Machine learning

Subterms

    More stories

    • 150 Shares169 Views

      in Environment

      Detailed images from space offer clearer picture of drought effects on plants

      by

      “MIT is a place where dreams come true,” says César Terrer, an assistant professor in the Department of Civil and Environmental Engineering. Here at MIT, Terrer says he’s given the resources needed to explore ideas he finds most exciting, and at the top of his list is climate science. In particular, he is interested in plant-soil interactions, and how the two can mitigate impacts of climate change. In 2022, Terrer received seed grant funding from the Abdul Latif Jameel Water and Food Systems Lab (J-WAFS) to produce drought monitoring systems for farmers. The project is leveraging a new generation of remote sensing devices to provide high-resolution plant water stress at regional to global scales.

      Growing up in Granada, Spain, Terrer always had an aptitude and passion for science. He studied environmental science at the University of Murcia, where he interned in the Department of Ecology. Using computational analysis tools, he worked on modeling species distribution in response to human development. Early on in his undergraduate experience, Terrer says he regarded his professors as “superheroes” with a kind of scholarly prowess. He knew he wanted to follow in their footsteps by one day working as a faculty member in academia. Of course, there would be many steps along the way before achieving that dream. 

      Upon completing his undergraduate studies, Terrer set his sights on exciting and adventurous research roles. He thought perhaps he would conduct field work in the Amazon, engaging with native communities. But when the opportunity arose to work in Australia on a state-of-the-art climate change experiment that simulates future levels of carbon dioxide, he headed south to study how plants react to CO2 in a biome of native Australian eucalyptus trees. It was during this experience that Terrer started to take a keen interest in the carbon cycle and the capacity of ecosystems to buffer rising levels of CO2 caused by human activity.

      Around 2014, he began to delve deeper into the carbon cycle as he began his doctoral studies at Imperial College London. The primary question Terrer sought to answer during his PhD was “will plants be able to absorb predicted future levels of CO2 in the atmosphere?” To answer the question, Terrer became an early adopter of artificial intelligence, machine learning, and remote sensing to analyze data from real-life, global climate change experiments. His findings from these “ground truth” values and observations resulted in a paper in the journal Science. In it, he claimed that climate models most likely overestimated how much carbon plants will be able to absorb by the end of the century, by a factor of three. 

      After postdoctoral positions at Stanford University and the Universitat Autonoma de Barcelona, followed by a prestigious Lawrence Fellowship, Terrer says he had “too many ideas and not enough time to accomplish all those ideas.” He knew it was time to lead his own group. Not long after applying for faculty positions, he landed at MIT. 

      New ways to monitor drought

      Terrer is employing similar methods to those he used during his PhD to analyze data from all over the world for his J-WAFS project. He and postdoc Wenzhe Jiao collect data from remote sensing satellites and field experiments and use machine learning to come up with new ways to monitor drought. Terrer says Jiao is a “remote sensing wizard,” who fuses data from different satellite products to understand the water cycle. With Jiao’s hydrology expertise and Terrer’s knowledge of plants, soil, and the carbon cycle, the duo is a formidable team to tackle this project.

      According to the U.N. World Meteorological Organization, the number and duration of droughts has increased by 29 percent since 2000, as compared to the two previous decades. From the Horn of Africa to the Western United States, drought is devastating vegetation and severely stressing water supplies, compromising food production and spiking food insecurity. Drought monitoring can offer fundamental information on drought location, frequency, and severity, but assessing the impact of drought on vegetation is extremely challenging. This is because plants’ sensitivity to water deficits varies across species and ecosystems. 

      Terrer and Jiao are able to obtain a clearer picture of how drought is affecting plants by employing the latest generation of remote sensing observations, which offer images of the planet with incredible spatial and temporal resolution. Satellite products such as Sentinel, Landsat, and Planet can provide daily images from space with such high resolution that individual trees can be discerned. Along with the images and datasets from satellites, the team is using ground-based observations from meteorological data. They are also using the MIT SuperCloud at MIT Lincoln Laboratory to process and analyze all of the data sets. The J-WAFS project is among one of the first to leverage high-resolution data to quantitatively measure plant drought impacts in the United States with the hopes of expanding to a global assessment in the future.

      Assisting farmers and resource managers 

      Every week, the U.S. Drought Monitor provides a map of drought conditions in the United States. The map has zero resolution and is more of a drought recap or summary, unable to predict future drought scenarios. The lack of a comprehensive spatiotemporal evaluation of historic and future drought impacts on global vegetation productivity is detrimental to farmers both in the United States and worldwide.  

      Terrer and Jiao plan to generate metrics for plant water stress at an unprecedented resolution of 10-30 meters. This means that they will be able to provide drought monitoring maps at the scale of a typical U.S. farm, giving farmers more precise, useful data every one to two days. The team will use the information from the satellites to monitor plant growth and soil moisture, as well as the time lag of plant growth response to soil moisture. In this way, Terrer and Jiao say they will eventually be able to create a kind of “plant water stress forecast” that may be able to predict adverse impacts of drought four weeks in advance. “According to the current soil moisture and lagged response time, we hope to predict plant water stress in the future,” says Jiao. 

      The expected outcomes of this project will give farmers, land and water resource managers, and decision-makers more accurate data at the farm-specific level, allowing for better drought preparation, mitigation, and adaptation. “We expect to make our data open-access online, after we finish the project, so that farmers and other stakeholders can use the maps as tools,” says Jiao. 

      Terrer adds that the project “has the potential to help us better understand the future states of climate systems, and also identify the regional hot spots more likely to experience water crises at the national, state, local, and tribal government scales.” He also expects the project will enhance our understanding of global carbon-water-energy cycle responses to drought, with applications in determining climate change impacts on natural ecosystems as a whole. More

    • 175 Shares189 Views

      in Energy

      Computers that power self-driving cars could be a huge driver of global carbon emissions

      by

      In the future, the energy needed to run the powerful computers on board a global fleet of autonomous vehicles could generate as many greenhouse gas emissions as all the data centers in the world today.

      That is one key finding of a new study from MIT researchers that explored the potential energy consumption and related carbon emissions if autonomous vehicles are widely adopted.

      The data centers that house the physical computing infrastructure used for running applications are widely known for their large carbon footprint: They currently account for about 0.3 percent of global greenhouse gas emissions, or about as much carbon as the country of Argentina produces annually, according to the International Energy Agency. Realizing that less attention has been paid to the potential footprint of autonomous vehicles, the MIT researchers built a statistical model to study the problem. They determined that 1 billion autonomous vehicles, each driving for one hour per day with a computer consuming 840 watts, would consume enough energy to generate about the same amount of emissions as data centers currently do.

      The researchers also found that in over 90 percent of modeled scenarios, to keep autonomous vehicle emissions from zooming past current data center emissions, each vehicle must use less than 1.2 kilowatts of power for computing, which would require more efficient hardware. In one scenario — where 95 percent of the global fleet of vehicles is autonomous in 2050, computational workloads double every three years, and the world continues to decarbonize at the current rate — they found that hardware efficiency would need to double faster than every 1.1 years to keep emissions under those levels.

      “If we just keep the business-as-usual trends in decarbonization and the current rate of hardware efficiency improvements, it doesn’t seem like it is going to be enough to constrain the emissions from computing onboard autonomous vehicles. This has the potential to become an enormous problem. But if we get ahead of it, we could design more efficient autonomous vehicles that have a smaller carbon footprint from the start,” says first author Soumya Sudhakar, a graduate student in aeronautics and astronautics.

      Sudhakar wrote the paper with her co-advisors Vivienne Sze, associate professor in the Department of Electrical Engineering and Computer Science (EECS) and a member of the Research Laboratory of Electronics (RLE); and Sertac Karaman, associate professor of aeronautics and astronautics and director of the Laboratory for Information and Decision Systems (LIDS). The research appears today in the January-February issue of IEEE Micro.

      Modeling emissions

      The researchers built a framework to explore the operational emissions from computers on board a global fleet of electric vehicles that are fully autonomous, meaning they don’t require a back-up human driver.

      The model is a function of the number of vehicles in the global fleet, the power of each computer on each vehicle, the hours driven by each vehicle, and the carbon intensity of the electricity powering each computer.

      “On its own, that looks like a deceptively simple equation. But each of those variables contains a lot of uncertainty because we are considering an emerging application that is not here yet,” Sudhakar says.

      For instance, some research suggests that the amount of time driven in autonomous vehicles might increase because people can multitask while driving and the young and the elderly could drive more. But other research suggests that time spent driving might decrease because algorithms could find optimal routes that get people to their destinations faster.

      In addition to considering these uncertainties, the researchers also needed to model advanced computing hardware and software that doesn’t exist yet.

      To accomplish that, they modeled the workload of a popular algorithm for autonomous vehicles, known as a multitask deep neural network because it can perform many tasks at once. They explored how much energy this deep neural network would consume if it were processing many high-resolution inputs from many cameras with high frame rates, simultaneously.

      When they used the probabilistic model to explore different scenarios, Sudhakar was surprised by how quickly the algorithms’ workload added up.

      For example, if an autonomous vehicle has 10 deep neural networks processing images from 10 cameras, and that vehicle drives for one hour a day, it will make 21.6 million inferences each day. One billion vehicles would make 21.6 quadrillion inferences. To put that into perspective, all of Facebook’s data centers worldwide make a few trillion inferences each day (1 quadrillion is 1,000 trillion).

      “After seeing the results, this makes a lot of sense, but it is not something that is on a lot of people’s radar. These vehicles could actually be using a ton of computer power. They have a 360-degree view of the world, so while we have two eyes, they may have 20 eyes, looking all over the place and trying to understand all the things that are happening at the same time,” Karaman says.

      Autonomous vehicles would be used for moving goods, as well as people, so there could be a massive amount of computing power distributed along global supply chains, he says. And their model only considers computing — it doesn’t take into account the energy consumed by vehicle sensors or the emissions generated during manufacturing.

      Keeping emissions in check

      To keep emissions from spiraling out of control, the researchers found that each autonomous vehicle needs to consume less than 1.2 kilowatts of energy for computing. For that to be possible, computing hardware must become more efficient at a significantly faster pace, doubling in efficiency about every 1.1 years.

      One way to boost that efficiency could be to use more specialized hardware, which is designed to run specific driving algorithms. Because researchers know the navigation and perception tasks required for autonomous driving, it could be easier to design specialized hardware for those tasks, Sudhakar says. But vehicles tend to have 10- or 20-year lifespans, so one challenge in developing specialized hardware would be to “future-proof” it so it can run new algorithms.

      In the future, researchers could also make the algorithms more efficient, so they would need less computing power. However, this is also challenging because trading off some accuracy for more efficiency could hamper vehicle safety.

      Now that they have demonstrated this framework, the researchers want to continue exploring hardware efficiency and algorithm improvements. In addition, they say their model can be enhanced by characterizing embodied carbon from autonomous vehicles — the carbon emissions generated when a car is manufactured — and emissions from a vehicle’s sensors.

      While there are still many scenarios to explore, the researchers hope that this work sheds light on a potential problem people may not have considered.

      “We are hoping that people will think of emissions and carbon efficiency as important metrics to consider in their designs. The energy consumption of an autonomous vehicle is really critical, not just for extending the battery life, but also for sustainability,” says Sze.

      This research was funded, in part, by the National Science Foundation and the MIT-Accenture Fellowship. More

    • 50 Shares149 Views

      in Energy

      Strengthening electron-triggered light emission

      by

      The way electrons interact with photons of light is a key part of many modern technologies, from lasers to solar panels to LEDs. But the interaction is inherently a weak one because of a major mismatch in scale: A wavelength of visible light is about 1,000 times larger than an electron, so the way the two things affect each other is limited by that disparity.

      Now, researchers at MIT and elsewhere have come up with an innovative way to make much stronger interactions between photons and electrons possible, in the process producing a hundredfold increase in the emission of light from a phenomenon called Smith-Purcell radiation. The finding has potential implications for both commercial applications and fundamental scientific research, although it will require more years of research to make it practical.

      The findings are reported today in the journal Nature, in a paper by MIT postdocs Yi Yang (now an assistant professor at the University of Hong Kong) and Charles Roques-Carmes, MIT professors Marin Soljačić and John Joannopoulos, and five others at MIT, Harvard University, and Technion-Israel Institute of Technology.

      In a combination of computer simulations and laboratory experiments, the team found that using a beam of electrons in combination with a specially designed photonic crystal — a slab of silicon on an insulator, etched with an array of nanometer-scale holes — they could theoretically predict stronger emission by many orders of magnitude than would ordinarily be possible in conventional Smith-Purcell radiation. They also experimentally recorded a one hundredfold increase in radiation in their proof-of-concept measurements.

      Unlike other approaches to producing sources of light or other electromagnetic radiation, the free-electron-based method is fully tunable — it can produce emissions of any desired wavelength, simply by adjusting the size of the photonic structure and the speed of the electrons. This may make it especially valuable for making sources of emission at wavelengths that are difficult to produce efficiently, including terahertz waves, ultraviolet light, and X-rays.

      The team has so far demonstrated the hundredfold enhancement in emission using a repurposed electron microscope to function as an electron beam source. But they say that the basic principle involved could potentially enable far greater enhancements using devices specifically adapted for this function.

      The approach is based on a concept called flatbands, which have been widely explored in recent years for condensed matter physics and photonics but have never been applied to affecting the basic interaction of photons and free electrons. The underlying principle involves the transfer of momentum from the electron to a group of photons, or vice versa. Whereas conventional light-electron interactions rely on producing light at a single angle, the photonic crystal is tuned in such a way that it enables the production of a whole range of angles.

      The same process could also be used in the opposite direction, using resonant light waves to propel electrons, increasing their velocity in a way that could potentially be harnessed to build miniaturized particle accelerators on a chip. These might ultimately be able to perform some functions that currently require giant underground tunnels, such as the 30-kilometer-wide Large Hadron Collider in Switzerland.

      “If you could actually build electron accelerators on a chip,” Soljačić says, “you could make much more compact accelerators for some of the applications of interest, which would still produce very energetic electrons. That obviously would be huge. For many applications, you wouldn’t have to build these huge facilities.”

      The new system could also potentially provide a highly controllable X-ray beam for radiotherapy purposes, Roques-Carmes says.

      And the system could be used to generate multiple entangled photons, a quantum effect that could be useful in the creation of quantum-based computational and communications systems, the researchers say. “You can use electrons to couple many photons together, which is a considerably hard problem if using a purely optical approach,” says Yang. “That is one of the most exciting future directions of our work.”

      Much work remains to translate these new findings into practical devices, Soljačić cautions. It may take some years to develop the necessary interfaces between the optical and electronic components and how to connect them on a single chip, and to develop the necessary on-chip electron source producing a continuous wavefront, among other challenges.

      “The reason this is exciting,” Roques-Carmes adds, “is because this is quite a different type of source.” While most technologies for generating light are restricted to very specific ranges of color or wavelength, and “it’s usually difficult to move that emission frequency. Here it’s completely tunable. Simply by changing the velocity of the electrons, you can change the emission frequency. … That excites us about the potential of these sources. Because they’re different, they offer new types of opportunities.”

      But, Soljačić concludes, “in order for them to become truly competitive with other types of sources, I think it will require some more years of research. I would say that with some serious effort, in two to five years they might start competing in at least some areas of radiation.”

      The research team also included Steven Kooi at MIT’s Institute for Soldier Nanotechnologies, Haoning Tang and Eric Mazur at Harvard University, Justin Beroz at MIT, and Ido Kaminer at Technion-Israel Institute of Technology. The work was supported by the U.S. Army Research Office through the Institute for Soldier Nanotechnologies, the U.S. Air Force Office of Scientific Research, and the U.S. Office of Naval Research. More

    • 100 Shares129 Views

      in Energy

      Machine learning facilitates “turbulence tracking” in fusion reactors

      by

      Fusion, which promises practically unlimited, carbon-free energy using the same processes that power the sun, is at the heart of a worldwide research effort that could help mitigate climate change.

      A multidisciplinary team of researchers is now bringing tools and insights from machine learning to aid this effort. Scientists from MIT and elsewhere have used computer-vision models to identify and track turbulent structures that appear under the conditions needed to facilitate fusion reactions.

      Monitoring the formation and movements of these structures, called filaments or “blobs,” is important for understanding the heat and particle flows exiting from the reacting fuel, which ultimately determines the engineering requirements for the reactor walls to meet those flows. However, scientists typically study blobs using averaging techniques, which trade details of individual structures in favor of aggregate statistics. Individual blob information must be tracked by marking them manually in video data. 

      The researchers built a synthetic video dataset of plasma turbulence to make this process more effective and efficient. They used it to train four computer vision models, each of which identifies and tracks blobs. They trained the models to pinpoint blobs in the same ways that humans would.

      When the researchers tested the trained models using real video clips, the models could identify blobs with high accuracy — more than 80 percent in some cases. The models were also able to effectively estimate the size of blobs and the speeds at which they moved.

      Because millions of video frames are captured during just one fusion experiment, using machine-learning models to track blobs could give scientists much more detailed information.

      “Before, we could get a macroscopic picture of what these structures are doing on average. Now, we have a microscope and the computational power to analyze one event at a time. If we take a step back, what this reveals is the power available from these machine-learning techniques, and ways to use these computational resources to make progress,” says Theodore Golfinopoulos, a research scientist at the MIT Plasma Science and Fusion Center and co-author of a paper detailing these approaches.

      His fellow co-authors include lead author Woonghee “Harry” Han, a physics PhD candidate; senior author Iddo Drori, a visiting professor in the Computer Science and Artificial Intelligence Laboratory (CSAIL), faculty associate professor at Boston University, and adjunct at Columbia University; as well as others from the MIT Plasma Science and Fusion Center, the MIT Department of Civil and Environmental Engineering, and the Swiss Federal Institute of Technology at Lausanne in Switzerland. The research appears today in Nature Scientific Reports.

      Heating things up

      For more than 70 years, scientists have sought to use controlled thermonuclear fusion reactions to develop an energy source. To reach the conditions necessary for a fusion reaction, fuel must be heated to temperatures above 100 million degrees Celsius. (The core of the sun is about 15 million degrees Celsius.)

      A common method for containing this super-hot fuel, called plasma, is to use a tokamak. These devices utilize extremely powerful magnetic fields to hold the plasma in place and control the interaction between the exhaust heat from the plasma and the reactor walls.

      However, blobs appear like filaments falling out of the plasma at the very edge, between the plasma and the reactor walls. These random, turbulent structures affect how energy flows between the plasma and the reactor.

      “Knowing what the blobs are doing strongly constrains the engineering performance that your tokamak power plant needs at the edge,” adds Golfinopoulos.

      Researchers use a unique imaging technique to capture video of the plasma’s turbulent edge during experiments. An experimental campaign may last months; a typical day will produce about 30 seconds of data, corresponding to roughly 60 million video frames, with thousands of blobs appearing each second. This makes it impossible to track all blobs manually, so researchers rely on average sampling techniques that only provide broad characteristics of blob size, speed, and frequency.

      “On the other hand, machine learning provides a solution to this by blob-by-blob tracking for every frame, not just average quantities. This gives us much more knowledge about what is happening at the boundary of the plasma,” Han says.

      He and his co-authors took four well-established computer vision models, which are commonly used for applications like autonomous driving, and trained them to tackle this problem.

      Simulating blobs

      To train these models, they created a vast dataset of synthetic video clips that captured the blobs’ random and unpredictable nature.

      “Sometimes they change direction or speed, sometimes multiple blobs merge, or they split apart. These kinds of events were not considered before with traditional approaches, but we could freely simulate those behaviors in the synthetic data,” Han says.

      Creating synthetic data also allowed them to label each blob, which made the training process more effective, Drori adds.

      Using these synthetic data, they trained the models to draw boundaries around blobs, teaching them to closely mimic what a human scientist would draw.

      Then they tested the models using real video data from experiments. First, they measured how closely the boundaries the models drew matched up with actual blob contours.

      But they also wanted to see if the models predicted objects that humans would identify. They asked three human experts to pinpoint the centers of blobs in video frames and checked to see if the models predicted blobs in those same locations.

      The models were able to draw accurate blob boundaries, overlapping with brightness contours which are considered ground-truth, about 80 percent of the time. Their evaluations were similar to those of human experts, and successfully predicted the theory-defined regime of the blob, which agrees with the results from a traditional method.

      Now that they have shown the success of using synthetic data and computer vision models for tracking blobs, the researchers plan to apply these techniques to other problems in fusion research, such as estimating particle transport at the boundary of a plasma, Han says.

      They also made the dataset and models publicly available, and look forward to seeing how other research groups apply these tools to study the dynamics of blobs, says Drori.

      “Prior to this, there was a barrier to entry that mostly the only people working on this problem were plasma physicists, who had the datasets and were using their methods. There is a huge machine-learning and computer-vision community. One goal of this work is to encourage participation in fusion research from the broader machine-learning community toward the broader goal of helping solve the critical problem of climate change,” he adds.

      This research is supported, in part, by the U.S. Department of Energy and the Swiss National Science Foundation. More

    • 88 Shares99 Views

      in Energy

      Computing for the health of the planet

      by

      The health of the planet is one of the most important challenges facing humankind today. From climate change to unsafe levels of air and water pollution to coastal and agricultural land erosion, a number of serious challenges threaten human and ecosystem health.

      Ensuring the health and safety of our planet necessitates approaches that connect scientific, engineering, social, economic, and political aspects. New computational methods can play a critical role by providing data-driven models and solutions for cleaner air, usable water, resilient food, efficient transportation systems, better-preserved biodiversity, and sustainable sources of energy.

      The MIT Schwarzman College of Computing is committed to hiring multiple new faculty in computing for climate and the environment, as part of MIT’s plan to recruit 20 climate-focused faculty under its climate action plan. This year the college undertook searches with several departments in the schools of Engineering and Science for shared faculty in computing for health of the planet, one of the six strategic areas of inquiry identified in an MIT-wide planning process to help focus shared hiring efforts. The college also undertook searches for core computing faculty in the Department of Electrical Engineering and Computer Science (EECS).

      The searches are part of an ongoing effort by the MIT Schwarzman College of Computing to hire 50 new faculty — 25 shared with other academic departments and 25 in computer science and artificial intelligence and decision-making. The goal is to build capacity at MIT to help more deeply infuse computing and other disciplines in departments.

      Four interdisciplinary scholars were hired in these searches. They will join the MIT faculty in the coming year to engage in research and teaching that will advance physical understanding of low-carbon energy solutions, Earth-climate modeling, biodiversity monitoring and conservation, and agricultural management through high-performance computing, transformational numerical methods, and machine-learning techniques.

      “By coordinating hiring efforts with multiple departments and schools, we were able to attract a cohort of exceptional scholars in this area to MIT. Each of them is developing and using advanced computational methods and tools to help find solutions for a range of climate and environmental issues,” says Daniel Huttenlocher, dean of the MIT Schwarzman College of Computing and the Henry Warren Ellis Professor of Electrical Engineering and Computer Science. “They will also help strengthen cross-departmental ties in computing across an important, critical area for MIT and the world.”

      “These strategic hires in the area of computing for climate and the environment are an incredible opportunity for the college to deepen its academic offerings and create new opportunity for collaboration across MIT,” says Anantha P. Chandrakasan, dean of the MIT School of Engineering and the Vannevar Bush Professor of Electrical Engineering and Computer Science. “The college plays a pivotal role in MIT’s overarching effort to hire climate-focused faculty — introducing the critical role of computing to address the health of the planet through innovative research and curriculum.”

      The four new faculty members are:

      Sara Beery will join MIT as an assistant professor in the Faculty of Artificial Intelligence and Decision-Making in EECS in September 2023. Beery received her PhD in computing and mathematical sciences at Caltech in 2022, where she was advised by Pietro Perona. Her research focuses on building computer vision methods that enable global-scale environmental and biodiversity monitoring across data modalities, tackling real-world challenges including strong spatiotemporal correlations, imperfect data quality, fine-grained categories, and long-tailed distributions. She partners with nongovernmental organizations and government agencies to deploy her methods in the wild worldwide and works toward increasing the diversity and accessibility of academic research in artificial intelligence through interdisciplinary capacity building and education.

      Priya Donti will join MIT as an assistant professor in the faculties of Electrical Engineering and Artificial Intelligence and Decision-Making in EECS in academic year 2023-24. Donti recently finished her PhD in the Computer Science Department and the Department of Engineering and Public Policy at Carnegie Mellon University, co-advised by Zico Kolter and Inês Azevedo. Her work focuses on machine learning for forecasting, optimization, and control in high-renewables power grids. Specifically, her research explores methods to incorporate the physics and hard constraints associated with electric power systems into deep learning models. Donti is also co-founder and chair of Climate Change AI, a nonprofit initiative to catalyze impactful work at the intersection of climate change and machine learning that is currently running through the Cornell Tech Runway Startup Postdoc Program.

      Ericmoore Jossou will join MIT as an assistant professor in a shared position between the Department of Nuclear Science and Engineering and the faculty of electrical engineering in EECS in July 2023. He is currently an assistant scientist at the Brookhaven National Laboratory, a U.S. Department of Energy-affiliated lab that conducts research in nuclear and high energy physics, energy science and technology, environmental and bioscience, nanoscience, and national security. His research at MIT will focus on understanding the processing-structure-properties correlation of materials for nuclear energy applications through advanced experiments, multiscale simulations, and data science. Jossou obtained his PhD in mechanical engineering in 2019 from the University of Saskatchewan.

      Sherrie Wang will join MIT as an assistant professor in a shared position between the Department of Mechanical Engineering and the Institute for Data, Systems, and Society in academic year 2023-24. Wang is currently a Ciriacy-Wantrup Postdoctoral Fellow at the University of California at Berkeley, hosted by Solomon Hsiang and the Global Policy Lab. She develops machine learning for Earth observation data. Her primary application areas are improving agricultural management and forecasting climate phenomena. She obtained her PhD in computational and mathematical engineering from Stanford University in 2021, where she was advised by David Lobell. More

    • 213 Shares199 Views

      in Security

      J-WAFS awards $150K Solutions grant to Patrick Doyle and team for rapid removal of micropollutants from water

      by

      The Abdul Latif Jameel Water and Food Systems Lab (J-WAFS) has awarded a 2022 J-WAFS Solutions grant to Patrick S. Doyle, the Robert T. Haslam Professor of Chemical Engineering at MIT, for his innovative system to tackle water pollution. Doyle will be working with co-Principal Investigator Rafael Gomez-Bombarelli, assistant professor in materials processing in the Department of Materials Science, as well as PhD students Devashish Gokhale and Tynan Perez. Building off of findings from a 2019 J-WAFS seed grant, Doyle and the research team will create cost-effective industry-scale processes to remove micropollutants from water. Project work will commence this month.

      The J-WAFS Solutions program provides one-year, renewable, commercialization grants to help move MIT technology from the laboratory to market. Grants of up to $150,000 are awarded to researchers with breakthrough technologies and inventions in water or food. Since its launch in 2015, J-WAFS Solutions grants have led to seven spinout companies and helped commercialize two products as open-source technologies. The grant program is supported by Community Jameel.

      A widespread problem 

      Micropollutants are contaminants that occur in low concentrations in the environment, yet continuous exposure and bioaccumulation of micropollutants make them a cause for concern. According to the U.S. Environmental Protection Agency, the plastics derivative Bisphenol A (BPA), the “forever chemicals” per-and polyfluoroalkyl substances (PFAS), and heavy metals like lead are common micropollutants known to be found in more than 85 percent of rivers, ponds, and lakes in the United States. Many of these bodies of water are sources of drinking water. Over long periods of time, exposure to micropollutants through drinking water can cause physiological damage in humans, increasing the risk of cancer, developmental disorders, and reproductive failure.

      Since micropollutants occur in low concentrations, it is difficult to detect and monitor their presence, and the chemical diversity of micropollutants makes it difficult to inexpensively remove them from water. Currently, activated carbon is the industry standard for micropollutant elimination, but this method cannot efficiently remove contaminants at parts-per-billion and parts-per-trillion concentrations. There are also strong sustainability concerns associated with activated carbon production, which is energy-intensive and releases large volumes of carbon dioxide.

      A solution with societal and economic benefits

      Doyle and his team are developing a technology that uses sustainable hydrogel microparticles to remove micropollutants from water. The polymeric hydrogel microparticles use chemically anchored structures including micelles and other chelating agents that act like a sponge by absorbing organic micropollutants and heavy metal ions. The microparticles are large enough to separate from water using simple gravitational settling. The system is sustainable because the microparticles can be recycled for continuous use. In testing, the long-lasting, reusable microparticles show quicker removal of contaminants than commercial activated carbon. The researchers plan to utilize machine learning to find optimal microparticle compositions that maximize performance on complex combinations of micropollutants in simulated and real wastewater samples.

      Economically, the technology is a new offering that has applications in numerous large markets where micropollutant elimination is vital, including municipal and industrial water treatment equipment, as well as household water purification systems. The J-WAFS Solutions grant will allow the team to build and test prototypes of the water treatment system, identify the best use cases and customers, and perform technoeconomic analyses and market research to formulate a preliminary business plan. With J-WAFS commercialization support, the project could eventually lead to a startup company.

      “Emerging micropollutants are a growing threat to drinking water supplies worldwide,” says J-WAFS Director John H. Lienhard, the Abdul Latif Jameel Professor of Water at MIT. “Cost-effective and scalable technologies for micropollutant removal are urgently needed. This project will develop and commercialize a promising new tool for water treatment, with the goal of improving water quality for millions of people.” More

    • 100 Shares99 Views

      in Energy

      Designing zeolites, porous materials made to trap molecules

      by

      Zeolites are a class of minerals used in everything from industrial catalysts and chemical filters to laundry detergents and cat litter. They are mostly composed of silicon and aluminum — two abundant, inexpensive elements — plus oxygen; they have a crystalline structure; and most significantly, they are porous. Among the regularly repeating atomic patterns in them are tiny interconnected openings, or pores, that can trap molecules that just fit inside them, allow smaller ones to pass through, or block larger ones from entering. A zeolite can remove unwanted molecules from gases and liquids, or trap them temporarily and then release them, or hold them while they undergo rapid chemical reactions.

      Some zeolites occur naturally, but they take unpredictable forms and have variable-sized pores. “People synthesize artificial versions to ensure absolute purity and consistency,” says Rafael Gómez-Bombarelli, the Jeffrey Cheah Career Development Chair in Engineering in the Department of Materials Science and Engineering (DMSE). And they work hard to influence the size of the internal pores in hopes of matching the molecule or other particle they’re looking to capture.

      The basic recipe for making zeolites sounds simple. Mix together the raw ingredients — basically, silicon dioxide and aluminum oxide — and put them in a reactor for a few days at a high temperature and pressure. Depending on the ratio between the ingredients and the temperature, pressure, and timing, as the initial gel slowly solidifies into crystalline form, different zeolites emerge.

      But there’s one special ingredient to add “to help the system go where you want it to go,” says Gómez-Bombarelli. “It’s a molecule that serves as a template so that the zeolite you want will crystallize around it and create pores of the desired size and shape.”

      The so-called templating molecule binds to the material before it solidifies. As crystallization progresses, the molecule directs the structure, or “framework,” that forms around it. After crystallization, the temperature is raised and the templating molecule burns off, leaving behind a solid aluminosilicate material filled with open pores that are — given the correct templating molecule and synthesis conditions — just the right size and shape to recognize the targeted molecule.

      The zeolite conundrum

      Theoretical studies suggest that there should be hundreds of thousands of possible zeolites. But despite some 60 years of intensive research, only about 250 zeolites have been made. This is sometimes called the “zeolite conundrum.” Why haven’t more been made — especially now, when they could help ongoing efforts to decarbonize energy and the chemical industry?

      One challenge is figuring out the best recipe for making them: Factors such as the best ratio between the silicon and aluminum, what cooking temperature to use, and whether to stir the ingredients all influence the outcome. But the real key, the researchers say, lies in choosing a templating molecule that’s best for producing the intended zeolite framework. Making that match is difficult: There are hundreds of known templating molecules and potentially a million zeolites, and researchers are continually designing new molecules because millions more could be made and might work better.

      For decades, the exploration of how to synthesize a particular zeolite has been done largely by trial and error — a time-consuming, expensive, inefficient way to go about it. There has also been considerable effort to use “atomistic” (atom-by-atom) simulation to figure out what known or novel templating molecule to use to produce a given zeolite. But the experimental and modeling results haven’t generated reliable guidance. In many cases, researchers have carefully selected or designed a molecule to make a particular zeolite, but when they tried their molecule in the lab, the zeolite that formed wasn’t what they expected or desired. So they needed to start over.

      Those experiences illustrate what Gómez-Bombarelli and his colleagues believe is the problem that’s been plaguing zeolite design for decades. All the efforts — both experimental and theoretical — have focused on finding the templating molecule that’s best for forming a specific zeolite. But what if that templating molecule is also really good — or even better — at forming some other zeolite?

      To determine the “best” molecule for making a certain zeolite framework, and the “best” zeolite framework to act as host to a particular molecule, the researchers decided to look at both sides of the pairing. Daniel Schwalbe-Koda PhD ’22, a former member of Gómez-Bombarelli’s group and now a postdoc at Lawrence Livermore National Laboratory, describes the process as a sort of dance with molecules and zeolites in a room looking for partners. “Each molecule wants to find a partner zeolite, and each zeolite wants to find a partner molecule,” he says. “But it’s not enough to find a good dance partner from the perspective of only one dancer. The potential partner could prefer to dance with someone else, after all. So it needs to be a particularly good pairing.” The upshot: “You need to look from the perspective of each of them.”

      To find the best match from both perspectives, the researchers needed to try every molecule with every zeolite and quantify how well the pairings worked.

      A broader metric for evaluating pairs

      Before performing that analysis, the researchers defined a new “evaluating metric” that they could use to rank each templating molecule-zeolite pair. The standard metric for measuring the affinity between a molecule and a zeolite is “binding energy,” that is, how strongly the molecule clings to the zeolite or, conversely, how much energy is required to separate the two. While recognizing the value of that metric, the MIT-led team wanted to take more parameters into account.

      Their new evaluating metric therefore includes not only binding energy but also the size, shape, and volume of the molecule and the opening in the zeolite framework. And their approach calls for turning the molecule to different orientations to find the best possible fit.

      Affinity scores for all molecule-zeolite pairs based on that evaluating metric would enable zeolite researchers to answer two key questions: What templating molecule will form the zeolite that I want? And if I use that templating molecule, what other zeolites might it form instead? Using the molecule-zeolite affinity scores, researchers could first identify molecules that look good for making a desired zeolite. They could then rule out the ones that also look good for forming other zeolites, leaving a set of molecules deemed to be “highly selective” for making the desired zeolite.  

      Validating the approach: A rich literature

      But does their new metric work better than the standard one? To find out, the team needed to perform atomistic simulations using their new evaluating metric and then benchmark their results against experimental evidence reported in the literature. There are many thousands of journal articles reporting on experiments involving zeolites — in many cases, detailing not only the molecule-zeolite pairs and outcomes but also synthesis conditions and other details. Ferreting out articles with the information the researchers needed was a job for machine learning — in particular, for natural language processing.

      For that task, Gómez-Bombarelli and Schwalbe-Koda turned to their DMSE colleague Elsa Olivetti PhD ’07, the Esther and Harold E. Edgerton Associate Professor in Materials Science and Engineering. Using a literature-mining technique that she and a group of collaborators had developed, she and her DMSE team processed more than 2 million materials science papers, found some 90,000 relating to zeolites, and extracted 1,338 of them for further analysis. The yield was 549 templating molecules tested, 209 zeolite frameworks produced, and 5,663 synthesis routes followed.

      Based on those findings, the researchers used their new evaluating metric and a novel atomistic simulation technique to examine more than half-a-million templating molecule-zeolite pairs. Their results reproduced experimental outcomes reported in more than a thousand journal articles. Indeed, the new metric outperformed the traditional binding energy metric, and their simulations were orders of magnitude faster than traditional approaches.

      Ready for experimental investigations

      Now the researchers were ready to put their approach to the test: They would use it to design new templating molecules and try them out in experiments performed by a team led by Yuriy Román-Leshkov, the Robert T. Haslam (1911) Professor of Chemical Engineering, and a team from the Instituto de Tecnologia Química in Valencia, Spain, led by Manuel Moliner and Avelino Corma.

      One set of experiments focused on a zeolite called chabazite, which is used in catalytic converters for vehicles. Using their techniques, the researchers designed a new templating molecule for synthesizing chabazite, and the experimental results confirmed their approach. Their analyses had shown that the new templating molecule would be good for forming chabazite and not for forming anything else. “Its binding strength isn’t as high as other molecules for chabazite, so people hadn’t used it,” says Gómez-Bombarelli. “But it’s pretty good, and it’s not good for anything else, so it’s selective — and it’s way cheaper than the usual ones.”

      In addition, in their new molecule, the electrical charge is distributed differently than in the traditional ones, which led to new possibilities. The researchers found that by adjusting both the shape and charge of the molecule, they could control where the negative charge occurs on the pore that’s created in the final zeolite. “The charge placement that results can make the chabazite a much better catalyst than it was before,” says Gómez-Bombarelli. “So our same rules for molecule design also determine where the negative charge is going to end up, which can lead to whole different classes of catalysts.”

      Schwalbe-Koda describes another experiment that demonstrates the importance of molecular shape as well as the types of new materials made possible using the team’s approach. In one striking example, the team designed a templating molecule with a height and width that’s halfway between those of two molecules that are now commonly used—one for making chabazite and the other for making a zeolite called AEI. (Every new zeolite structure is examined by the International Zeolite Association and — once approved — receives a three-letter designation.)

      Experiments using that in-between templating molecule resulted in the formation of not one zeolite or the other, but a combination of the two in a single solid. “The result blends two different structures together in a way that the final result is better than the sum of its parts,” says Schwalbe-Koda. “The catalyst is like the one used in catalytic converters in today’s trucks — only better.” It’s more efficient in converting nitrogen oxides to harmless nitrogen gases and water, and — because of the two different pore sizes and the aluminosilicate composition — it works well on exhaust that’s fairly hot, as during normal operation, and also on exhaust that’s fairly cool, as during startup.

      Putting the work into practice

      As with all materials, the commercial viability of a zeolite will depend in part on the cost of making it. The researchers’ technique can identify promising templating molecules, but some of them may be difficult to synthesize in the lab. As a result, the overall cost of that molecule-zeolite combination may be too high to be competitive.

      Gómez-Bombarelli and his team therefore include in their assessment process a calculation of cost for synthesizing each templating molecule they identified — generally the most expensive part of making a given zeolite. They use a publicly available model devised in 2018 by Connor Coley PhD ’19, now the Henri Slezynger (1957) Career Development Assistant Professor of Chemical Engineering at MIT. The model takes into account all the starting materials and the step-by-step chemical reactions needed to produce the targeted templating molecule.

      However, commercialization decisions aren’t based solely on cost. Sometimes there’s a trade-off between cost and performance. “For instance, given our chabazite findings, would customers or the community trade a little bit of activity for a 100-fold decrease in the cost of the templating molecule?” says Gómez-Bombarelli. “The answer is likely yes. So we’ve made a tool that can help them navigate that trade-off.” And there are other factors to consider. For example, is this templating molecule truly novel, or have others already studied it — or perhaps even hold a patent on it?

      “While an algorithm can guide development of templating molecules and quantify specific molecule-zeolite matches, other types of assessments are best left to expert judgment,” notes Schwalbe-Koda. “We need a partnership between computational analysis and human intuition and experience.”

      To that end, the MIT researchers and their colleagues decided to share their techniques and findings with other zeolite researchers. Led by Schwalbe-Koda, they created an online database that they made publicly accessible and easy to use — an unusual step, given the competitive industries that rely on zeolites. The interactive website — zeodb.mit.edu — contains the researchers’ final metrics for templating molecule-zeolite pairs resulting from hundreds of thousands of simulations; all the identified journal articles, along with which molecules and zeolites were examined and what synthesis conditions were used; and many more details. Users are free to search and organize the data in any way that suits them.

      Gómez-Bombarelli, Schwalbe-Koda, and their colleagues hope that their techniques and the interactive website will help other researchers explore and discover promising new templating molecules and zeolites, some of which could have profound impacts on efforts to decarbonize energy and tackle climate change.

      This research involved a team of collaborators at MIT, the Instituto de Tecnologia Química (UPV-CSIC), and Stockholm University. The work was supported in part by the MIT Energy Initiative Seed Fund Program and by seed funds from the MIT International Science and Technology Initiative. Daniel Schwalbe-Koda was supported by an ExxonMobil-MIT Energy Fellowship in 2020–21.

      This article appears in the Spring 2022 issue of Energy Futures, the magazine of the MIT Energy Initiative. More

    • 100 Shares149 Views

      in Energy

      Taking a magnifying glass to data center operations

      by

      When the MIT Lincoln Laboratory Supercomputing Center (LLSC) unveiled its TX-GAIA supercomputer in 2019, it provided the MIT community a powerful new resource for applying artificial intelligence to their research. Anyone at MIT can submit a job to the system, which churns through trillions of operations per second to train models for diverse applications, such as spotting tumors in medical images, discovering new drugs, or modeling climate effects. But with this great power comes the great responsibility of managing and operating it in a sustainable manner — and the team is looking for ways to improve.

      “We have these powerful computational tools that let researchers build intricate models to solve problems, but they can essentially be used as black boxes. What gets lost in there is whether we are actually using the hardware as effectively as we can,” says Siddharth Samsi, a research scientist in the LLSC. 

      To gain insight into this challenge, the LLSC has been collecting detailed data on TX-GAIA usage over the past year. More than a million user jobs later, the team has released the dataset open source to the computing community.

      Their goal is to empower computer scientists and data center operators to better understand avenues for data center optimization — an important task as processing needs continue to grow. They also see potential for leveraging AI in the data center itself, by using the data to develop models for predicting failure points, optimizing job scheduling, and improving energy efficiency. While cloud providers are actively working on optimizing their data centers, they do not often make their data or models available for the broader high-performance computing (HPC) community to leverage. The release of this dataset and associated code seeks to fill this space.

      “Data centers are changing. We have an explosion of hardware platforms, the types of workloads are evolving, and the types of people who are using data centers is changing,” says Vijay Gadepally, a senior researcher at the LLSC. “Until now, there hasn’t been a great way to analyze the impact to data centers. We see this research and dataset as a big step toward coming up with a principled approach to understanding how these variables interact with each other and then applying AI for insights and improvements.”

      Papers describing the dataset and potential applications have been accepted to a number of venues, including the IEEE International Symposium on High-Performance Computer Architecture, the IEEE International Parallel and Distributed Processing Symposium, the Annual Conference of the North American Chapter of the Association for Computational Linguistics, the IEEE High-Performance and Embedded Computing Conference, and International Conference for High Performance Computing, Networking, Storage and Analysis. 

      Workload classification

      Among the world’s TOP500 supercomputers, TX-GAIA combines traditional computing hardware (central processing units, or CPUs) with nearly 900 graphics processing unit (GPU) accelerators. These NVIDIA GPUs are specialized for deep learning, the class of AI that has given rise to speech recognition and computer vision.

      The dataset covers CPU, GPU, and memory usage by job; scheduling logs; and physical monitoring data. Compared to similar datasets, such as those from Google and Microsoft, the LLSC dataset offers “labeled data, a variety of known AI workloads, and more detailed time series data compared with prior datasets. To our knowledge, it’s one of the most comprehensive and fine-grained datasets available,” Gadepally says. 

      Notably, the team collected time-series data at an unprecedented level of detail: 100-millisecond intervals on every GPU and 10-second intervals on every CPU, as the machines processed more than 3,000 known deep-learning jobs. One of the first goals is to use this labeled dataset to characterize the workloads that different types of deep-learning jobs place on the system. This process would extract features that reveal differences in how the hardware processes natural language models versus image classification or materials design models, for example.   

      The team has now launched the MIT Datacenter Challenge to mobilize this research. The challenge invites researchers to use AI techniques to identify with 95 percent accuracy the type of job that was run, using their labeled time-series data as ground truth.

      Such insights could enable data centers to better match a user’s job request with the hardware best suited for it, potentially conserving energy and improving system performance. Classifying workloads could also allow operators to quickly notice discrepancies resulting from hardware failures, inefficient data access patterns, or unauthorized usage.

      Too many choices

      Today, the LLSC offers tools that let users submit their job and select the processors they want to use, “but it’s a lot of guesswork on the part of users,” Samsi says. “Somebody might want to use the latest GPU, but maybe their computation doesn’t actually need it and they could get just as impressive results on CPUs, or lower-powered machines.”

      Professor Devesh Tiwari at Northeastern University is working with the LLSC team to develop techniques that can help users match their workloads to appropriate hardware. Tiwari explains that the emergence of different types of AI accelerators, GPUs, and CPUs has left users suffering from too many choices. Without the right tools to take advantage of this heterogeneity, they are missing out on the benefits: better performance, lower costs, and greater productivity.

      “We are fixing this very capability gap — making users more productive and helping users do science better and faster without worrying about managing heterogeneous hardware,” says Tiwari. “My PhD student, Baolin Li, is building new capabilities and tools to help HPC users leverage heterogeneity near-optimally without user intervention, using techniques grounded in Bayesian optimization and other learning-based optimization methods. But, this is just the beginning. We are looking into ways to introduce heterogeneity in our data centers in a principled approach to help our users achieve the maximum advantage of heterogeneity autonomously and cost-effectively.”

      Workload classification is the first of many problems to be posed through the Datacenter Challenge. Others include developing AI techniques to predict job failures, conserve energy, or create job scheduling approaches that improve data center cooling efficiencies.

      Energy conservation 

      To mobilize research into greener computing, the team is also planning to release an environmental dataset of TX-GAIA operations, containing rack temperature, power consumption, and other relevant data.

      According to the researchers, huge opportunities exist to improve the power efficiency of HPC systems being used for AI processing. As one example, recent work in the LLSC determined that simple hardware tuning, such as limiting the amount of power an individual GPU can draw, could reduce the energy cost of training an AI model by 20 percent, with only modest increases in computing time. “This reduction translates to approximately an entire week’s worth of household energy for a mere three-hour time increase,” Gadepally says.

      They have also been developing techniques to predict model accuracy, so that users can quickly terminate experiments that are unlikely to yield meaningful results, saving energy. The Datacenter Challenge will share relevant data to enable researchers to explore other opportunities to conserve energy.

      The team expects that lessons learned from this research can be applied to the thousands of data centers operated by the U.S. Department of Defense. The U.S. Air Force is a sponsor of this work, which is being conducted under the USAF-MIT AI Accelerator.

      Other collaborators include researchers at MIT Computer Science and Artificial Intelligence Laboratory (CSAIL). Professor Charles Leiserson’s Supertech Research Group is investigating performance-enhancing techniques for parallel computing, and research scientist Neil Thompson is designing studies on ways to nudge data center users toward climate-friendly behavior.

      Samsi presented this work at the inaugural AI for Datacenter Optimization (ADOPT’22) workshop last spring as part of the IEEE International Parallel and Distributed Processing Symposium. The workshop officially introduced their Datacenter Challenge to the HPC community.

      “We hope this research will allow us and others who run supercomputing centers to be more responsive to user needs while also reducing the energy consumption at the center level,” Samsi says. More