiology or Medicine for his discovery of ion channels in cell membranes that open and close in response to mechanical pressure. These channels, named PIEZO1 and PIEZO2, act as a cell’s gatekeepers, controlling what goes in and what comes out of a cell. Both types of channels, Patapoutian found, are regulated by their respective genes, also named PIEZO1 and PIEZO2. 

After his talk, Raman showed Patapoutian her group’s experimental results, which showed a connection between blood vessel growth and mechanical stimulation. Patapoutian in turn proposed that the explanation could be the PIEZO1 channel; by mechanically exercising the central artery, Raman may have been stimulating ion channels in the artery’s cells to open, triggering new blood vessels to grow. 

To test this hypothesis, Raman looked to knock down the PIEZO1 gene. If this gene were less active, and fewer blood vessels grew as a result, then it would mean that blood vessels do indeed grow in response to mechanical stimulation, and specifically, through the activation of PIEZO1 ion channels. 

The team repeated their experiments, this time with endothelial cells that were genetically edited to suppress the PIEZO1 gene. Sure enough, they observed that significantly fewer new blood vessels sprouted, even as they mechanically exercised the central artery.

Now that the team has found a way to grow and control blood vessel growth, they plan to apply the protocol to grow organized networks of vessels to supply artificial organs and tissues. “We are now investigating how precisely patterning blood vessel growth can help improve muscle function,” says co-author Jessica Shah.

This work was supported, in part, by the U.S. Department of War Army Research Office Early Career Program and PECASE Grant, and a Department of War DURIP Program Grant.

Professor Arthur Bahr has been named head of the MIT Literature Section, effective July 1. 

“Arthur is an exceptional scholar and a proven leader. I am confident that he will guide the unit with judgment, insight, and a deep commitment to its continued success,” says Agustín Rayo, the Kenan Sahin Dean of the School of Humanities, Arts, and Social Sciences. “I very much look forward to having him join the school’s leadership team.”

Bahr’s work blends formalist and materialist approaches to find literary resonance in the physical particularities of medieval manuscripts. He joined the MIT faculty in 2007 and helped lead the Ancient and Medieval Studies program in 2009-18 and 2022-23, working with colleagues from across the Institute to strengthen and expand the program. He has also been curriculum chair and undergraduate officer of the Literature Section.

“Lit@MIT has some of the world’s most innovative literary scholars and some of the Institute’s most dedicated teachers,” Bahr says. “It has also been my home for nearly 20 years, and I feel both humbled and energized by the opportunity to help shape its future. 

“Literature creates opportunities to slow down and reflect on what really matters, and in a fast-paced, increasingly automated world, those skills are more vital than ever,” he continues.

Bahr succeeds Associate Professor Sandy Alexandre, who served as head of the unit since July 2025.

Bahr is the author of “Chasing the Pearl-Manuscript: Speculation, Shapes, Delight” (University of Chicago Press, 2025); “Fragments and Assemblages: Forming Compilations of Medieval London” (University of Chicago Press, 2013); and co-editor of “Medieval English Manuscripts: Form, Aesthetics, and the Literary Text,” a special volume of The Chaucer Review (47.4, April 2013). His essays have appeared in ELH, Studies in the Age of Chaucer, Studies in Philology, and The Chaucer Review, among others. 

Bahr has been named a SHASS Faculty Fellow for the spring 2027 semester. His next project combines his interest in manuscripts with his training as a figure skating judge to explore analogies between sheets of parchment and sheets of ice, as sites of performance, inscription, and erasure.  

Bahr was named a MacVicar Faculty Fellow in 2015. He received the James A. (’48) and Ruth Levitan Award for Excellence in Teaching in 2012.

Bahr has served MIT as chair of the Committee on the Undergraduate Program from 2019 to 2021, and served on the pandemic-era Academic Policies and Regulations Team. He was also a subcommittee chair of the Education Group of Task Force 2021 and Beyond, and member of the subsequent Refinement and Implementation Committee on the Undergraduate Program. 

Bahr earned his undergraduate degree from Amherst College and his PhD in English Language and Literature from the University of California at Berkeley.

In May 2019, the government of Baltimore, Maryland, fell into chaos. Cybercriminals had locked the city out of many of its critical files and demanded payment to decrypt them. The city refused to pay ransom. The attack incapacitated a swath of services, including real estate transactions and bill payment, and recovery costs soared into the millions.

The syllabus of class 11.074/11.274 (Cybersecurity Clinic), a course in the MIT Department of Urban Studies and Planning (DUSP), includes a case study on Baltimore’s situation as an example of increasingly common ransomware attacks on municipal governments and other public agencies. To counter such threats, Lecturer Jungwoo Chun and Ford Professor of Urban and Environmental Planning Lawrence Susskind launched the MIT Cybersecurity Clinic in 2019. They have offered the course nearly every semester since.

Much like a legal or medical clinic, the course doubles as hands-on training for students and a pro-bono service to at-risk communities. After completing instructional modules and passing a certification exam, students are assigned in teams to a client. By the end of the semester, each team creates a report assessing the client’s vulnerabilities to cyberattack and recommending steps to improve protection. So far, the clinic has provided more than 40 assessments, confidential and free of charge, primarily for New England municipalities and health-care organizations.

In 2025, the FBI’s Internet Crime Complaint Center documented an average of 2,765 cyberattacks targeting Americans every day. When these attacks strike cities and towns, the fallout goes beyond finances, says Chun: “There’s a terrifying, cascading effect on every dimension of our lives.” 

In recent years, cyberattacks targeting the kinds of client communities served by MIT’s clinic have imperiled water supplies, impeded 911 and police services, and exposed citizens’ personal data.

Despite being gateways to essential infrastructure, many small municipalities and hospitals lack in-house staff trained in cybersecurity. Demand for such experts far exceeds supply in today’s labor market, and public sector budgets rarely can match the high salaries private companies offer qualified candidates.

According to Comparitech, from 2018 to 2024, there have been 525 ransomware attacks on U.S. government entities, approximately one every five days, leading to an estimated $1.09 billion in downtime costs.  

“Underfunded public and not-for-profit bodies need to follow a self-help pathway,” Susskind says. “There are many low-cost moves that these organizations can implement with a little coaching from a free-service clinic.”

Defensive social engineering

Some might be surprised to find a university cybersecurity program housed outside the computer science department. Chun is an applied social scientist with expertise in public policy and planning, and Susskind is a leading scholar of conflict resolution and consensus building. They call the approach they’ve developed for the clinic “defensive social engineering” to emphasize that cybersecurity isn’t solely a technical challenge.

Chun acknowledges that the rapid development of artificial intelligence has created alarming new tools for criminals — “now AI can not only identify the vulnerability, but do the attack itself, which is really scary” — and an ever-evolving menu of software claims to guard against these attacks. Accordingly, the course spends considerable time on the technical aspects of cybersecurity. “But at the end of the day,” Chun says, “the biggest attack vector is still through humans.”

The term “social engineering” commonly refers to ways cybercrime victims are manipulated into compromising security (for example, by sending money to a scammer, downloading malicious code, or disclosing sensitive information). Susskind and Chun’s concept of defensive social engineering is similarly grounded in human psychology. The approach emphasizes that cybersecurity must be part of everyone’s job, technical or otherwise.

“It’s about people knowing what to do, people making the right choices,” says Chun. “It’s helping them use the resources and budget they have now on things that can be long-lasting, rather than just spending on the latest antivirus software.”

“Students with computer science backgrounds are surprised by the importance we attach to helping clients build organizational capacity,” says Susskind. “Students need to understand the leadership dynamics in their client communities. The IT director can’t just do what she or he wants. They depend on the local government for their budget. They need approval to hire new staff.”

On the other hand, Susskind says, students from planning or social science backgrounds often study smart city innovations without learning much about the technologies needed to manage the associated risks. And there are aspects of AI and advanced system design — along with cyber law and other topics critical to cybersecurity — that engineering students may not learn in their other courses. The Cybersecurity Clinic aims to round out the knowledge of students from every discipline. The course aims to broaden those students’ knowledge, too, by inviting at least half a dozen guest speakers each semester from industry, other universities and MIT academic departments, industry, and/or relevant public agencies.

This past spring, for example, the lineup of lecturers included Dan Ricci, the founder of Industrial Data Works, on the modeling of risk in energy systems within budget-constrained environments; Gus Serino, president of I&C Secure Inc., on operational-technology cybersecurity for industrial control systems; and representatives from the MassCyberCenter and the Cybersecurity Infrastructure Security Agency providing overviews of their respective state- and federal-level organizations’ programs and initiatives.

“There are highly specialized things to learn, especially about the ways AI is changing cybersecurity, that we need help teaching,” Susskind says. “The rate at which the field of cybersecurity is changing means that most academics will have a very hard time keeping up.”

A roadmap for improvement

Clinic students spend the first four weeks of the semester preparing for field assignments. A series of online modules, supplemented by class discussion, outline the scope and nature of cyberattacks against critical urban infrastructure; review the 23 risk areas most relevant to their type of clients; and provide guidance for each step of the assessment process. This includes simulations of tricky client interactions. What if clients don’t take students seriously, or fail to provide the necessary information? What if they argue to receive a more positive assessment than the facts warrant?

“I’ve never ever had a class that prepared us for such realistic scenarios before,” says Diego Contreras, a rising senior majoring in computer science and engineering who completed the course this spring.

The modules culminate in an exam students must pass on their first try to receive a field assignment. For the remainder of the semester, they’ll receive continued support via weekly class meetings and get faculty input on their drafted reports, but the onus is on students to coordinate their team’s activities and build client trust.

“You represent MIT, and that is quite the responsibility,” Contreras says. “This course has given me people skills I wouldn’t have developed in any other context.”

“The most delicate aspect of the project was balancing our assessment findings,” says Zev Moore ’26, who took the class last fall as a senior studying mathematical economics and finance. “Our approach was to provide important feedback while simultaneously validating the positive security measures our client already had in place, which ensured our report felt like a collaborative roadmap for improvement.”

Certain key recommendations show up in the majority of reports. For example, clients are advised to inventory all hardware and software tied into their network and track who has access; patch software and back up data regularly; require multi-factor authentication and frequent password updates; train employees not to open attachments from unknown parties; prepare an attack response plan that clarifies lines of authority and includes the organization’s stance on paying ransoms; and only use vendors with good cybersecurity hygiene.

“None of these items is costly,” Susskind says. “Together, they will probably avoid 80 percent or more of the possible cost and danger of cyberattacks.”

Spreading the model

To date, more than 120 students have completed the full course at MIT. The online modules that prepare students for certification are freely available to the public as a massive open online course on MITx called Cybersecurity for Critical Urban Infrastructure, which has attracted tens of thousands of learners. The modules are also used by universities with their own cybersecurity clinics — a growing cohort, thanks in part to a consortium (with 61 member institutions and counting) co-founded by MIT in 2021 with the University of California at Berkeley, Indiana University, and the University of Alabama.

Most student teams wrap up client work after finalizing their recommendations; a few have volunteered to stay on after semester’s end to advise on implementation. In either case, Susskind and Chun check in periodically with clients for at least two years following each engagement.

“We often hear of the vulnerability assessment report serving as the organization's blueprint for their short-term, mid-term, and long-term agenda to be more prepared for future attacks,” says Chun. “We primarily work with IT directors or chief technology officers, and many of them have been telling us post-engagement that they shared the MIT report with the city or town leadership and were able to convince them they needed extra budget or a specific line item. They were using the student report as leverage to say, ‘it’s not just me saying it. We have a credible team who dedicated their time and these are the findings.’

“It's really a humbling experience,” Chun adds, “when some of our past clients reach out to us again after some time to say: ‘Now we have different people, we just purchased new equipment. Can we do this all over again?’”

Robots walking down the street, surrounded by astounded onlookers, is an increasingly common sight. But these machines aren’t yet the do-it-all assistants you’d want working in a kitchen or factory, and a major bottleneck is data. Much like humans, robots learn best by experience. The challenge is that it’s labor-intensive and time-consuming to physically teach these machines so many actions across different settings. 

“One natural idea is to use simulation as a training ground. While there has been significant progress over the last few years in the physics engines that power robotics simulators, one of the remaining challenges has been creating sufficiently rich and diverse simulation content to capture the complexity of the real world,” says Russ Tedrake, the Toyota Professor of Electrical Engineering and Computer Science (EECS), Aeronautics and Astronautics, and Mechanical Engineering at MIT, and a principal investigator at MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL).

It turns out that AI agents, or semi-autonomous programs that “think” and complete well-defined tasks, could help produce the lifelike virtual settings that robots need. The new “SceneSmith” system developed by researchers at MIT CSAIL and Toyota Research Institute uses three agents to piece together the objects, walls, and overall look of a 3D scene. Its recreations of indoor spaces such as restaurants, bedrooms, and hotels are more realistic and detailed than prior systems, helping robots practice skills and try out different ways of doing tasks before they’re powered on. In turn, engineers save time on real-world testing.

The agents have a sense of how everyday places are supposed to look because they each call on a multi-modal system called a vision-language model (VLM), specifically the state-of-the-art VLM GPT-5.2. It’s trained on lots of text and images from the internet to handle more visual prompts. This advanced model gives each agent a sort of spatial knowledge: First, a “designer” agent generates the elements of a scene, then a “critic” advises whether it looks realistic, and finally, an “orchestrator” manages their back-and-forth, deciding when the design is done. Once the three VLMs wrap up their creative collaboration, the scene is ready to load directly into physics simulation software.

“We’ve found that the system can construct 3D scenes the way a human designer would,” says MIT EECS PhD student Nicholas Pfaff, a CSAIL researcher and a lead author on a paper with Tedrake presenting the work. “We made over 1,300 scenes using a leading VLM that has internet-scale priors, and it made insanely creative and diverse arrangements. I hadn’t taught the system to do that in the prompts; it just improvised.”

Talk to my agent

Thanks to VLM agents, you can ask SceneSmith to do things like “generate a garage with a car, a workbench, tires stacked in the corner, and a ladder against the wall,” and get a virtual playground rich with objects a robot can tinker with. These rooms are decorated with up to six times more items per scene than prior methods, making them great for helping robots learn skills such as putting a cup in the sink, placing fruit on plates, and moving a soda can from a shelf to a table.

With so many rich virtual environments handy, you can evaluate whether your robot is ready for deployment without so much trial and error in the physical world. The researchers tested out different action plans (also called “policies”) in SceneSmith’s digital worlds, generating 100 unique spaces in the process. A VLM agent evaluated each attempt, and it found the robot’s plans were faulty, with the machine often failing at its chores. Humans agreed with the model’s verdicts over 99 percent of the time, which could help roboticists weed out flawed approaches in simulation before a robot moves in the real world.

But how realistic are these virtual worlds, really? It can be difficult to prove outright, so the researchers approached the question from several angles. The most telling test: they dropped a pretrained robot policy — an AI controller trained largely on real-world data, which had never seen a SceneSmith scene — into the generated environments. In one test, users told the system to “take the apple from the bowl and place it onto the cutting board,” and the simulated robot did exactly that. If the scenes didn’t closely resemble the real settings the policy had learned from, it simply wouldn’t have worked. 

The team also teleoperated robots through the virtual spaces, guiding them to open cabinets, put away bottles, and navigate between rooms. Their experiments revealed that the environments hold up under sustained physical interaction, expanding beyond visual inspection.

Behind the scenes

The agents that SceneSmith uses each have a well-defined role in the generative process, fleshing out scenes in stages. They essentially create a floor plan and bring it to life. 

Let’s say you wanted to create a scene similar to the first floor of a house. The “designer” VLM would start with a general layout, which the “critic” reviews, and then the “orchestrator” signs off. The agents repeat this approach for each step: adding furniture, placing objects on walls and then ceilings, and finally, dropping in objects that robots can manipulate. For example, the VLMs can add cabinets that the robots can open and close — an articulated item, which prior baselines didn’t often have.

At each stage, the second VLM ensures the scene is practical, advising that a bathtub is removed from a living room, for example. The third VLM ensures a high-quality scene is generated, even taking the design process a few turns back if the visuals aren’t up to par. Once the three VLMs wrap up their creative collaboration, the mechanics of the physical world are added via simulation software.

With a sound understanding of how rooms should look, where objects should be placed, and real-world physics, SceneSmith has a noticeable edge over prior methods. Compared to scene-generation baselines such as “HSM” and “Holodeck,” SceneSmith made environments with more objects, including a private office, a pottery store, and even a Minecraft-themed gaming room.

SceneSmith was also a favorite among over 200 users. They found the system’s visuals to be more realistic over 90 percent of the time. They also observed that, generally speaking, it followed prompts more closely than other approaches did. In other words, it was the best at generating the virtual playgrounds users actually wanted to see.

A system of many talents

Realism, diversity, and richness are all strong suits for SceneSmith, even when it comes to generating individual 3D objects. You can prompt it to create a rolling serving cart, and it’ll make a 2D image that it then turns into a detailed model with physical properties like mass, friction, and inertia.

Such a detailed process does come with a speed trade-off, though. It can take multiple hours to produce a single scene because the agents are creating and closely scrutinizing each object. With more computing power, the system could see dramatic increases in efficiency. CSAIL engineers are also hoping to expand to deformable objects (like sponges), should extensive 3D libraries become available.

“SceneSmith represents a significant advance in this regard by providing an agentic framework for generating simulation-ready indoor environments just from a simple text prompt,” says Jeremy Binagia, an applied scientist at Amazon Robotics who wasn’t involved in the research. “It advances the state of the art in several ways, including pushing the limits of the density of objects in the simulated environment, ensuring that all of the objects are physically accurate (versus just being visually realistic), and creating assets that are not constrained to a fixed library, since they can be generated via text-to-3D.”

Pfaff and Tedrake wrote the paper with Thomas Cohn SM ’24, an MIT PhD student and CSAIL researcher; and Toyota Research Institute roboticists Sergey Zakharov and Rick Cory SM ’08, PhD ’10. Their work was supported, in part, by Amazon, the U.S. Office of Naval Research, the Toyota Research Institute, and the U.S. National Science Foundation.

The team presented their findings as a spotlight at last week’s International Conference on Machine Learning. 

With the exploding popularity of generative artificial intelligence, many open-source models are now available online for anyone to adapt for their task, such as generating product renderings in a certain artistic style. 

But these models also find their way into the hands of nefarious actors who may optimize them to produce illegal content, like hate speech or child sexual abuse material (CSAM). This is a growing problem — the National Center for Missing and Exploited Children received more than 1.5 million reports of AI-generated CSAM in 2025, an increase from 67,000 in 2024.

Engineers usually test AI for harmful capabilities by prompting the model and inspecting its outputs, but this is impossible for CSAM, since it is illegal in the U.S to generate such content, regardless of intent.

To avoid this dilemma and improve AI safety, a team of MIT scientists, led by graduate student Vinith Suriyakumar and associate professors Ashia Wilson and Marzyeh Ghassemi, joined forces with researchers from Thorn to develop a new auditing approach that determines whether a model can produce CSAM, without prompting it. Thorn is a child safety nonprofit whose mission is to transform how children are protected from sexual abuse and exploitation in the digital age.

Their technique examines how the inner workings of a model have been adapted, but it never generates an output. By examining hidden representations, it can reliably infer whether a model has been specialized to produce harmful imagery.

When tested, the auditing procedure identified model variations that had been specialized to generate CSAM with 100 percent accuracy. A hosting platform could use this technique to flag unsafe models and quickly remove them or prevent them from being uploaded in the first place.

“This unlocks a new avenue for platforms that host open-source models and for law enforcement to actually test whether a model is capable of generating CSAM. Before, we had no way of measuring this. It was a huge blind spot that some people were taking advantage of. Now, we can address an AI safety problem that is having severe negative impacts,” says Vinith Suriyakumar, an MIT electrical engineering and computer science (EECS) graduate student and lead author of a paper on this technique.

Suriyakamur and Wilson, the Lister Borthers Career Develop Professor in EECS and a principal investigator in the Laboratory for Information and Decision Systems (LIDS), are joined on the paper by Lena Stempfle, an MIT postdoc; Ghassemi, an associate professor in EECS and a member of the Institute of Medical Engineering Sciences (IMES) and LIDS; and others at Boston University and Thorn. The paper was be presented as a spotlight at the “Trustworthy AI for Good” workshop at the International Conference on Machine Learning.

Auditing adaptations

Recent techniques have made it easier for users to specialize a generative AI model for their task through a process known as fine-tuning. 

Rather than retraining the entire model on a task-specific dataset, individuals can utilize an algorithm called low-rank adaptation (LoRA) to specialize the model in a more efficient manner.

This has led to a wave of new generative AI model variants for a variety of purposes, like producing watercolor images that mimic an artistic movement. But it has also enabled malicious actors to create models that can generate high-quality CSAM and other harmful imagery.

To audit a model, engineers typically prompt it for harmful content and check its outputs, but this manual auditing procedure is not scalable. In addition, repeatedly generating heinous images can have negative psychological impacts on human evaluators. 

This evaluation method quickly falls apart when testing CSAM, which is illegal to generate for any purpose in the U.S. and many other international jurisdictions.

“We are in this very difficult situation where, based on the law itself, we cannot use the de facto means of evaluation. We had to throw out the entire toolkit and take a different approach,” Suriyakumar says.

After learning about this conundrum, the researchers joined forces with Thorn, to address this issue.

A nongenerative solution

Instead of focusing on outputs, the researchers targeted the modifications a LoRA algorithm makes during fine-tuning. 

Their technique probes these modifications, called LoRA adaptors, to determine whether a model has been specialized for a harmful capability, without generating an output.

Using a technique called Gaussian probing, the researchers feed the model a set of random data points and analyze how it manipulates those data within its multilayer internal structure. 

“We never run the model all the way to the end or prompt the model, so we never generate images,” Suriyakumar explains.

The researchers capture those modifications at multiple time points within the model’s inner structure and average them to summarize how the LoRA adaptor changed the model’s computation. They found these responses to be a strong signal of how a model had been specialized.

They tested their method on variations of three types of models, comparing the results to ground-truth data from LoRA adaptors known for generating CSAM, other harmful images, and safe content. 

Their method was 100 percent accurate in identifying models that had been adapted to generate CSAM. 

“There is a huge bucket of child safety concerns with AI, and these are real concerns that need to be addressed. A lot of children are being harmed by AI deepfakes. We’ve shown that Gaussian probing can be a very useful tool, and we hope the research community really pours more attention into this problem,” Wilson says.

Importantly, their technique is scalable and would be relatively inexpensive to implement. Since thousands of model variations are published online every month, scalability is key to help auditors remove harmful adaptations before they are widely distributed.

Gaussian probing is also more robust than some other auditing techniques, since a nefarious actor would need to carefully alter the inner workings of the base model to avoid detection.

In the future, the researchers want to evaluate their technique on a larger set of model variations and explore whether Gaussian probing can detect harmful capabilities in base models before they are adapted.

“Now we have a technological approach to partially address this concern. So much effort was poured into this collaboration, which enabled us to tackle a really hard problem that is harming so many children, nationally and around the world. Hopefully, we can have a transformative impact in this area,” Ghassemi says.

This work was supported, in part, by the Bridgewater AIA Labs Research Fellowship.

Infrared cameras can be used to spot useful information that our eyes can’t see, such as gases escaping from a pipeline, chemicals in the atmosphere, or heat leaking from a building. But sensing infrared light in sophisticated ways still requires expensive and bulky systems.

Now MIT researchers have created a chip-based optical device that can dynamically control incoming infrared light, to act as a tunable lens that gathers additional information for infrared cameras. Each microscopic pixel of the device’s lens can control infrared light independently, allowing it to change its focus and help cameras detect different signals without moving parts.

The system is described in a paper published in Nature Communications. The researchers also explain how they built a lab-scale demonstration using mostly conventional manufacturing processes in a semiconductor chip factory, suggesting the approach could be implemented at industrial scales.

The technology could enable compact, tunable infrared cameras for more dynamic thermal imaging, chemical sensing, pollution monitoring, and even new kinds of optical computing.

“This could give us more information as we study space, or help with environmental protections where you want to monitor for specific compounds in the atmosphere,” explains first author Cosmin-Constantin Popescu PhD ’25. “Thermal imaging is another application, and you can think of military applications where night vision goggles are currently being used. Basically, a lot of organic molecules absorb in the mid-infrared wavelength, and you could use this system to detect them.”

Joining Popescu on the paper are MIT PhD students Maarten Robbert Anton Peters and Khoi Phuong Dao; Dynasil company scientists Oleg Maksimov and Harish Bhandari; University of Central Florida PhD candidate Kathleen Richardson and scientist Rashi Sharma; University of Washington Professor Arka Majumdar; Korea Advanced Institute of Science and Technology Associate Professor Hyun Jung Kim; MIT postdoc Rui Chen; Luigi Ranno PhD ’25; Brian Mills ’20, PhD ’26; Draper Laboratory scientist Dennis Calahan; MIT principal investigator Tian Gu; and Juejun Hu, MIT’s John F. Elliott Professor of Materials Science and Engineering.

Chip-based lenses

In recent years, researchers have developed ways to dynamically control light by etching tiny, precise patterns on transparent materials known as “metasurfaces,” which could enable more compact, programmable cameras and other advanced optical devices.

Hu’s research group at MIT has experimented with a class of metasurfaces that shift from solid to liquid after heat is applied. The phase changes can be harnessed to control how the materials interact with light. In 2021, Hu and collaborators created a miniature lens that could adjust its focus to different depths through such phase changes.

The device worked reliably, but it could only adjust focus all at once across the entire material, which is how most metasurfaces work. For their new study the researchers wanted to build on that approach to control light independently at each microscopic pixel of the material.

“Most active metasurfaces trying to do single-pixel tuning need wires going to every pixel, and how you route the wires becomes a big issue,” Hu explains. “The best approach so far has been one-dimensional pixel control with a bunch of wires.”

The researchers also wanted to create a system that worked with the mid-infrared wavelength of light, which the human eye can’t see but is useful for detecting heat signatures and molecules including methane and propane. Mid-infrared detection devices are already used to detect gas leaks and study Earth’s atmosphere, and for a number of defense and aerospace applications.

To build their system, the researchers adapted an approach commonly used in displays in which two layers of neatly packed copper wires are placed on top of each other perpendicularly. Below the wires is a layer of doped silicon that generates heat at the cross points of the wires and sits on top of the phase-change material. The silicon’s heat is used to switch each pixel of the material between crystalline and amorphous structures, which changes how the material interacts with the infrared light coming in. The silicon also includes a diode selector, which helps prevent unintended currents from leaking through neighboring pixels.

“We did some calculations showing this architecture allows us to scale to potentially millions of pixels without having any issues with the [unintended] currents,” Hu says. “The key innovation is this crossbar architecture, which creates a scalable way to increase the pixel-level switching of metasurfaces. We didn’t invent this architecture — it’s used in displays — but it’s the first time anyone’s used it for active phase-change metasurfaces to show you can get pixel-level control. People have been working toward two-dimensional pixel-level control for a long time, and it’s the first time anyone’s implemented it.”

The researchers worked with equipment in MIT.nano and with a factory that manufactures semiconductor chips, ultimately creating a two-dimensional system that featured a 6-by-6 metasurface pixel array. They tested their system and found it could switch on and off reliably.

“We found this mesh architecture to be very resilient,” Popescu says. “You don’t want these materials to switch once and not work anymore. You want it to switch a large number of times: maybe tens of thousands of times or more.”

Scaling up

The researchers say integrating part of their system’s design into existing semiconductor manufacturing should help it move beyond a research prototype.

“As you want to scale up, you need something that’s part of a consistent process, and that’s why chip foundry manufacturing becomes so important,” Hu says. “Working with a semiconductor foundry with well-defined process control is very powerful. It also allows you to implement each of the components into a single efficient process.”

The researchers are working to add more pixels to their array and develop more robust versions of their system so that it can start capturing more infrared information.

“In lots of cases when you’re taking images, you have prior knowledge of what you’re looking for,” Hu says. “You might be looking for a human in a dark room, or some specific features in an image, like a tree, and that prior information can be useful because now you can configure this system to specifically highlight those features.”

Hu also notes that researchers have used metasurfaces to emulate computational neural networks that power AI systems, though he notes that applications could be farther away from taking hold.

“This could enable more effective optical computing, where metasurfaces are used to encode network weights in neural networks,” Hu explains. “When light passes through the material, it interacts with the metasurface, and that information gets encoded in such a way that you can infer computational results. Researchers have already used this approach to emulate very complex neural networks.”

The work was supported, in part, by the U.S. Air Force, the U.S. National Science Foundation, the National Research Foundation of Korea, and the Draper Scholar Program.

A new study led by MIT researchers could drive the development of more energy-efficient digital displays — such as flat-screen TVs, augmented and virtual reality headsets, smartphone screens, medical imaging devices, and even large-area ambient lighting surfaces — that also generate richer, brighter colors.

The MIT scientists, in collaboration with researchers at Samsung, studied the microscopic changes that occur inside LEDs that utilize electrically excited quantum dots, which are precisely shaped nanoscale semiconductor particles that emit extremely pure colored light. 

Quantum dots are currently used in some of the computer and television displays with the best picture quality available. The efficiency of these displays could be further improved, and their manufacturing process further simplified, if the quantum dots could be electrically excited, as was first demonstrated in the quantum dot LED (QD-LED) structures over 20 years ago. 

But limitations on the operating lifespans of these QD-LEDs have prevented their widespread use in commercial applications.

The new study shows how encapsulating QD-LEDs in an acrylate-based resin can extend their lifespan by minimizing the physical degradation that would otherwise occur during QD-LED operation. 

The researchers demonstrated that encapsulating QD-LEDs with a resin layer using a simple, scalable process boosts stability and performance. In some devices, resin encapsulation enabled a 5,000-fold lifespan improvement. Importantly, their study reveals the fundamental reasons resin encapsulation is effective.

“The insights into how and why quantum dot LEDs get modified during their operation open the possibility of fixing everything that holds back commercialization of QD-LED displays. This technology can provide a light source like never before — pure in color, paper thin, and of large area, transforming how we produce both displays and general lighting,” says Vladimir Bulović, the Fariborz Maseeh (1990) Professor of Emerging Technology, principal investigator in the Research Laboratory of Electronics (RLE), director of MIT.nano, and senior author of this study.

He is joined on the paper by lead author Ruiqi Zhang, an electrical engineering and computer science graduate student; Moungi Bawendi, the Lester Wolfe Professor of Chemistry; and other colleagues at MIT and Samsung SAIT. The research appears today in Science Advances.

A blue bottleneck

This paper draws on foundational work by Bawendi, who shared the Nobel Prize in Chemistry in 2023 for discovering and synthesizing quantum dots, and engineering work by Bulović, who joined MIT in 2000, when he began collaborating with Bawendi to make efficient LED displays using quantum dots. 

Conventional LED displays utilize thousands of tiny lightbulbs that generate the red, green, and blue light needed to create the perception of any color on the visible spectrum. More advanced OLED screens, which Bulović was developing through his graduate work at Princeton University, utilize electrically excited, glowing organic molecules instead of light bulbs.

Bulović, Bawendi, and others at MIT sought to replace the organic molecules with quantum dots, which emit purer red, green, and blue light in a more energy-efficient manner.

“With quantum dots, the color quality of the screen would be more visually appealing and more optically flexible. One can mix and match those quantum dot colors more precisely to generate any color that is needed,” says Bulović.

Their collaboration generated a series of inventions on quantum dot LED technologies, leading to the launch of the startup QD Vision, which successfully commercialized the first-ever displays containing quantum dots. In 2016, QD Vision was acquired by Samsung, which incorporated a less efficient form of quantum dot technology into their “QLED” displays.

Although they are more energy-efficient, electrically excited QD-LEDs have still not been commercialized, particularly since the limited lifetime of the blue QD-LED does not meet the requirements of commercial displays.

“The blue quantum dot LEDs are 50 to 100 times less stable than their red and green counterparts. If you use them in an LED display, your TV might last for just a few months before it stops working. We wanted to understand what is different about the blue quantum dot LEDs,” Zhang says.

A nanoscale investigation

He and his collaborators developed a technique to slice a tiny QD-LED in nanoscale-thin slivers, revealing the device cross-section. They examined these cross-sections under extremely powerful microscopes at MIT.nano. This precise method allowed them to see what happens at the nanoscale to the ultrathin layers of materials stacked inside the QD-LED.

They explored the structural and chemical changes that occurred in each layer of red and blue QD-LEDs by comparing cross-sections of freshly made devices to cross-sections of devices that were operated on overdrive. The researchers found that during operation, the three core functional layers that enable blue QD-LEDs to glow are degraded, with modified morphology and reduced thickness. 

The distinct quantum dots also get merged together, losing their shape. This layer thinning and coarsening is caused, in part, by the release of extra hydrogen and oxygen during operation.

“We don’t yet know exactly where these extra elements are coming from — there are so many possibilities. But we definitely don’t want extra hydrogen and oxygen in the device,” Zhang says.

To prevent this degradation, they utilized a technique sometimes adopted by industry. They encapsulated the QD-LEDs with an acrylate-based resin.

They discovered that this encapsulation technique suppresses the release of the hydrogen and oxygen and inhibits some of the degradation that changes the morphology of the layers of the blue QD-LED. 

“For the first time, we have insights into the details of what happens inside these structures of many mixed and layered materials that form the QD-LED. No one knew this before,” Bulović says.

This encapsulation strategy, which is a cost-effective and scalable technique, led to an eightfold improvement in the lifetime of red QD-LEDs and more than a 5,000-fold lifetime improvement in blue QD-LEDs.

The researchers believe the resin prevents the formation of moisture in the cloud of gases that surrounds the quantum dot. That moisture likely causes the QD-LED to degrade. 

However, their experiments revealed that resin encapsulation does not eliminate all sources of degradation. 

The researchers are now exploring the addition of extra layers to QD-LEDs that could further improve efficiency and lifespan. They also plan to build on the lessons learned in this study to increase the stability of QD-LEDs for other applications. 

“This version of quantum dot LEDs would be better than anything that exists now — simpler to make, more efficient, and higher performing. This could open vistas into many more ways of thinking about this technology, not just for the sake of displays or lighting, but also for sensors, lasers, and so on,” says Bulović.

This work was funded by the Samsung Advanced Institute of Technology. The research was carried out, in part, using MIT.nano facilities.

Loons, gulls, puffins, and petrels are some of the 100 species of birds that can both fly and swim. These diving birds can plunge in water to swim after prey, and leap back into the air to fly away. 

Inspired by these naturally aquatic aviators, engineers at MIT and EPFL in Lausanne, Switzerland, have designed a robot that can swim underwater, then flap out of the water to continue flying through air, much like diving birds. 

The “flapping-wing aerial-aquatic vehicle,” or FAAV, weighs less than 300 grams (about half a pound) and is designed to help scientists study the mechanics that enable diving birds to fly through air and water. 

The robot has a central body, or fuselage; two flexible, flapping wings; and a steerable tail. The wings and tail can be swapped out for different sizes. In experiments carried out in a water tank and at a local lake, the engineers identified combinations of wing size, flapping frequency, and tail angle that enable the robot to smoothly transition from swimming through water to breaking through the surface to flying through the air.

Their results, which appear today in the journal Science, could help scientists understand how diving birds adapt their flight mechanics to move through air and water — mediums with very different physical properties. The design could also launch a new class of aerial-aquatic drones and vehicles. The researchers envision such winged robots could be deployed in oceanography to fly to and sample from aquatic regions that would otherwise be too dangerous for traditional ocean vessels to access.

“Our dream vision is for oceanographers, marine biologists, and members of coastal communities to launch this robot from a boat, or from shore, and it would fly close to the area of interest, such as an iceberg or a port facility, or over a pod of whales,” says Raphael Zufferey, assistant professor of mechanical engineering at MIT. “It would dive into the water to take a measurement or collect a sample, and fly back to deliver the data at a fraction of the cost of traditional methods. Then it could go back out to dive for more.” 

Zufferey is the lead author of the new study, which includes co-authors from EPFL and Northwest Indian College in Bellingham, Washington.

Flight mechanics

At MIT, Zufferey heads up the AURA Lab, where he and his students engineer aerial and aquatic vehicles inspired by biomechanics in nature. The robots they build are small in size and designed to unobtrusively explore and monitor the health of oceans and waterways. 

For their new work, the team aimed to design a vehicle that can fly in the air and underwater. Any such vehicle would have to adapt to and transition between two very different substances. Water is 1,000 times denser than air, and moving through one or the other requires very different mechanics. Or so people might assume.

“You have to do some adaptation to make that transition work. But there’s a solution that exists in nature,” Zufferey says. “Birds like puffins can fly very fast through the air, and can dive and swim through water at speeds of 3 meters per second. They’re able to do pretty amazing things. So we knew is was possible. Just no one had tried this in a mobile robotic system.”

To get an idea for how diving birds fly, the team looked through the scientific literature and pulled together available data on puffins, petrels, kingfishers, and other diving birds. They observed that smaller birds flap their wings around 10 times per second when flying through air, and around four times per second when swimming through water. Larger birds have a slightly lower flapping frequency through both air and water due to their wider wingspans. 

With the biomechanics of birds in mind, the team developed a winged robot designed to flap at similar frequencies to that of actual diving birds. 

Making the leap

The new robot roughly resembles a bird, with a body, two wings, and a tail. The body contains a battery and waterproof electric motor that drives a crankshaft, which in turn pumps the wings up and down at preset frequencies. The wings are made of thin membranes that are coated with hydrophobic nanoparticles to help wick away water. And the tail is motorized, enabling it to change its angle to help the robot fly up or dive down. 

The wings can be swapped out for different sizes. The researchers fabricated and tested three sets of wings: small (60 centimeters wide), medium (80 centimeters), and large (100 centimeters). They carried out experiments first in a small water tank, then in Lake Geneva in Switzerland.

In their tests, they placed the robot underwater, about half a meter below the surface. They programmed the wings to flap at certain frequencies and the tail to pitch at certain angles throughout the robot’s flight. They then observed under what conditions the robot successfully swam up toward the surface, out of the water and into the air. 

The robot flew multiple flights with different wing sizes, flapping frequencies, and tail angles. Overall, the team found the robot was able to reliably fly, swim, and transition between water and air when it flew with medium-sized wings. Flexibility in the wings is key; the wings need to be flexible enough to minimize flapping amplitude in water and also firm enough to keep the robot aloft in the air. 

The researchers also found the robot could swim through water at speeds of almost 1 meter per second when it flapped with a frequency of around 5 herz, or five flaps per second. The robot could fly through the air at around 6 meters per second, when flapping at a similar frequency. The speeds and flapping frequencies of the robot were similar to that of actual diving birds. 

To make the leap from water to air, they found the robot should be pitched at 70 degrees — a relatively steep angle that keeps the robot’s wingtips from touching the water’s surface as it flaps up and into the air. Any steeper, and the robot would tip back into the water.

Interestingly, this combination of wing size, flap frequency, and tail pitch enabled the robot to swim underwater, launch off the surface, and fly, without something that many diving birds require: feet. When birds such as puffins and ducks take off from the water’s surface, they paddle their feet, along with flapping their wings and pitching their tails. Surprisingly, Zufferey and his colleagues found that, at least in robotics, the act of flying out of water doesn’t necessarily require a paddling maneuver. 

“If you look at birds, most birds need to paddle at the surface to take off. And the question was, do we need the same for robots? And it turns out we don’t,” Zufferey says.

Going forward, the team is improving the design of the wings to enable them to turn in addition to flapping up and down. They will also test the robot’s performance under turbulent conditions, such as swimming out of choppy waters and flying through wind. Then, they hope to deploy the vehicle to help answer questions in ocean science.

“One of the major challenges in ocean science is collecting data both frequently and across many locations, which is something this robot could do in the future,” Zufferey says. “You could send this out not just every week, but every hour. It could fly out at high speeds, dive in fly back, deliver its data, and go back out, multiple times.”

This work was supported, in part, by a Marie Skłodowska-Curie Actions fellowship grant.

In southwestern Bangladesh, where extreme heat and severe tropical cyclones threaten the lives of millions of people, a new kind of climate refuge has opened its doors.

At the Baradal Aftab Uddin Collegiate School in the Satkhira district, the Jameel Observatory Climate Resilience Early Warning System Network (Jameel Observatory-CREWSnet) opened its first “adaptation fortress,” a solar-powered community shelter designed to protect residents from extreme heat and tropical storms.

A year-round refuge

When the heat arrives in southwestern Bangladesh, people have traditionally looked for relief under the shade of trees or near bodies of water. Now, during heatwaves, temperatures can reach 44 degrees Celsius (111 degrees Fahrenheit), levels at which shade is no longer enough.

A school by day and refuge from disaster, the adaptation fortress transforms the traditional concept of a cyclone shelter into a permanent year-round community resilience hub.

The facility offers residents protection from two of the region’s fastest-growing climate threats. During government-declared heat emergencies, it can host up to 200 people in four air-conditioned rooms supplied with clean drinking water. As a cyclone shelter, it can accommodate up to 500 people in additional rooms.

For the 30 million residents in southwestern Bangladesh, caught in a compounding cycle of cyclones and record-breaking heatwaves, the fortress represents something larger: a shift from reacting to disasters to preparing for them.

From forecast to fortress

That shift is the founding premise of the Jameel Observatory-CREWSnet project, which develops climate-resilience solutions that help vulnerable communities prepare for and adapt to life-altering conditions. 

The opening of the adaptation fortress marks a milestone for the project, and for MIT’s broader climate mission. Jameel Observatory-CREWSnet was one of MIT's five Climate Grand Challenges flagship projects, selected to translate climate research into tangible solutions for underserved communities facing some of the world’s most urgent climate threats. 

The project started in 2022 with Community Jameel and a research team at MIT led by Elfatih Eltahir, the H.M. King Bhumibol Professor of Hydrology and Climate in the Department of Civil and Environmental Engineering, along with John Aldridge, assistant leader of the Human Resilience Technology Group at MIT Lincoln Laboratory, and Deborah Campbell, senior staff scientist at MIT Lincoln Laboratory.  

Working in collaboration with BRAC International, a Bangladesh-founded nonprofit organization, the project combines advanced climate and socioeconomic forecasting with practical adaptation solutions. The adaptation fortress extends the project’s mission from forecasting climate threats to building permanent protection against them.

“When we launched the Jameel Observatory-CREWSnet, our goal was to close the gap between what climate science tells us is coming and what communities can actually do about it,” says Eltahir. “The adaptation fortress is that idea made concrete. Our models project more intense heatwaves for this region, and now residents of Satkhira have a place built to withstand them.”

The project’s climate modeling gives the fortress its urgency. Developed over decades in Eltahir’s research group, the models predict increasingly intense heatwaves across southwestern Bangladesh in the years ahead — dangerous heat layered on top of the cyclone risks they already endure. 

That same evidence shaped who gets through the door first. A priority access list focuses on those the heat endangers most: the elderly, people with respiratory conditions such as asthma, expectant mothers and mothers with infants, and students of the Baradal school.

Built to outlast the grid

The building was designed to weather climate shocks. A rooftop solar array powers the building as its primary energy source, with a battery backup that keeps it fully operational during grid outages. Solar grid-based air conditioning units combat extreme heat, and windows of glass encased in iron protect against breakage while sealing in the cool air.

The facility also integrates rainwater harvesting to mitigate the severe salinity that plagues local groundwater, and is designed to help cover its own upkeep. A net-metering interface allows surplus electricity generated during low-occupancy periods to be sold back to the national grid, creating a circular revenue stream that funds long-term maintenance.

The fortress is built with the community. A school committee oversees day-to-day operations and emergency protocols in partnership with BRAC, formalized through a signed memorandum of understanding to ensure long-term sustainability. The facility is supported by a comprehensive user guide translated into Bangla to empower local management.

Engineered to scale 

The Satkhira adaptation fortress is a pilot, and will be rigorously assessed. Remote sensors will track temperature, humidity, and power consumption. The findings will directly inform a second adaptation fortress planned for a secondary school in the Jashore district, where construction is scheduled to begin before the end of 2026.

If the evidence supports the model’s effectiveness, the concept could ultimately scale to as many as 1,250 fortresses across southwestern Bangladesh.

“From the start, our vision for this project has been a capability that could extend far beyond any single community,” says Campbell. “The adaptation fortress is a model we can learn from and refine in Satkhira, then carry to the many other places facing these same compounding climate threats.”

The work is supported by Community Jameel for Jameel Observatory CREWSnet, and by MIT Climate Grand Challenges.

The path to launching and growing a startup can be full of twists and turns. For a budding entrepreneur, gaining perspective from those who have already experienced the journey can be incredibly valuable, and highly inspirational. 

“There are so many amazing entrepreneurial stories among our alumni. We want to bring those stories to our students and our community and build networks with our incredible alumni founders,” says John Hart, the Class of 1922 Professor and head of the Department of Mechanical Engineering (MechE). “Through the Founder’s Journey class and other new programs, we want to cultivate interest in entrepreneurship among our students and expand opportunities to bring MechE-born technologies to the world.” 

According to a 2015 report on MIT’s global entrepreneurial impact, there are more than 30,000 active companies founded by MIT alumni worldwide, employing some 4.6 million people. Marina Hatsopoulos SM ’93, founding CEO of Z Corp., an early market leader in 3D printing, said one of the aims of the course was to show students they don’t need to reinvent everything. “So much of this has been done before. I want them to understand that this is a well-trod path.” 

Class 2.S977/2.S979 (Founder’s Journey: Launching and Scaling Hardware Startups) explores real-life challenges of startups focused on building and scaling hardware technologies. First held in spring 2025, the inaugural class invited students to “find and activate their entrepreneurial energy” through the lens of challenges faced by founders and their teams at various stages in development of new hardware-focused companies — ranging from fundraising to supply chain development, and much more. 

Each week of the class was structured around a key challenge faced during the development and growth of a hardware startup, presented by the instructors and guest speaker. The speakers were founders of companies in robotics, energy, 3D printing, consumer products, and other frontier technologies. Students engaged through preparing questions for the speakers and participating in follow-on discussions and reflective exercises throughout the semester. 

Ken Zolot, senior lecturer at MIT, and Hatsopoulous co-led the class and developed it along with Hart. Hart, who was among the alumni speakers in the course’s first iteration, also spoke to the class about his experience as a co-founder of VulcanForms, which began through collaboration with fellow co-founder Martin Feldmann MEng ’14. 

The other alumni speakers included Mick Mountz (Kiva/Amazon); Jon Hirschtick (Solidworks/Onshape); Max Lobovsky (Formlabs); Elise Strobach (Aeroshield); Greg Mark (Markforged); Seemantini Nadkarni (Coalesenz); Eran Egozy (Harmonix); Renuka Babu (DOTS Technology); Davide Marini (Inkbit); Loewen Cavill (Amira); and Colin Angle (iRobot).

Colin Angle ’89, SM ’91, co-founder of iRobot

Colin Angle ’89, SM ’91, co-founder and former CEO of iRobot, now CEO and co-founder of Familiar Machines and Magic, identified a passion for building things early on. 

“This idea that you can create something from nothing, that you can have an idea and not just draw it, but build it and make it real, is something I’ve always loved,” he says. “MIT had such a strong, hands-on ethos, and that really, powerfully resonated.”

While living in the Alpha Delta Phi Fraternity house at MIT, Angle watched several companies get their start (by his count, five multimillion-dollar companies were started by his fraternity brothers during his time in the house). Seeing others do it helped to demystify the process. 

He started iRobot in his living room, beginning at first not with a product concept, but a grand vision. “We’re supposed to have robots. So, if not us, who? And if not now, when? It was a magical day.” 

iRobot may be best known for the Roomba, an autonomous robotic vacuum cleaner, but through the years the company also sent robots to Afghanistan (saving thousands of lives with the Pack Bot tactical mobile robot) and explored the Great Pyramid in Giza live on National Geographic. 

“The joy I have taken from my entrepreneurial journey has been the ability to build bigger things, from building teams to building a company capable of building something far beyond what I could have ever imagined doing myself … we created inventions that no one thought possible, simply because we believed we could.”

Elise Strobach SM ’17, PhD ’20, CEO and co-founder of AeroShield

Elise Strobach SM ’17, PhD ’20 is CEO and co-founder of AeroShield Materials. The company, co-founded with Kyle Wilke PhD ’19 and Aaron Baskerville-Bridges SM ’20, MBA ’20, develops super-insulating transparent window inserts with technology based on transparent silica aerogels developed by Strobach while she was completing her PhD in Professor Evelyn Wang’s lab.

“I wasn’t thinking of myself as an entrepreneur at that time, but looking back, that’s definitely where that seed was planted,” says Strobach. As entrepreneurs, she says, “We have the … freedom to find the best problem to solve and to continue to seek the best way to solve that problem.”

Aerogels, which were first invented almost 100 years ago and were first commercialized by NASA to insulate equipment in space, had a hazy blue tint that limited their use in certain applications. The aerogel material created by Strobach and her team is completely see-through, creating a variety of new everyday applications. The company recently achieved another milestone, with their work on display at the Smithsonian National Air and Space Museum in Washington.

“You don’t have to know everything to start. You just have to know that this is what you want to do and just get started.”

Maxim Lobovsky SM ’11, CEO and co-founder of FormLabs

Maxim Lobovsky SM ’11 was already working on 3D printers when he came to MIT to study at the MIT Media Lab. As he was finishing his master’s degree, he saw an opportunity to build something new.  

Lobovsky, with fellow Media Lab graduates David Cranor SM ’11 and Natan Linder SM ’11, founded Formlabs, a developer and manufacturer of 3D printing technology. The trio set out to build a professional-level 3D printer, but a significant cost reduction and one that would be easier to use than what was then available on the market. At a time when 3D printers could cost $100,000 or more, Formlabs’ product started around $3,000.

“We definitely built Formlabs in a classic, disruptive innovation path,” Lobovsky says. They achieved the cost reduction through several different ways, including replacing technology developed in the 1980s with modern consumer electronics components like the laser diodes that were developed for Blu-ray Disc players, and with “just a lot of clever engineering.” 

It was a long grind to raise the first round of funding, he says. The team participated in MIT’s 100K competition and pitched their idea to many potential investors (with limited success, initially). Their big break came in the form of an overheard conversation. 

“As someone who is naturally introverted, shy engineer … a really important lesson [was] that, sometimes, you can get lucky,” he says. “Sometimes talking loudly at a restaurant is actually a good way to get things going.” 

Lobovsky and one of his co-founders were having dinner with a potential investor at Legal Seafoods in Harvard Square. The pitch to the initial investor didn’t go well, but Mitch Kapor, the founder of Lotus Software and an early pioneer in the PC industry overheard the conversation, and he ended up leading Formlabs’ first round of funding. 

Today, Formlabs is the largest supplier of professional stereolithography and selective laser sintering 3D printers in the world. 

Jon Hirschtick ’83, SM ’83, co-founder of SolidWorks and Onshape

Jon Hirschtick ’83, SM ’83, co-founder of SolidWorks and Onshape, says the first time he can remember thinking about starting a company was when he was an undergraduate. 

“I had heard about startups, and it sounded like a lot of things that I was drawn to … a sense of being able to realize your vision, express yourself; a sense of excitement, of making money, and even the idea of a chaotic environment,” he says. 

Hirschtick has spent over four decades building computer-aided design (CAD) software, starting as an intern at MIT in 1981 and continuing that work today. “I thought, ‘hey, the world could use this software.’ It’ll be a better place with the software that I envisioned.”

He refers to CAD as a meta product design. “We’re designing a product that other people use to design products, and that’s just really cool to me.” 

“I think startups just fit me,” he says. “The excitement, the idea of trying to solve a lot of problems at the same time. MIT is a place of problem-solving ... and a startup is a place where there’s lots of problems to solve.” He adds that a lot of big companies are doing new things, but “startups are always doing things.”

He says most anything today that is a manufactured product is modeled in CAD first. “If you’re interested and excited by product development, then building a CAD system lets you get involved in the world’s product development.”

“Nobody knows for sure when they start a company whether it’s going to be successful or not. If it were, if there was a way of knowing for sure, then there wouldn’t be all these classes in entrepreneurship. They’d just tell you the secret. There’s always risk. Visions and hallucinations, they look and feel the same. You only find out which is which once you really try to realize them.”

A version of this story appears in the 2026 issue of MechE Connects, the Department of Mechanical Engineering’s magazine. 

Researchers at MIT Lincoln Laboratory have designed a throwable, baseball-sized sensor that can remotely detect hazardous vapors and aerosols. 

Called the Tactical Optical Spherical Sensor for Interrogating Threats (TOSSIT), the sensor is designed to alert military service members, first responders, and law enforcement to the presence of chemical threats like nerve and blister agents, industrial chemical accidents, or fentanyl dust. 

Users can simply toss, drone-drop, or launch TOSSIT into an area of concern. To detect chemicals, the sensor samples the air and uses an internal camera to observe color changes on a removable dye card.

If certain chemicals are present, TOSSIT alerts users via an app or alarms in the sensor.

"TOSSIT fills an unmet need, providing a low-cost sensing option for vapors and solid aerosol threats — think toxic dust particles — that would otherwise not be detectable by small deployed sensor systems,” says principal investigator Richard Kingsborough.

After extensive testing in the field, the technology is being transferred to the U.S. military.

Most people think of the waterfront as the edge of the city. A team of MIT researchers sees it as a dynamic, Lego-like construction site.

Their new system, called “FloatForm,” is a swarm of small square robotic boats that assemble themselves into larger structures on the water, break apart, and reassemble into something new, all with minimal human direction. 

Each robot, about the size of a dinner plate at 21 centimeters square, is a self-contained vessel with its own thrusters, sensors, and magnetic latches. Together, they hint at a future in which floating infrastructure could become more adaptive: a temporary platform after an emergency, a market on a canal, or a stage that appears for a festival and dissolves when the crowd goes home.

“Our FloatForm projects envisions a future where the waterfront becomes a programmable extension of the city, where autonomous boats can self-organize into bridges, platforms, and other useful structures on demand,” says Daniela Rus, the Panasonic Professor of Electrical Engineering and Computer Science at MIT and director of MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL). “This kind of distributed robotics opens new possibilities for mobility, emergency response, public space, and infrastructure on water.”

“With FloatForm, we are essentially turning static water surfaces into dynamic, programmable spaces,” says Wei Wang, lead author of a new paper on the project and a former MIT research scientist who now leads the Marine Robotics Lab at the University of Wisconsin at Madison. “Imagine an urban environment where public space isn’t fixed, but can autonomously expand, contract, or reconfigure on demand.” 

“We see it as forming infrastructure on the water, using a modular system to create one larger system,” says Alejandro Gonzalez-Garcia, a former researcher with MIT CSAIL and the Senseable City Lab. “If there’s an emergency, you could form a new bridge to alleviate traffic in the city. Or you could create floating markets and floating stages. If you want a more livable city, you want to use the water, too.”

The open-access work, published today in Nature Communications, comes from the labs of Rus and Carlo Ratti, professor of practice of urban technologies and planning at MIT and director of the Senseable City Lab, and grows out of Roboat, their joint project with the Amsterdam Institute for Advanced Metropolitan Solutions that put full-size autonomous vessels on Amsterdam’s canals. Those canals once carried the city’s goods; today, they mostly carry tourists. 

“We explored whether the canals could be used for waste collection, or for transport, to offload some of the stress on the roads back onto the water,” says Niklas Hagemann, an MIT graduate student in architecture, CSAIL affiliate, and former Senseable City Lab researcher who has worked on the project since its early stages. “Urban areas are getting denser, so could you expand public space onto water that’s currently underutilized?”

FloatForm shrinks that vision down to tabletop scale to answer a harder question: How do you get dozens, and eventually thousands, of floating robots to organize themselves?

Lessons from the ant raft

The team found its answer in biology. Fire ants famously survive floods by linking their bodies into living rafts, with no leader choreographing the assembly. Each ant follows simple local rules, and a resilient structure emerges.

“Each ant is an independent agent,” says Gonzalez-Garcia. “We wanted each robot to have its own capabilities, the same way ant colonies form a raft.”

Most existing self-assembling robot systems, on water and elsewhere, rely on a central computer dictating every move. That approach is vulnerable to single points of failure and scales poorly: The planning math balloons as robots are added, and the swarm must assemble sequentially, with most robots idling while they wait their turn. FloatForm flips the balance. A lightweight central planner steps in only sparingly, assigning each robot a final position to perfect the lattice, a level of geometric precision that purely distributed methods struggle to guarantee. Everything else, including navigating toward the target shape, avoiding collisions, and adapting to disturbances, runs on the robots themselves, which coordinate by exchanging positions with their immediate neighbors. The whole swarm moves at once.

That parallelism is what sets the work apart. The planning complexity of FloatForms approach depends only on a robot’s local neighbors, not the total size of the swarm. “What we’re trying to do is to have minimal central intervention, and have them all move together at the same time,” says Gonzalez-Garcia.

In experiments at MIT, a fleet of eight robots repeatedly gathered from random positions into a target shape, latched into a rigid structure, broke apart on command, reassembled into a new configuration, and then drove across the pool as a single vessel, with each run taking four to eight minutes. In that final mode, called collective transport, a planner charts a trajectory for the whole structure and each robot computes its own contribution. “Every robot becomes an actuator,” Gonzalez-Garcia explains. Simulations showed the framework scaling smoothly to swarms of 64.

“The beauty of this largely decentralized approach is that the computation doesn’t get bogged down as the swarm grows,” says Wang. “Whether you are working with eight boats or 80, the entire fleet coordinates and moves simultaneously. Because the overall assembly time doesn’t significantly increase in principle, the system remains highly scalable.” 

There's a physical payoff to sticking together, too. “Our boats become more stable by joining together, like the ant raft, if you have waves or currents,” Hagemann says.

An origami handshake

The robots connect through a latching mechanism hidden entirely inside each hull. A single servo motor at the center drives an origami-inspired auxetic structure, a geometry that contracts uniformly in all directions at once, pulling permanent magnets on all four sides inward to release, or pushing them outward to grab a neighbor across gaps of 10 to 15 centimeters. The magnets are arranged with alternating polarities, so the boats reliably click into clean square lattices.

The elegant part is what the mechanism doesn’t do: consume (much) power. A 3D-printed gearbox holds the latch in either state with the motor switched off. “It uses energy to latch and de-latch, but in between those states, it doesn’t use any energy,” says Hagemann. For infrastructure that might hold a configuration for hours, that matters. “Because the robots are so small, you can only have a battery so big,” adds Gonzalez-Garcia. “If they use less energy on latching, they can use more on computation, or on actually moving.”

Getting there took some humbling engineering. Four miniature thrusters arranged in an “X” give each robot omnidirectional motion, including turning in place, but they pack large forces relative to the robots’ tiny inertia, which made early prototypes twitchy and prone to aggressive spins at low speeds. The team added stabilizing fins to increase hydrodynamic drag and tuned the controllers to stay robust across robots that, at this scale, are never quite identical. The magnets posed their own problem: They held on so well that de-latching sometimes required the robots to twist themselves free.

From the tank to the canal

Across 10 trials, the system completed its missions without human intervention 90 percent of the time with four robots and 70 percent with eight. When things did go wrong, the architecture showed its resilience: A robot that briefly lost its bearings could rejoin the structure on its own, without bringing the whole swarm to a halt, and robots stuck in formation deadlocks learned to shake themselves free and retry.

Moving from a controlled indoor tank to a real canal or harbor will take more than confidence. “There’s always a relationship between the size of a boat and the magnitude of the disturbance it can handle,” says Gonzalez-Garcia. “These boats are very small, so in very disturbed water, they cannot work.” Scaling up will mean reinforcing the latches, potentially with mechanical interlocking like the full-size Roboat used, and trading the lab’s ultrasonic indoor positioning for GPS or vision-based sensing. Helpfully, the coordination algorithm was designed to be sensor-agnostic: swap the sensors, keep the logic.

The team envisions applications well beyond city canals, from forming temporary platforms for offshore inspection and maintenance to adaptive sensor networks for studying migratory species to reconfigurable docking stations for emergency response in hard-to-reach areas. There is also potential for offshore and remote operations, from temporary construction platforms to environmental monitoring and scientific expeditions.

And the geography is wide open. “Venice, the Netherlands, Belgium, the fjords and lakes of Norway, really any city with a river can take advantage of this,” says Gonzalez-Garcia. “The project uses spaces where water is already important, but it also raises the question: Where else can water be used for something more?” 

“This is an exciting step forward in realizing distributed collective behaviors on water,” says University of Michigan Assistant Professor Steven Ceron, who wasn’t involved in the research. “Assembly, self-reconfiguration, and collective motion are difficult enough in dry environments, but achieving these behaviors in a predominantly distributed fashion on water represents a serious additional challenge, and this team has credibly overcome it. By shifting the computational burden onto the robots themselves, they have built a more resilient system that in the near future could enable robot collectives like this to be deployed in open-water environments for search operations, environmental monitoring, and reconfigurable marine infrastructure.”

Gonzalez-Garcia, Hagemann, and Wang wrote the paper with senior authors Ratti, who is also a professor at Politecnico di Milano, and Rus. Gonzalez-Garcia is additionally affiliated with the MECO Research Team at KU Leuven. The research was supported by a grant from the Amsterdam Institute for Advanced Metropolitan Solutions, with additional support from the University of Wisconsin at Madison. The team thanks MIT Sea Grant and Professor Michael Triantafyllou for providing the test tank.