Environment

Subterms

    Latest story

    50 Shares139 Views

    Mind, hand, and harvest

    by

    On a sunny, warm Sunday MIT students, staff, and faculty spread out across the fields of Hannan Healthy Foods in Lincoln, Massachusetts. Some of these volunteers pluck tomatoes from their vines in a patch a few hundred feet from the cars whizzing by on Route 117. Others squat in the shade cast by the greenhouse to snip chives. Still others slice heads of Napa cabbage from their roots in a bed nearer the woods. Everything being harvested today will wind up in Harvest Boxes, which will be sold at a pop-up farm stand the next day in the lobby of the Stata Center back on the MIT campus.This initiative — a pilot collaboration between MIT’s Office of Sustainability (MITOS), MIT Anthropology, Hannan Healthy Foods, and the nascent MIT Farm student organization — sold six-pound boxes of fresh, organic produce to the MIT community for $10 per box — half off the typical wholesale price. The weekly farm stands ran from Sept. 15 through Oct. 27.“There is a documented need for accessible, affordable, fresh food on college campuses,” says Heather Paxson, William R. Kenan, Jr. Professor of Anthropology and one of the organizers of the program. “The problems for a small farmer in finding a sufficient market … are connected to the challenges of food insecurity in even wealthy areas. And so, it really is about connecting those dots.”Through the six weeks of the project, farm stand shoppers purchased more than 2,000 pounds of fresh produce that they wouldn’t otherwise have had access to. Hannan, Paxson, and the team hope that this year’s pilot was successful enough to continue into future growing seasons, either in this farm stand form or as something else that can equally serve the campus community.“This year we decided to pour our heart, soul, and resources into this vision and prove what’s possible,” says Susy Jones, senior sustainability project manager at MITOS. “How can we do it in a way that is robust and goes through the official MIT channels, and yet pushes the boundaries of what’s possible at MIT?”A growing ideaMohammed Hannan, founder of Hannan Healthy Foods, first met Paxson and Jones in 2022. Jones was looking for someone local who grew vegetables common in Asian cuisine in response to a student request. Paxson wanted a small farm to host a field trip for her subject 21A.155 (Food, Culture and Politics). In July, Paxson and Jones learned about an article in the Boston Globe featuring Hannan as an example of a small farmer hit hard by federal budget cuts.They knew right away they wanted to help. They pulled in Zachary Rapaport and Aleks Banas, architecture master’s students and the co-founders of MIT Farm, an organization dedicated to getting the MIT community off campus and onto local farms. This MIT contingent connected with Hannan to come up with a plan.“These projects — when they flow, they flow,” says Jones. “There was so much common ground and excitement that we were all willing to jump on calls at 7 p.m. many nights to figure it out.”After a series of rapid-fire brainstorming sessions, the group decided to host weekly volunteer sessions at Hannan’s farm during the autumn growing season and sell the harvest at a farm stand on campus.“It fits in seamlessly with the MIT motto, ‘mind and hand,’ ‘mens et manus,’ learning by doing, as well as the heart, which has been added unofficially — mind, hand, heart,” says Paxson.Jones tapped into the MITOS network for financial, operational, student, and city partners. Rapaport and Banas put out calls for volunteers. Paxson incorporated a volunteer trip into her syllabus and allocated discretionary project funding to subsidize the cost of the produce, allowing the food to be sold at 50 percent of the wholesale price that Hannan was paid for it.“The fact that MIT students, faculty, and staff could come out to the farm, and that our harvest would circulate back to campus and into the broader community — there’s an energy around it that’s very different from academics. It feels essential to be part of something so tangible,” says Rapaport.The volunteer sessions proved to be popular. Throughout the pilot, about 75 students and half a dozen faculty and staff trekked out to Lincoln from MIT’s Cambridge, Massachusetts, campus at least once to clear fields and harvest vegetables. Hannan hopes the experience will change the way they think about their food.“Harvesting the produce, knowing the operation, knowing how hard it is, it’ll stick in their brain,” he says.On that September Sunday, second-year electrical engineering and computer science major Abrianna Zhang had come out with a friend after seeing a notification on the dormspam email lists. Zhang grew up in a California suburb big on supporting local farmers, but volunteering showed her a different side of the job.“There’s a lot of work that goes into raising all these crops and then getting all this manual labor,” says Zhang. “It makes me think about the economy of things. How is this even possible … for us to gain access to organic fruits or produce at a reasonable price?”Setting up shopSince mid-September, Monday has been Farm Stand day at MIT. Tables covered in green gingham tablecloths strike through the Stata Center lobby, holding stacks of cardboard boxes filled with produce. Customers wait in line to claim their piece of the fresh harvest — carrots, potatoes, onions, tomatoes, herbs, and various greens.Many of these students typically head to off-campus grocery stores to get their fresh produce. Katie Stabb, a sophomore civil and environmental engineering major and self-proclaimed “crazy plant lady,” grows her own food in the summer, but travels far from campus to shop for her vegetables during the school year. Having this stand right at MIT gives her time back, and she’s been spreading the news to her East Campus dorm mates — even picking boxes up for them when they can’t make it themselves and helping them figure out what to do with their excess ingredients.“I have encountered having way too many chives before, but that’s new for some folks,” she says. “Last week we pooled all of our chives and I made chive pancakes, kind of like scallion pancakes.”Stabb is not alone. In a multi-question customer survey conducted at the close of the Farm Stand season, 62 percent of respondents said the Harvest Box gave them the chance to try new foods and 49 percent experimented with new recipes. Seventy percent said this project helped them increase their vegetable intake.Nearly 60 percent of the survey respondents were graduate students living off campus. Banas, one of the MIT Farm co-leads, is one of those grad students enjoying the benefits.“I was cooking and making food that I bought from the farm stand and thought, ‘Oh, this is very literally influencing my life in a positive way.’ And I’m hoping that this has a similar impact for other people,” she says.The impact goes beyond the ability of students to nourish themselves with fresh vegetables. New communities have grown from this collaboration. Jones, for example, expanded her network at MITOS by tapping into expertise and resources from MIT Dining, the Vice President for Finance Merchant Services, and the MIT Federal Credit Union.“There were just these pockets of people in every corner of MIT who know how to do these very specific things that might seem not very glamorous, but make something like this possible,” says Jones. “It’s such a positive, affirming moment when you’re starting from scratch and someone’s like, ‘This is such a cool idea, how can I help?’”Strengthening communityInviting people from MIT to connect across campus and explore beyond Cambridge has helped students and employees alike feel like they’re part of something bigger.“The community that’s grown around this work is what keeps me so engaged,” says Rapaport. “MIT can have a bit of a siloing effect. It’s easy to become so focused on your classes and academics that your world revolves around them. Farm club grew out of wanting to build connections across the student body and to see ourselves and MIT as part of a larger network of people, communities, and relationships.”This particular connection will continue to grow, as Rapaport and Banas will use their architectural expertise to lead a design-build team in developing a climate-adaptive and bio-based root cellar at Hannan Healthy Foods, to improve the farm’s winter vegetable storage conditions. Community engagement is an ethos Hannan has embraced since the start of his farming journey in 2018, motivated by a desire to provision first his family and then others with healthy food.“One thing I have done over the years, I was not trying to do farming by myself,” he says. “I always reached out to as many people as I could. The idea is, if community is not involved, they just see it as an individual business.”It’s why he gifts his volunteers huge bags of tomatoes at the end of a shift, or donates some of his harvest to food banks, or engages an advisory committee of local residents to ensure he’s filling the right needs.“There’s a reciprocal dimension to gifting that needs to continue,” says Paxson. “That is what builds and maintains community — it’s classic anthropology.”And much of what’s exchanged in this type of reciprocity can’t be charted or graded or marked on a spreadsheet. It’s cooking pancakes with dorm mates. It’s meeting and appreciating new colleagues. It’s grabbing a friend to harvest cabbage on a beautiful autumn Sunday.“Seeing a student who volunteered over the weekend harvesting chives come to the market on Monday and then want to take a selfie with those chives,” says Jones. “To me, that’s a cool moment.” More

    More stories

    • 100 Shares113 Views

      in Environment

      How a building creates and defines a region

      by

      As an undergraduate majoring in architecture, Dong Nyung Lee ’21 wasn’t sure how to respond when friends asked him what the study of architecture was about.“I was always confused about how to describe it myself,” he says with a laugh. “I would tell them that it wasn’t just about a building, or a city, or a community. It’s a balance across different scales, and it has to touch everything all at once.”As a graduate student enrolled in a design studio course last spring — 4.154 (Territory as Interior) — Lee and his classmates had to design a building that would serve a specific community in a specific location. The course, says Lee, gave him clarity as to “what architecture is all about.”Designed by Roi Salgueiro Barrio, a lecturer in the MIT School of Architecture and Planning’s Department of Architecture, the coursework combines ecological principles, architectural design, urban economics, and social considerations to address real-world problems in marginalized or degraded areas.“When we build, we always impact economies, mostly by the different types of technologies we use and their dependence on different types of labor and materials,” says Salgueiro Barrio. “The intention here was to think at both levels: the activities that can be accommodated, and how we can actually build something.”Research firstStudents were tasked with repurposing an abandoned fishing industry building on the Barbanza Peninsula in Galicia, Spain, and proposing a new economic activity for the building that would help regenerate the local economy. Working in groups, they researched the region’s material resources and fiscal sectors and designed detailed maps. This approach to constructing a building was new for Vincent Jackow a master’s student in architecture.“Normally in architecture, we work at the scale of one-to-100 meters,” he says. But this process allowed me to connect the dots between what the region offered and what could be built to support the economy.”The aim of revitalizing this area is also a goal of Fundación RIA (FRIA), a nonprofit think tank established by Pritzker Prize-winning architect David Chipperfield. FRIA generates research and territorial planning with the goal of long-term sustainability of the built and natural environment in the Galicia region. During their spring break in March, the students traveled to Galicia, met with Chipperfield, business owners, fishermen, and farmers, and explored a variety of sites. They also consulted with the owner of the building they were to repurpose.Returning to MIT, the students constructed nine detailed models. Master’s student Aleks Banaś says she took the studio because it required her to explore the variety of scales in an architectural project from territorial analysis to building detail, all while keeping the socio-economic aspect of design decisions in mind.“I’m interested in how architecture can support local economies,” says Banaś. “Visiting Galicia was very special because of the communities we interacted with. We were no longer looking at articles and maps of the region; we were learning about day-to-day life. A lot of people shared with us the value of their work, which is not economically feasible.”Banaś was impressed by the region’s strong maritime history and the generations of craftspeople working on timber boat-making. Inspired by the collective spirit of the region, she designed “House of Sea,” transforming the former cannery into a hub for community gathering and seafront activities. The reimagined building would accommodate a variety of functions including a boat-building workshop for the Ribeira carpenters’ association, a restaurant, and a large, covered section for local events such as the annual barnacle festival.“I wanted to demonstrate how we can create space for an alternative economy that can host and support these skills and traditions,” says Banaś. Jackow’s building — “La Nueva Cordelería,” or “New Rope Making” — was a facility using hemp to produce rope and hempcrete blocks (a construction material). The production of both “is very on-trend in the E.U.” and provides an alternative to petrochemical-based ropes for the region’s marine uses, says Jackow. The building would serve as a cultural hub, incorporating a café, worker housing, and offices. Even its very structure would also make use of the rope by joining timber with knots allowing the interior spaces to be redesigned.Lee’s building was designed to engage with the forestry and agricultural industries.“What intrigued me was that Galicia is heavily dependent on pulp production and wood harvesting,” he says. “I wanted to give value to the post-harvest residue.”Lee designed a biochar plant using some of the concrete and terra cotta blocks on site. Biochar is made by heating the harvested wood residue through pyrolysis — thermal decomposition in an environment with little oxygen. The resulting biochar would be used by farmers for soil enhancement.“The work demonstrated an understanding of the local resources and using them to benefit the revitalization of the area,” says Salgueiro Barrio, who was pleased with the results. FRIA was so impressed with the work that they held an exhibition at their gallery in Santiago de Compostela in August and September to highlight the importance of connecting academic research with the territory through student projects. Banaś interned with FRIA over the summer working on multiple projects, including the plan and design for the exhibition. The challenge here, she says, was to design an exhibition of academic work for a general audience. The final presentation included maps, drawings, and photographs by the students.For Lee, the course was more meaningful than any he has taken to date. Moving between the different scales of the project illustrated, for him, “the biggest challenge for a designer and an architect. Architecture is universal, and very specific. Keeping those dualities in focus was the biggest challenge and the most interesting part of this project. It hit at the core of what architecture is.” More

    • 125 Shares89 Views

      in Environment

      Ultrasonic device dramatically speeds harvesting of water from the air

      by

      Feeling thirsty? Why not tap into the air? Even in desert conditions, there exists some level of humidity that, with the right material, can be soaked up and squeezed out to produce clean drinking water. In recent years, scientists have developed a host of promising sponge-like materials for this “atmospheric water harvesting.”But recovering the water from these materials usually requires heat — and time. Existing designs rely on heat from the sun to evaporate water from the materials and condense it into droplets. But this step can take hours or even days. Now, MIT engineers have come up with a way to quickly recover water from an atmospheric water harvesting material. Rather than wait for the sun to evaporate water out, the team uses ultrasonic waves to shake the water out.The researchers have developed an ultrasonic device that vibrates at high frequency. When a water-harvesting material, known as a “sorbent,” is placed on the device, the device emits ultrasound waves that are tuned to shake water molecules out of the sorbent. The team found that the device recovers water in minutes, versus the tens of minutes or hours required by thermal designs.

      Play video

      MIT engineers design an ultrasonic system to “shake” water out of an atmospheric water harvester. The new design can recover captured water in minutes rather than hours.

      Unlike heat-based designs, the device does require a power source. The team envisions that the device could be powered by a small solar cell, which could also act as a sensor to detect when the sorbent is full. It could also be programmed to automatically turn on whenever a material has harvested enough moisture to be extracted. In this way, a system could soak up and shake out water from the air over many cycles in a single day.“People have been looking for ways to harvest water from the atmosphere, which could be a big source of water particularly for desert regions and places where there is not even saltwater to desalinate,” says Svetlana Boriskina, principal research scientist in MIT’s Department of Mechanical Engineering. “Now we have a way to recover water quickly and efficiently.”Boriskina and her colleagues report on their new device in a study appearing today in the journal Nature Communications. The study’s first author is Ikra Iftekhar Shuvo, an MIT graduate student in media arts and sciences, along with Carlos Díaz-Marín, Marvin Christen, Michael Lherbette, and Christopher Liem.Precious hoursBoriskina’s group at MIT develops materials that interact with the environment in novel ways. Recently, her group explored atmospheric water harvesting (AWH), and ways that materials can be designed to efficiently absorb water from the air. The hope is that, if they can work reliably, AWH systems would be of most benefit to communities where traditional sources of drinking water — and even saltwater — are scarce.Like other groups, Boriskina’s lab had generally assumed that an AWH system in the field would absorb moisture during the night, and then use the heat from the sun during the day to naturally evaporate the water and condense it for collection.“Any material that’s very good at capturing water doesn’t want to part with that water,” Boriskina explains. “So you need to put a lot of energy and precious hours into pulling water out of the material.”She realized there could be a faster way to recover water after Ikra Shuvo joined her group. Shuvo had been working with ultrasound for wearable medical device applications. When he and Boriskina considered ideas for new projects, they realized that ultrasound could be a way to speed up the recovery step in atmospheric water harvesting.“It clicked: We have this big problem we’re trying to solve, and now Ikra seemed to have a tool that can be used to solve this problem,” Boriskina recalls.Water danceUltrasound, or ultrasonic waves, are acoustic pressure waves that travel at frequencies of over 20 kilohertz (20,000 cycles per second). Such high-frequency waves are not visible or audible to humans. And, as the team found, ultrasound vibrates at just the right frequency to shake water out of a material.“With ultrasound, we can precisely break the weak bonds between water molecules and the sites where they’re sitting,” Shuvo says. “It’s like the water is dancing with the waves, and this targeted disturbance creates momentum that releases the water molecules, and we can see them shake out in droplets.”Shuvo and Boriskina designed a new ultrasonic actuator to recover water from an atmospheric water harvesting material. The heart of the device is a flat ceramic ring that vibrates when voltage is applied. This ring is surrounded by an outer ring that is studded with tiny nozzles. Water droplets that shake out of a material can drop through the nozzle and into collection vessels attached above and below the vibrating ring.They tested the device on a previously designed atmospheric water harvesting material. Using quarter-sized samples of the material, the team first placed each sample in a humidity chamber, set to various humidity levels. Over time, the samples absorbed moisture and became saturated. The researchers then placed each sample on the ultrasonic actuator and powered it on to vibrate at ultrasonic frequencies. In all cases, the device was able to shake out enough water to dry out each sample in just a few minutes.The researchers calculate that, compared to using heat from the sun, the ultrasonic design is 45 times more efficient at extracting water from the same material.“The beauty of this device is that it’s completely complementary and can be an add-on to almost any sorbent material,” says Boriskina, who envisions a practical, household system might consist of a fast-absorbing material and an ultrasonic actuator, each about the size of a window. Once the material is saturated, the actuator would briefly turn on, powered by a solar cell, to shake out the water. The material would then be ready to harvest more water, in multiple cycles throughout a single day.“It’s all about how much water you can extract per day,” she says. “With ultrasound, we can recover water quickly, and cycle again and again. That can add up to a lot per day.”This work was supported, in part, by the MIT Abdul Latif Jameel Water and Food Systems Lab and the MIT-Israel Zuckerman STEM Fund. More

    • 75 Shares137 Views

      in Environment

      Small, inexpensive hydrophone boosts undersea signals

      by

      Researchers at MIT Lincoln Laboratory have developed a first-of-its-kind hydrophone built around a simple, commercially available microphone. The device, leveraging a common microfabrication process known as microelectromechanical systems (MEMS), is significantly smaller and less expensive than current hydrophones, yet has equal or exceeding sensitivity. The hydrophone could have applications for the U.S. Navy, as well as industry and the scientific research community.”Given the broad interest from the Navy in low-cost hydrophones, we were surprised that this design had not been pursued before,” says Daniel Freeman, who leads this work in the Advanced Materials and Microsystems Group. “Hydrophones are critical for undersea sensing in a variety of applications and platforms. Our goal was to demonstrate that we could develop a device at reduced size and cost without sacrificing performance.”Essentially an underwater microphone, a hydrophone is an instrument that converts sound waves into electrical signals, allowing us to “hear” and record sounds in the ocean and other bodies of water. These signals can later be analyzed and interpreted, providing valuable information about the underwater environment.MEMS devices are incredibly small systems — ranging from a few millimeters down to microns (smaller than a human hair) — with tiny moving parts. They are used in a variety of sensors, including microphones, gyroscopes, and accelerometers. The small size of MEMS sensors has made them crucial in various applications, from smartphones to medical devices. Currently, no commercially available hydrophones utilize MEMS technology, so the team set out to understand whether such a design was possible.With funding from the Office of the Under Secretary of War for Research and Engineering to develop a novel hydrophone, the team first planned to use microfabrication, an area of expertise at the laboratory, to develop their device. However, that approach proved to be too costly and involved to pursue. This obstacle led the team to pivot and build their hydrophone around a commercially available MEMS microphone. “We had to come up with an inexpensive alternative without giving up performance, and this is what led us to build the design around a microphone, which to our knowledge is a novel approach,” Freeman explains.In collaboration with researchers at Tufts University, as well as industry partners SeaLandAire Technologies and Navmar Applied Sciences Corp., the team made the hydrophone by encapsulating the MEMS microphone in a polymer with low permeability to water while leaving an air cavity around the microphone’s diaphragm (the component of the microphone that vibrates in response to sound waves). One key challenge that they faced was the possibility of losing too much signal to the packaging and the air cavity around the MEMS microphone. After a substantial amount of simulation, design iterations, and testing, the team found that the signal lost from incorporating air into the device was compensated for by the very high sensitivity of the MEMS microphone itself. As a result, the device was able to perform at a sensitivity comparable to high-end hydrophones at depths down to 400 feet and temperatures as low as 40 degrees Fahrenheit. To date, the collaborative effort has involved computational modeling, system electronics design and fabrication, prototype unit manufacturing, and calibrator and pool testing.In July, eight researchers traveled to Seneca Lake in New York to test a variety of devices. The hydrophones were lowered to increasing depths in the water — 100 feet at first, then incrementally lower down to 400 feet. At each depth, acoustic signals of varying frequencies were transmitted for the instrument to record. The transmitted signals were calibrated to a known level so they could then measure the actual sensitivity of the hydrophones across different frequencies. When the sound hits the hydrophone’s diaphragm, it generates an electrical signal that is amplified, digitized, and transmitted to a recording device at the surface for post-test data analysis. The team utilized both commercial underwater cables as well as Lincoln Laboratory’s fiber-based sensing arrays.”This was our first field test in deep water, and therefore it was an important milestone in demonstrating the ability to operate in a realistic environment, rather than the water chambers that we’d been using,” Freeman says. “Our hope was that the performance of our device would match what we’ve seen in our water tank, where we tested at high hydrostatic pressure across a range of frequencies. In other words, we hoped this test would provide results that confirm our predictions based on lab-based testing.”The test results were excellent, showing that the sensitivity and the signal-to-noise was within a few decibels of the quietest ocean state, known as sea state zero. Moreover, this performance was achieved in deep water, at 400 feet, and with very low temperatures, around 40 degrees Fahrenheit.The prototype hydrophone has applications across a wide variety of commercial and military use-cases owing to its small size, efficient power draw, and low cost.”We’re in discussion with the Department of War about transitioning this technology to the U.S. government and industry,” says Freeman. “There is still some room for optimizing the design, but we think we’ve demonstrated that this hydrophone has the key benefits of being robust, high performance, and very low cost.” More

    • 175 Shares107 Views

      in Energy

      From nanoscale to global scale: Advancing MIT’s special initiatives in manufacturing, health, and climate

      by

      “MIT.nano is essential to making progress in high-priority areas where I believe that MIT has a responsibility to lead,” opened MIT president Sally Kornbluth at the 2025 Nano Summit. “If we harness our collective efforts, we can make a serious positive impact.”It was these collective efforts that drove discussions at the daylong event hosted by MIT.nano and focused on the importance of nanoscience and nanotechnology across MIT’s special initiatives — projects deemed critical to MIT’s mission to help solve the world’s greatest challenges. With each new talk, common themes were reemphasized: collaboration across fields, solutions that can scale up from lab to market, and the use of nanoscale science to enact grand-scale change.“MIT.nano has truly set itself apart, in the Institute’s signature way, with an emphasis on cross-disciplinary collaboration and open access,” said Kornbluth. “Today, you’re going to hear about the transformative impact of nanoscience and nanotechnology, and how working with the very small can help us do big things for the world together.”Collaborating on healthAngela Koehler, faculty director of the MIT Health and Life Sciences Collaborative (MIT HEALS) and the Charles W. and Jennifer C. Johnson Professor of Biological Engineering, opened the first session with a question: How can we build a community across campus to tackle some of the most transformative problems in human health? In response, three speakers shared their work enabling new frontiers in medicine.Ana Jaklenec, principal research scientist at the Koch Institute for Integrative Cancer Research, spoke about single-injection vaccines, and how her team looked to the techniques used in fabrication of electrical engineering components to see how multiple pieces could be packaged into a tiny device. “MIT.nano was instrumental in helping us develop this technology,” she said. “We took something that you can do in microelectronics and the semiconductor industry and brought it to the pharmaceutical industry.”While Jaklenec applied insight from electronics to her work in health care, Giovanni Traverso, the Karl Van Tassel Career Development Professor of Mechanical Engineering, who is also a gastroenterologist at Brigham and Women’s Hospital, found inspiration in nature, studying the cephalopod squid and remora fish to design ingestible drug delivery systems. Representing the industry side of life sciences, Mirai Bio senior vice president Jagesh Shah SM ’95, PhD ’99 presented his company’s precision-targeted lipid nanoparticles for therapeutic delivery. Shah, as well as the other speakers, emphasized the importance of collaboration between industry and academia to make meaningful impact, and the need to strengthen the pipeline for young scientists.Manufacturing, from the classroom to the workforcePaving the way for future generations was similarly emphasized in the second session, which highlighted MIT’s Initiative for New Manufacturing (MIT INM). “MIT’s dedication to manufacturing is not only about technology research and education, it’s also about understanding the landscape of manufacturing, domestically and globally,” said INM co-director A. John Hart, the Class of 1922 Professor and head of the Department of Mechanical Engineering. “It’s about getting people — our graduates who are budding enthusiasts of manufacturing — out of campus and starting and scaling new companies,” he said.On progressing from lab to market, Dan Oran PhD ’21 shared his career trajectory from technician to PhD student to founding his own company, Irradiant Technologies. “How are companies like Dan’s making the move from the lab to prototype to pilot production to demonstration to commercialization?” asked the next speaker, Elisabeth Reynolds, professor of the practice in urban studies and planning at MIT. “The U.S. capital market has not historically been well organized for that kind of support.” She emphasized the challenge of scaling innovations from prototype to production, and the need for workforce development.“Attracting and retaining workforce is a major pain point for manufacturing businesses,” agreed John Liu, principal research scientist in mechanical engineering at MIT. To keep new ideas flowing from the classroom to the factory floor, Liu proposes a new worker type in advanced manufacturing — the technologist — someone who can be a bridge to connect the technicians and the engineers.Bridging ecosystems with nanoscienceBridging people, disciplines, and markets to affect meaningful change was also emphasized by Benedetto Marelli, mission director for the MIT Climate Project and associate professor of civil and environmental engineering at MIT.“If we’re going to have a tangible impact on the trajectory of climate change in the next 10 years, we cannot do it alone,” he said. “We need to take care of ecology, health, mobility, the built environment, food, energy, policies, and trade and industry — and think about these as interconnected topics.”Faculty speakers in this session offered a glimpse of nanoscale solutions for climate resiliency. Michael Strano, the Carbon P. Dubbs Professor of Chemical Engineering, presented his group’s work on using nanoparticles to turn waste methane and urea into renewable materials. Desirée Plata, the School of Engineering Distinguished Climate and Energy Professor, spoke about scaling carbon dioxide removal systems. Mechanical engineering professor Kripa Varanasi highlighted, among other projects, his lab’s work on improving agricultural spraying so pesticides adhere to crops, reducing agricultural pollution and cost.In all of these presentations, the MIT faculty highlighted the tie between climate and the economy. “The economic systems that we have today are depleting to our resources, inherently polluting,” emphasized Plata. “The goal here is to use sustainable design to transition the global economy.”What do people do at MIT.nano?This is where MIT.nano comes in, offering shared access facilities where researchers can design creative solutions to these global challenges. “What do people do at MIT.nano?” asked associate director for Fab.nano Jorg Scholvin ’00, MNG ’01, PhD ’06 in the session on MIT.nano’s ecosystem. With 1,500 individuals and over 20 percent of MIT faculty labs using MIT.nano, it’s a difficult question to quickly answer. However, in a rapid-fire research showcase, students and postdocs gave a response that spanned 3D transistors and quantum devices to solar solutions and art restoration. Their work reflects the challenges and opportunities shared at the Nano Summit: developing technologies ready to scale, uniting disciplines to tackle complex problems, and gaining hands-on experience that prepares them to contribute to the future of hard tech.The researchers’ enthusiasm carried the excitement and curiosity that President Kornbluth mentioned in her opening remarks, and that many faculty emphasized throughout the day. “The solutions to the problems we heard about today may come from inventions that don’t exist yet,” said Strano. “These are some of the most creative people, here at MIT. I think we inspire each other.”Robert N. Noyce (1953) Cleanroom at MIT.nanoCollaborative inspiration is not new to the MIT culture. The Nano Summit sessions focused on where we are today, and where we might be going in the future, but also reflected on how we arrived at this moment. Honoring visionaries of nanoscience and nanotechnology, President Emeritus L. Rafael Reif delivered the closing remarks and an exciting announcement — the dedication of the MIT.nano cleanroom complex. Made possible through a gift by Ray Stata SB ’57, SM ’58, this research space, 45,000 square feet of ISO 5, 6, and 7 cleanrooms, will be named the Robert N. Noyce (1953) Cleanroom.“Ray Stata was — and is — the driving force behind nanoscale research at MIT,” said Reif. “I want to thank Ray, whose generosity has allowed MIT to honor Robert Noyce in such a fitting way.”Ray Stata co-founded Analog Devices in 1965, and Noyce co-founded Fairchild Semiconductor in 1957, and later Intel in 1968. Noyce, widely regarded as the “Mayor of Silicon Valley,” became chair of the Semiconductor Industry Association in 1977, and over the next 40 years, semiconductor technology advanced a thousandfold, from micrometers to nanometers.“Noyce was a pioneer of the semiconductor industry,” said Stata. “It is due to his leadership and remarkable contributions that electronics technology is where it is today. It is an honor to be able to name the MIT.nano cleanroom after Bob Noyce, creating a permanent tribute to his vision and accomplishments in the heart of the MIT campus.”To conclude his remarks and the 2025 Nano Summit, Reif brought the nano journey back to today, highlighting technology giants such as Lisa Su ’90, SM ’91, PhD ’94, for whom Building 12, the home of MIT.nano, is named. “MIT has educated a large number of remarkable leaders in the semiconductor space,” said Reif. “Now, with the Robert Noyce Cleanroom, this amazing MIT community is ready to continue to shape the future with the next generation of nano discoveries — and the next generation of nano leaders, who will become living legends in their own time.” More

    • 138 Shares167 Views

      in Environment

      MIT senior turns waste from the fishing industry into biodegradable plastic

      by

      Sometimes the answers to seemingly intractable environmental problems are found in nature itself. Take the growing challenge of plastic waste. Jacqueline Prawira, an MIT senior in the Department of Materials Science and Engineering (DMSE), has developed biodegradable, plastic-like materials from fish offal, as featured in a recent segment on the CBS show “The Visioneers with Zay Harding.” “We basically made plastics to be too good at their job. That also means the environment doesn’t know what to do with this, because they simply won’t degrade,” Prawira told Harding. “And now we’re literally drowning in plastic. By 2050, plastics are expected to outweigh fish in the ocean.” “The Visioneers” regularly highlights environmental innovators. The episode featuring Prawira premiered during a special screening at Climate Week NYC on Sept. 24.Her inspiration came from the Asian fish market her family visits. Once the fish they buy are butchered, the scales are typically discarded. “But I also started noticing they’re actually fairly strong. They’re thin, somewhat flexible, and pretty lightweight, too, for their strength,” Prawira says. “And that got me thinking: Well, what other material has these properties? Plastics.” She transformed this waste product into a transparent, thin-film material that can be used for disposable products such as grocery bags, packaging, and utensils. Both her fish-scale material and a composite she developed don’t just mimic plastic — they address one of its biggest flaws. “If you put them in composting environments, [they] will degrade on their own naturally without needing much, if any, external help,” Prawira says. This isn’t Prawira’s first environmental innovation. Working in DMSE Professor Yet-Ming Chiang’s lab, she helped develop a low-carbon process for making cement — the world’s most widely used construction material, and a major emitter of carbon dioxide. The process, called silicate subtraction, enables compounds to form at lower temperatures, cutting fossil fuel use. Prawira and her co-inventors in the Chiang lab are also using the method to extract valuable lithium with zero waste. The process is patented and is being commercialized through the startup Rock Zero. For her achievements, Prawira recently received the Barry Goldwater Scholarship, awarded to undergraduates pursuing careers in science, mathematics, or engineering. In her “Visioneers” interview, she shared her hope for more sustainable ways of living. “I’m hoping that we can have daily lives that can be more in sync with the environment,” Prawira said. “So you don’t always have to choose between the convenience of daily life and having to help protect the environment.” More

    • 138 Shares199 Views

      in Energy

      Giving buildings an “MRI” to make them more energy-efficient and resilient

      by

      Older buildings let thousands of dollars-worth of energy go to waste each year through leaky roofs, old windows, and insufficient insulation. But even as building owners face mounting pressure to comply with stricter energy codes, making smart decisions about how to invest in efficiency is a major challenge.Lamarr.AI, born in part from MIT research, is making the process of finding ways to improve the energy efficiency of buildings as easy as clicking a button. When customers order a building review, it triggers a coordinated symphony of drones, thermal and visible-range cameras, and artificial intelligence designed to identify problems and quantify the impact of potential upgrades. Lamarr.AI’s technology also assesses structural conditions, creates detailed 3D models of buildings, and recommends retrofits. The solution is already being used by leading organizations across facilities management as well as by architecture, engineering, and construction firms.“We identify the root cause of the anomalies we find,” says CEO and co-founder Tarek Rakha PhD ’15. “Our platform doesn’t just say, ‘This is a hot spot and this is a cold spot.’ It specifies ‘This is infiltration or exfiltration. This is missing insulation. This is water intrusion.’ The detected anomalies are also mapped to a 3D model of the building, and there are deeper analytics, such as the cost of each retrofit and the return on investment.”To date, the company estimates its platform has helped clients across health care, higher education, and multifamily housing avoid over $3 million in unnecessary construction and retrofit costs by recommending targeted interventions over costly full-system replacements, while improving energy performance and extending asset life. For building owners managing portfolios worth hundreds of millions of dollars, Lamarr.AI’s approach represents a fundamental shift from reactive maintenance to strategic asset management.The founders, who also include MIT Professor John Fernández and Research Scientist Norhan Bayomi SM ’17, PhD ’21, are thrilled to see their technology accelerating the transition to more energy-efficient and higher-performing buildings.“Reducing carbon emissions in buildings gets you the greatest return on investment in terms of climate interventions, but what has been needed are the technologies and tools to help the real estate and construction sectors make the right decisions in a timely and economical way,” Fernández says.Automating building scansBayomi and Rakha completed their PhDs in the MIT Department of Architecture’s Building Technology Program. For her thesis, Bayomi developed technology to detect features of building exteriors and classify thermal anomalies through scans of buildings, with a specific focus on the impact of heat waves on low-income communities. Bayomi and her collaborators eventually deployed the system to detect air leaks as part of a partnership with a community in New York City.After graduating MIT, Rakha became an assistant professor at Syracuse University. In 2015, together with fellow Syracuse University Professor Senem Velipasalar, he began developing his concept for drone-based building analytics — an idea that later received support through a grant from New York State’s Department of Economic Development. In 2019, Bayomi and Fernández joined the project, and the team received a $1.8 million research award from the U.S. Department of Energy.“The technology is like giving a building an MRI using drones, infrared imaging, visible light imaging, and proprietary AI that we developed through computer vision technology, along with large language models for report generation,” Rakha explains.“When we started the research, we saw firsthand how vulnerable communities were suffering from inefficient buildings, but couldn’t afford comprehensive diagnostics,” Bayomi says. “We knew that if we could automate this process and reduce costs while improving accuracy, we’d unlock a massive market. Now we’re seeing demand from everyone, from municipal buildings to major institutional portfolios.”Lamarr.AI was officially founded in 2021 to commercialize the technology, and the founders wasted no time tapping into MIT’s entrepreneurial ecosystem. First, they received a small seed grant from the MIT Sandbox Innovation Fund. In 2022, they won the MITdesignX prize and were semifinalists in the MIT $100K Entrepreneurship Competition. The founders named the company after Hedy Lamarr, the famous actress and inventor of a patented technology that became the basis for many modern secure communications.Current methods for detecting air leaks in buildings utilize fan pressurizers or smoke. Contractors or building engineers may also spot-check buildings with handheld infrared cameras to manually identify temperature differences across individual walls, windows, and ductwork.Lamarr.AI’s system can perform building inspections far more quickly. Building managers can order the company’s scans online and select when they’d like the drone to fly. Lamarr.AI partners with drone companies worldwide to fly off-the-shelf drones around buildings, providing them with flight plans and specifications for success. Images are then uploaded onto Lamarr.AI’s platform for automated analysis.“As an example, a survey of a 180,000-square-foot building like the MIT Schwarzman College of Computing, which we scanned, produces around 2,000 images,” Fernández says. “For someone to go through those manually would take a couple of weeks. Our models autonomously analyze those images in a few seconds.”After the analysis, Lamarr.AI’s platform generates a report that includes the suspected root cause of every weak point found, an estimated cost to correct that problem, and its estimated return on investment using advanced building energy simulations.“We knew if we were able to quickly, inexpensively, and accurately survey the thermal envelope of buildings and understand their performance, we would be addressing a huge need in the real estate, building construction, and built environment sectors,” Fernández explains. “Thermal anomalies are a huge cause of unwanted heat loss, and more than 45 percent of construction defects are tied to envelope failures.”The ability to operate at scale is especially attractive to building owners and operators, who often manage large portfolios of buildings across multiple campuses.“We see Lamarr.AI becoming the premier solution for building portfolio diagnostics and prognosis across the globe, where every building can be equipped not just for the climate crisis, but also to minimize energy losses and be more efficient, safer, and sustainable,” Rakha says.Building science for everyoneLamarr.AI has worked with building operators across the U.S. as well as in Canada, the United Kingdom, and the United Arab Emirates.In June, Lamarr.AI partnered with the City of Detroit, with support from Newlab and Michigan Central, to inspect three municipal buildings to identify areas for improvement. Across two of the buildings, the system identified more than 460 problems like insulation gaps and water leaks. The findings were presented in a report that also utilized energy simulations to demonstrate that upgrades, such as window replacements and targeted weatherization, could reduce HVAC energy use by up to 22 percent.The entire process took a few days. The founders note that it was the first building inspection drone flight to utilize an off-site operator, an approach that further enhances the scalability of their platform. It also helps further reduce costs, which could make building scans available to a broader swath of people around the world.“We’re democratizing access to very high-value building science expertise that previously cost tens of thousands per audit,” Bayomi says. “Our platform makes advanced diagnostics affordable enough for routine use, not just one-time assessments. The bigger vision is automated, regular building health monitoring that keeps facilities teams informed in real-time, enabling proactive decisions rather than reactive crisis management. When building intelligence becomes continuous and accessible, operators can optimize performance systematically rather than waiting for problems to emerge.” More

    • 50 Shares109 Views

      in Energy

      Where climate meets community

      by

      The MIT Living Climate Futures Lab (LCFL) centers the human dimensions of climate change, bringing together expertise from across MIT to address one of the world’s biggest challenges.The LCFL has three main goals: “addressing how climate change plays out in everyday life, focusing on community-oriented partnerships, and encouraging cross-disciplinary conversations around climate change on campus,” says Chris Walley, the SHASS Dean’s Distinguished Professor of Anthropology and head of MIT’s Anthropology Section. “We think this is a crucial direction for MIT and will make a strong statement about the kind of human-centered, interdisciplinary work needed to tackle this issue.”Walley is faculty lead of LCFL, working in collaboration with a group of 19 faculty colleagues and researchers. The LCFL began to coalesce in 2022 when MIT faculty and affiliates already working with communities dealing with climate change issues organized a symposium, inviting urban farmers, place-based environmental groups, and others to MIT. Since then, the lab has consolidated the efforts of faculty and affiliates representing disciplines from across the MIT School of Humanities, Arts, and Social Sciences (SHASS) and the Institute.Amah Edoh, a cultural anthropologist and managing director of LCFL, says the lab’s collaboration with community organizations and development of experiential learning classes aims to bridge the gap that can exist between the classroom and the real world.“Sometimes we can find ourselves in a bubble where we’re only in conversation with other people from within academia or our own field of practice. There can be a disconnect between what students are learning somewhat abstractly and the ‘real world’ experience of the issues” Edoh says. “By taking up topics from the multidimensional approach that experiential learning makes possible, students learn to take complexity as a given, which can help to foster more critical thinking in them, and inform their future practice in profound ways.”Edoh points out that the effects of climate change play out in a huge array of areas: health, food security, livelihoods, housing, and governance structures, to name a few.“The Living Climate Futures Lab supports MIT researchers in developing the long-term collaborations with community partners that are essential to adequately identifying and responding to the challenges that climate change creates in everyday life,” she says.Manduhai Buyandelger, professor of anthropology and one of the participants in LCFL, developed the class 21A.S01 (Anthro-Engineering: Decarbonization at the Million-Person Scale), which has in turn sparked related classes. The goal is “to merge technological innovation with people-centered environments.” Working closely with residents of Ulaanbaatar, Mongolia, Buyandelger and collaborator Mike Short, the Class of 1941 Professor of Nuclear Science and Engineering, helped develop a molten salt heat bank as a reusable energy source.“My work with Mike Short on energy and alternative heating in Mongolia helps to cultivate a new generation of creative and socially minded engineers who prioritize people in thinking about technical solutions,” Buyandelger says, adding, “In our course, we collaborate on creating interdisciplinary methods where we fuse anthropological methods with engineering innovations so that we can expand and deepen our approach to mitigate climate change.”

      Play video

      MIT Living Climate Futures Lab LaunchVideo: MIT Anthropology

      Iselle Barrios ’25, says 21A.S01 was her first anthropology course. She traveled to Mongolia and was able to experience firsthand all the ways in which the air pollution and heating problem was much larger and more complicated than it seemed from MIT’s Cambridge, Massachusetts, campus.“It was my first exposure to anthropological and STS critiques of science and engineering, as well as international development,” says Barrios, a chemical engineering major. “It fundamentally reshaped the way I see the role of technology and engineers in the broader social context in which they operate. It really helped me learn to think about problems in a more holistic and people-centered way.”LCFL participant Alvin Harvey, a postdoc in the MIT Media Lab’s Space Enabled Research Group and a citizen of the Navajo Nation, works to incorporate traditional knowledge in engineering and science to “support global stewardship of earth and space ecologies.””I envision the Living Climate Futures Lab as a collaborative space that can be an igniter and sustainer of relationships, especially between MIT and those whose have generational and cultural ties to land and space that is being impacted by climate change,” Harvey says. “I think everyone in our lab understands that protecting our climate future is a collective journey.”Kate Brown, the Thomas M. Siebel Distinguished Professor in History of Science, is also a participant in LCFL. Her current interest is urban food sovereignty movements, in which working-class city dwellers used waste to create “the most productive agriculture in recorded human history,” Brown says. While pursuing that work, Brown has developed relationships and worked with urban farmers in Mansfield, Ohio, as well as in Washington and Amsterdam.Brown and Susan Solomon, the Lee and Geraldine Martin Professor of Environmental Studies and Chemistry, teach a class called STS.055 (Living Dangerously: Environmental Programs from 1900 to Today) that presents the environmental problems and solutions of the 20th century, and how some “solutions” created more problems over time. Brown also plans to teach a class on the history of global food production once she gets access to a small plot of land on campus for a lab site.“The Living Climate Futures Lab gives us the structure and flexibility to work with communities that are struggling to find solutions to the problems being created by the climate crisis,” says Brown.Earlier this year, the MIT Human Insight Collaborative (MITHIC) selected the Living Climate Futures Lab as its inaugural Faculty-Driven Initiative (FDI), which comes with a $500,000 seed grant.MIT Provost Anantha Chandrakasan, co-chair of MITHIC, says the LCFL exemplifies how we can confront the climate crisis by working in true partnership with the communities most affected.“By combining scientific insight with cultural understanding and lived experience, this initiative brings a deeper dimension to MIT’s climate efforts — one grounded in collaboration, empathy, and real-world impact,” says Chandrakasan.Agustín Rayo, the Kenan Sahin Dean of SHASS and co-chair of MITHIC, says the LCFL is precisely the type of interdisciplinary collaboration the FDI program was designed to support.”By bringing together expertise from across MIT, I am confident the Living Climate Futures Lab will make significant contributions in the Institute’s effort to address the climate crisis,” says Rayo.Walley said the seed grant will support a second symposium in 2026 to be co-designed with community groups, a suite of experiential learning classes, workshops, a speaker series, and other programming. Throughout this development phase, the lab will solicit donor support to build it into an ongoing MIT initiative and a leader in the response to climate change. More

    • 150 Shares169 Views

      in Environment

      3 Questions: How AI is helping us monitor and support vulnerable ecosystems

      by

      A recent study from Oregon State University estimated that more than 3,500 animal species are at risk of extinction because of factors including habitat alterations, natural resources being overexploited, and climate change.To better understand these changes and protect vulnerable wildlife, conservationists like MIT PhD student and Computer Science and Artificial Intelligence Laboratory (CSAIL) researcher Justin Kay are developing computer vision algorithms that carefully monitor animal populations. A member of the lab of MIT Department of Electrical Engineering and Computer Science assistant professor and CSAIL principal investigator Sara Beery, Kay is currently working on tracking salmon in the Pacific Northwest, where they provide crucial nutrients to predators like birds and bears, while managing the population of prey, like bugs.With all that wildlife data, though, researchers have lots of information to sort through and many AI models to choose from to analyze it all. Kay and his colleagues at CSAIL and the University of Massachusetts Amherst are developing AI methods that make this data-crunching process much more efficient, including a new approach called “consensus-driven active model selection” (or “CODA”) that helps conservationists choose which AI model to use. Their work was named a Highlight Paper at the International Conference on Computer Vision (ICCV) in October.That research was supported, in part, by the National Science Foundation, Natural Sciences and Engineering Research Council of Canada, and Abdul Latif Jameel Water and Food Systems Lab (J-WAFS). Here, Kay discusses this project, among other conservation efforts.Q: In your paper, you pose the question of which AI models will perform the best on a particular dataset. With as many as 1.9 million pre-trained models available in the HuggingFace Models repository alone, how does CODA help us address that challenge?A: Until recently, using AI for data analysis has typically meant training your own model. This requires significant effort to collect and annotate a representative training dataset, as well as iteratively train and validate models. You also need a certain technical skill set to run and modify AI training code. The way people interact with AI is changing, though — in particular, there are now millions of publicly available pre-trained models that can perform a variety of predictive tasks very well. This potentially enables people to use AI to analyze their data without developing their own model, simply by downloading an existing model with the capabilities they need. But this poses a new challenge: Which model, of the millions available, should they use to analyze their data? Typically, answering this model selection question also requires you to spend a lot of time collecting and annotating a large dataset, albeit for testing models rather than training them. This is especially true for real applications where user needs are specific, data distributions are imbalanced and constantly changing, and model performance may be inconsistent across samples. Our goal with CODA was to substantially reduce this effort. We do this by making the data annotation process “active.” Instead of requiring users to bulk-annotate a large test dataset all at once, in active model selection we make the process interactive, guiding users to annotate the most informative data points in their raw data. This is remarkably effective, often requiring users to annotate as few as 25 examples to identify the best model from their set of candidates. We’re very excited about CODA offering a new perspective on how to best utilize human effort in the development and deployment of machine-learning (ML) systems. As AI models become more commonplace, our work emphasizes the value of focusing effort on robust evaluation pipelines, rather than solely on training.Q: You applied the CODA method to classifying wildlife in images. Why did it perform so well, and what role can systems like this have in monitoring ecosystems in the future?A: One key insight was that when considering a collection of candidate AI models, the consensus of all of their predictions is more informative than any individual model’s predictions. This can be seen as a sort of “wisdom of the crowd:” On average, pooling the votes of all models gives you a decent prior over what the labels of individual data points in your raw dataset should be. Our approach with CODA is based on estimating a “confusion matrix” for each AI model — given the true label for some data point is class X, what is the probability that an individual model predicts class X, Y, or Z? This creates informative dependencies between all of the candidate models, the categories you want to label, and the unlabeled points in your dataset.Consider an example application where you are a wildlife ecologist who has just collected a dataset containing potentially hundreds of thousands of images from cameras deployed in the wild. You want to know what species are in these images, a time-consuming task that computer vision classifiers can help automate. You are trying to decide which species classification model to run on your data. If you have labeled 50 images of tigers so far, and some model has performed well on those 50 images, you can be pretty confident it will perform well on the remainder of the (currently unlabeled) images of tigers in your raw dataset as well. You also know that when that model predicts some image contains a tiger, it is likely to be correct, and therefore that any model that predicts a different label for that image is more likely to be wrong. You can use all these interdependencies to construct probabilistic estimates of each model’s confusion matrix, as well as a probability distribution over which model has the highest accuracy on the overall dataset. These design choices allow us to make more informed choices over which data points to label and ultimately are the reason why CODA performs model selection much more efficiently than past work.There are also a lot of exciting possibilities for building on top of our work. We think there may be even better ways of constructing informative priors for model selection based on domain expertise — for instance, if it is already known that one model performs exceptionally well on some subset of classes or poorly on others. There are also opportunities to extend the framework to support more complex machine-learning tasks and more sophisticated probabilistic models of performance. We hope our work can provide inspiration and a starting point for other researchers to keep pushing the state of the art.Q: You work in the Beerylab, led by Sara Beery, where researchers are combining the pattern-recognition capabilities of machine-learning algorithms with computer vision technology to monitor wildlife. What are some other ways your team is tracking and analyzing the natural world, beyond CODA?A: The lab is a really exciting place to work, and new projects are emerging all the time. We have ongoing projects monitoring coral reefs with drones, re-identifying individual elephants over time, and fusing multi-modal Earth observation data from satellites and in-situ cameras, just to name a few. Broadly, we look at emerging technologies for biodiversity monitoring and try to understand where the data analysis bottlenecks are, and develop new computer vision and machine-learning approaches that address those problems in a widely applicable way. It’s an exciting way of approaching problems that sort of targets the “meta-questions” underlying particular data challenges we face. The computer vision algorithms I’ve worked on that count migrating salmon in underwater sonar video are examples of that work. We often deal with shifting data distributions, even as we try to construct the most diverse training datasets we can. We always encounter something new when we deploy a new camera, and this tends to degrade the performance of computer vision algorithms. This is one instance of a general problem in machine learning called domain adaptation, but when we tried to apply existing domain adaptation algorithms to our fisheries data we realized there were serious limitations in how existing algorithms were trained and evaluated. We were able to develop a new domain adaptation framework, published earlier this year in Transactions on Machine Learning Research, that addressed these limitations and led to advancements in fish counting, and even self-driving and spacecraft analysis.One line of work that I’m particularly excited about is understanding how to better develop and analyze the performance of predictive ML algorithms in the context of what they are actually used for. Usually, the outputs from some computer vision algorithm — say, bounding boxes around animals in images — are not actually the thing that people care about, but rather a means to an end to answer a larger problem — say, what species live here, and how is that changing over time? We have been working on methods to analyze predictive performance in this context and reconsider the ways that we input human expertise into ML systems with this in mind. CODA was one example of this, where we showed that we could actually consider the ML models themselves as fixed and build a statistical framework to understand their performance very efficiently. We have been working recently on similar integrated analyses combining ML predictions with multi-stage prediction pipelines, as well as ecological statistical models. The natural world is changing at unprecedented rates and scales, and being able to quickly move from scientific hypotheses or management questions to data-driven answers is more important than ever for protecting ecosystems and the communities that depend on them. Advancements in AI can play an important role, but we need to think critically about the ways that we design, train, and evaluate algorithms in the context of these very real challenges. More