Leveraging the power of neurodiversity

How does a small startup go head to head with the country’s largest IT consulting firms, growing rapidly and securing clients like AIG, Berkshire Hathaway, and WarnerMedia in the process? For the quality engineering startup Ultranauts, the answer lies in the power of its cognitively diverse workforce.

More than 75 percent of the company’s employees are on the autism spectrum, allowing Ultranauts to tap into the unique strengths of each team member as it helps large enterprises and mature startups improve the quality of their data, analytics, and software.

“We’re able to create value for clients in a way that few other teams can through the simple premise of tapping into a much wider range of talent and different approaches to problem solving,” Ultranauts CEO Rajesh Anandan ’95, SM ’96 says.

Now that Ultranauts, which was founded in 2013, has proven the business viability of its approach, it is keen to help other companies build on its progress.

“If you’re trying to figure out how to build more inclusive engineering and data science teams, we’re always happy to share everything we know and the practices we’ve developed,” says Anandan, who co-founded the company with Art Shectman ’95.

Of course, the best way to encourage others to emulate your approach is to have success. Even amidst the disruptions of the Covid-19 pandemic, Anandan says Ultranauts has continued to grow its revenue by 50 percent a year, and it’s currently using a funding round raised last year to accelerate growth.

“Today the organization isn’t really Art and I — we’re the old guys who need to stay out of the way,” Anandan says. “We’ve got incredibly capable colleagues at all levels of the organization. We have autistic colleagues on the leadership team, running the recruiting team, managing engineering teams, and as a result, we have a truly diverse and inclusive group that does great work.”

Proving the value of neurodiversity

Anandan and Shectman were roommates at MIT and became fast friends. Anandan, who grew up in Sri Lanka and had never owned a computer before coming to MIT, said the experience of meeting so many like-minded people was “like finding your tribe.”

After a stint at Microsoft following graduation, Anandan went into consulting at Bain and Company, and then entered the nonprofit space, first with The Global Fund to Fight AIDS, Tuberculosis and Malaria, and then with the United Nations Children’s Fund (UNICEF).

“Inspired by a UNICEF report on children with disabilities, I did some research with [the management consulting firm] Stax Inc. around opportunity and employment for communities being left out because they were perceived to have a disability. I was sharing the findings with Art, around evidence of certain skills or abilities that were going unnoticed, and describing some of the more common strengths associated with individuals on the autism spectrum,” Anandan says. “At the same time, there’s a real societal challenge around this system that continues to penalize humans for being different.”

By 2012, Shectman had been running a software engineering firm, Elephant Ventures, for eight years and was familiar with the challenge of finding good testers for new software rollouts.

“I said, ‘Find two or three people this week and we’ll hire them on Wednesday,’” Shectman recalls.

The founders didn’t have a personal connection with anyone on the spectrum, but they gained a key early ally in the CEOs of the nonprofit autism advocacy groups Integrate Advisors and GRASP, who helped them put together a job description and distributed it through the GRASP network.

Within 72 hours, the founders had 150 applications. Many of those applicants had graduate degrees, but nobody had experience in software testing. Still, the founders brought on three people, committing to learning and supporting them as they worked.

“There were a couple open questions: Could we build a commercially viable business — not a charity — that could grow and be successful?” Shectman says. “The other was could we do it fully virtually? On both fronts, the so-called experts didn’t think it was possible.”

Within a few months it was clear to the founders that the new hires were performing as well as someone who had been in the industry for years. The duo officially launched Ultranauts in the summer of 2013.

Since then the company has developed recruiting processes that evaluate candidates based on assessments and work simulations rather than relying on previous experience. Ultranauts has also focused on making its work environment as supportive as possible by doing things like distributing meeting agendas ahead of time, captioning every video meeting, and asking for daily feedback from employees.

The aspect of the company the founders are most proud of is the “universal workplace” that Ultranauts is creating, based on flexible norms like desired time equivalent work weeks, transparent decision-making including public management dashboards, and inclusive business practices that enable their fully distributed and diverse teams to do their best work.

“As a mission-driven firm, we want to create an environment where everyone can thrive,” Shectman says. “We want to empower capable humans who just haven’t had a shot before to be successful.”

A thriving business

To date, Ultranauts’ quality engineers have tested a wide range of applications and platforms including insurance underwriting, commercial lending, health care reporting, and content streaming. They can also perform data-quality audits, validate analytics and models, and build end-to-end test automation of information supply chains.

“It’s a really complex ecosystem from ingestion of data into pipelines, into machine learning models, into a software product, into a user interface, and there are failure points all through that journey,” Anandan says. “Our teams are able to come in and understand that complexity and be super surgical about constructing quality checks that we’re able to automate so these disastrous failures are prevented.”

Anandan says Fortune 100 companies have chosen Ultranauts over the global consulting firms they were previously using. The startup was also recently named a finalist for MIT Solve, a social impact initiative for tech entrepreneurs.

“Looking back, I think we could have started any business based on the idea that if we create an environment where we bring different brain types together, with different learning styles and information processing models, to collaborate and focus on attacking the same problem, we’re just going to be better at it,” Anandan says.

For the founders, Ultranauts’ success stems from the company’s ability to create an inclusive work environment for diverse employees.

“That’s the secret sauce: If you can create an environment where you can take a whole bunch of people who are very different, and create conditions where they can truly use their strengths, you end up being better,” Anandan says. “It’s better for the team, it’s better for the business, and it’s better for clients.”

de MIT News https://ift.tt/36oWWa8

Shrinking massive neural networks used to model language

You don’t need a sledgehammer to crack a nut.

Jonathan Frankle is researching artificial intelligence — not noshing pistachios — but the same philosophy applies to his “lottery ticket hypothesis.” It posits that, hidden within massive neural networks, leaner subnetworks can complete the same task more efficiently. The trick is finding those “lucky” subnetworks, dubbed winning lottery tickets.

In a new paper, Frankle and colleagues discovered such subnetworks lurking within BERT, a state-of-the-art neural network approach to natural language processing (NLP). As a branch of artificial intelligence, NLP aims to decipher and analyze human language, with applications like predictive text generation or online chatbots. In computational terms, BERT is bulky, typically demanding supercomputing power unavailable to most users. Access to BERT’s winning lottery ticket could level the playing field, potentially allowing more users to develop effective NLP tools on a smartphone — no sledgehammer needed.

“We’re hitting the point where we’re going to have to make these models leaner and more efficient,” says Frankle, adding that this advance could one day “reduce barriers to entry” for NLP.

Frankle, a PhD student in Michael Carbin’s group at the MIT Computer Science and Artificial Intelligence Laboratory, co-authored the study, which will be presented next month at the Conference on Neural Information Processing Systems. Tianlong Chen of the University of Texas at Austin is the lead author of the paper, which included collaborators Zhangyang Wang, also of Texas A&M, as well as Shiyu Chang, Sijia Liu, and Yang Zhang, all of the MIT-IBM Watson AI Lab.

You’ve probably interacted with a BERT network today. It’s one of the technologies that underlies Google’s search engine, and it has sparked excitement among researchers since Google released BERT in 2018. BERT is a method of creating neural networks — algorithms that use layered nodes, or “neurons,” to learn to perform a task through training on numerous examples. BERT is trained by repeatedly attempting to fill in words left out of a passage of writing, and its power lies in the gargantuan size of this initial training dataset. Users can then fine-tune BERT’s neural network to a particular task, like building a customer-service chatbot. But wrangling BERT takes a ton of processing power.

“A standard BERT model these days — the garden variety — has 340 million parameters,” says Frankle, adding that the number can reach 1 billion. Fine-tuning such a massive network can require a supercomputer. “This is just obscenely expensive. This is way beyond the computing capability of you or me.”

Chen agrees. Despite BERT’s burst in popularity, such models “suffer from enormous network size,” he says. Luckily, “the lottery ticket hypothesis seems to be a solution.”

To cut computing costs, Chen and colleagues sought to pinpoint a smaller model concealed within BERT. They experimented by iteratively pruning parameters from the full BERT network, then comparing the new subnetwork’s performance to that of the original BERT model. They ran this comparison for a range of NLP tasks, from answering questions to filling the blank word in a sentence.

The researchers found successful subnetworks that were 40 to 90 percent slimmer than the initial BERT model, depending on the task. Plus, they were able to identify those winning lottery tickets before running any task-specific fine-tuning — a finding that could further minimize computing costs for NLP. In some cases, a subnetwork picked for one task could be repurposed for another, though Frankle notes this transferability wasn’t universal. Still, Frankle is more than happy with the group’s results.

“I was kind of shocked this even worked,” he says. “It’s not something that I took for granted. I was expecting a much messier result than we got.”

This discovery of a winning ticket in a BERT model is “convincing,” according to Ari Morcos, a scientist at Facebook AI Research. “These models are becoming increasingly widespread,” says Morcos. “So it’s important to understand whether the lottery ticket hypothesis holds.” He adds that the finding could allow BERT-like models to run using far less computing power, “which could be very impactful given that these extremely large models are currently very costly to run.”

Frankle agrees. He hopes this work can make BERT more accessible, because it bucks the trend of ever-growing NLP models. “I don’t know how much bigger we can go using these supercomputer-style computations,” he says. “We’re going to have to reduce the barrier to entry.” Identifying a lean, lottery-winning subnetwork does just that — allowing developers who lack the computing muscle of Google or Facebook to still perform cutting-edge NLP. “The hope is that this will lower the cost, that this will make it more accessible to everyone … to the little guys who just have a laptop,” says Frankle. “To me that’s really exciting.”

This research was funded, in part, by IBM.

de MIT News https://ift.tt/33zBfCy

Fikile Brushett is looking for new ways to store energy

Fikile Brushett, an MIT associate professor of chemical engineering, had an unusual source of inspiration for his career in the chemical sciences: the character played by Nicolas Cage in the 1996 movie “The Rock.” In the film, Cage portrays an FBI chemist who hunts down a group of rogue U.S. soldiers who have commandeered chemical weapons and taken over the island of Alcatraz.

“For a really long time, I really wanted to be a chemist and work for the FBI with chemical warfare agents. That was the goal: to be Nick Cage,” recalls Brushett, who first saw the movie as a high school student living in Silver Spring, Maryland, a suburb of Washington.

Though he did not end up joining the FBI or working with chemical weapons — which he says is probably for the best — Brushett did pursue his love of chemistry. In his lab at MIT, Brushett leads a group dedicated to developing more efficient and sustainable ways to store energy, including batteries that could be used to store the electricity generated by wind and solar power. He is also exploring new ways to convert carbon dioxide to useful fuels.

“The backbone of our global energy economy is based upon liquid fossil fuels right now, and energy demand is increasing,” he says. “The challenge we’re facing is that carbon emissions are tied very tightly to this increasing energy demand, and carbon emissions are linked to climate volatility, as well as pollution and health effects. To me, this is an incredibly urgent, important, and inspiring problem to go after.”

“A body of knowledge”

Brushett’s parents immigrated to the United States in the early 1980s, before he was born. His mother, an English as a second language teacher, is from South Africa, and his father, an economist, is from the United Kingdom. Brushett grew up mostly in the Washington area, with the exception of four years spent living in Zimbabwe, due to his father’s work at the World Bank.

Brushett remembers this as an idyllic time, saying, “School ended at 1 p.m., so you almost had the whole afternoon to do sports at school, or you could go home and just play in the garden.”

His family returned to the Washington area while he was in sixth grade, and in high school, he started to get interested in chemistry, as well as other scientific subjects and math.

At the University of Pennsylvania, he decided to major in chemical engineering because someone had advised him that if he liked chemistry and math, chemical engineering would be a good fit. While he enjoyed some of his chemical engineering classes, he struggled with others at first.

“I remember really having a hard time with chemE for a while, and I was fortunate enough to have a really good academic advisor who said, ‘Listen, chemE is hard for some people. Some people get it immediately, for some people it takes a little while for it to sink in,’” he says. Around his junior year, concepts started to fall into place, he recalls. “Rather than looking at courses as self-contained units, the units started coming together and flowing into a body of knowledge. I was able to see the interconnections between courses.”

While he was originally most interested in molecular biotechnology — the field of engineering proteins and other biological molecules — he ended up working in a reaction engineering lab with his academic advisor, John Vohs. There, he studied how catalytic surfaces influence chemical reactions. At Vohs’ recommendation, he applied to the University of Illinois at Urbana-Champaign for graduate school, where he worked on electrochemistry projects. With his PhD advisor, Paul Kenis, he developed microfluidic fuel cells that could run on a variety of different fuels as portable power sources.

During his third year of graduate school, he began applying for faculty positions and was offered a job at MIT, which he accepted but deferred for two years so he could do a postdoc at Argonne National Laboratory. There, he worked with scientists and engineers doing a wide range of research on electrochemical energy storage, and became interested in flow batteries, which is now one of the major focus areas of his lab at MIT.

Modeling new technology

Unlike the rechargeable lithium-ion batteries that power our cell phones and laptops, flow batteries use large tanks of liquid to store energy. Such batteries have traditionally been prohibitively expensive because they rely on pricey electroactive metal salts. Brushett is working on alternative approaches that use less expensive electroactive materials derived from organic compounds.

Such batteries could be used to store the power intermittently produced by wind turbines and solar panels, making them a more reliable, efficient, and cost-effective source of energy. His lab also works on new processes for converting carbon dioxide, a waste product and greenhouse gas, into useful fuels.

In a related area of research, Brushett’s lab performs “techno-economic” modeling of potential new technologies, to help them assess what aspects of the technology need the most improvement to make them economically feasible.

“With techno-economic modeling, we can devise targets for basic science,” he says. “We’re always looking for the rate-limiting step. What is it that’s preventing us from moving forward? In some cases it could be a catalyst, in other cases it could be a membrane. In other cases it could be the architecture for the device.”

Once those targets are identified, researchers working in those areas have a better idea of what they need to focus on to make a particular technology work, Brushett says.

“That’s the thing I’ve been most proud of from our research — hopefully opening up or demystifying the field and allowing a more diverse set of researchers to enter and to add value, which I think is important in terms of growing the science and developing new ideas,” he says.

de MIT News https://ift.tt/3q9Tvfg

Slowing the spread of Covid-19

An air of uncertainty descended on MIT’s campus in early March. Whispers and rumors about campus closing down swirled in the hallways. Students convened en masse on Killian Court to dance, hug, and cry as they were told they had until the end of the week to vacate campus. Within days, the Infinite Corridor’s usual stream of activity and noise was silenced.

While MIT’s dorms and classrooms became unnervingly quiet, there was a thrum of activity among faculty and researchers. Research teams across the Institute quickly swung into action, hatching plans and developing technologies to slow or stop the spread of the virus. These teams were among the only people allowed on campus this spring to work on Covid-19 related research.

The unprecedented nature of this global pandemic necessitates a diverse range of solutions. From designing low-cost ventilators to understanding how the virus is transmitted and manufacturing PPE, mechanical engineers have been a driving force in many research projects that seek to slow Covid-19’s spread and save lives.

“Mechanical engineers are used to developing concrete solutions for the grand challenges the world faces across a vast range of research areas,” says Evelyn Wang, Gail E. Kendall Professor and head of MIT’s Department of Mechanical Engineering. “This uniquely positioned our research community to serve as leaders in the global response to the Covid-19 pandemic.”

Since the beginning of the year, a number of mechanical engineering faculty and research staff at MIT have led collaborative research efforts in the fight against the virus. These projects have had a tangible impact — deepening our understanding of how the virus spreads, informing international guidelines, and protecting front-line workers and vulnerable populations.

Predicting the spread with machine learning

Earlier this year, as coronavirus cases spiked in countries like Italy, South Korea, and the United States, two main questions emerged: How many cases would there be in each country and what measures could be taken to stop the spread? George Barbastathis, professor of mechanical engineering, worked with Raj Dandekar, a PhD candidate studying civil and environmental engineering, to develop a model that could answer these questions.

The pair created the first-ever model that combined data from the spread of Covid-19 with a neural network to make predictions about the spread and determine which quarantine measures were effective. Dandekar first began developing the model as a project for MIT course 2.168 (Learning Machines), which Barbastathis teaches. He was inspired by a mathematical approach developed by Christopher Rackauckas, instructor of mathematics at MIT, that was published on a pre-print server in January of this year.

“I found it really interesting working in this new field of scientific machine learning, which combines machine learning with the physical world using real-life data,” says Dandekar. Their model enhanced the traditional SEIR model, which captures the number of “susceptible,” “exposed,” “infected,” and “recovered” individuals, by training a neural network to also identify those who were under quarantine and therefore no longer at risk to spread the virus. Using data after the 500th case was recorded in Wuhan, China; Italy; South Korea; and the United States, Barbastathis and Dandekar mapped the spread of the virus and derived what is known as the “quarantine control strength function.”

The result, perhaps unsurprisingly, demonstrated that the stronger the quarantine measures, the more effective a country was in slowing or stopping the spread. After releasing their model open-source on the web, Barbastathis reflected on the second wave that had just hit South Korea during an interview in early April. 

“If the U.S. were to follow the same policy of relaxing quarantine measures too soon, we have predicted that the consequences would be far more catastrophic,” Barbastathis said at the time. Weeks later, many states in the United States found these words to ring true as cases spiked.

Shortly after making their model publicly available, the research team was inundated with requests from Spain to Silicon Valley. Biopharmaceutical companies, government entities, and fellow academics were interested in applying the model to their own work.

Over the summer, Barbastathis and Dandekar began collaborating with Rackauckas and Emma Wang, a sophomore studying electrical engineering and computer science, to make their model even more useful to other researchers across the world. The result is a toolkit that offers both diagnostic and predictive data on a more granular level.

“With our new model, we are able to transform data about Covid-19 into data about how well quarantine measures succeeded in containing the spread per country, and even per state,” says Rackauckas. “Now we have a tool that can assign a global quarantine strength score that researchers can then use to correlate to all sorts of other social phenomenon.”

According to Barbastathis, the resulting model is a testament to what can be accomplished through interdisciplinary collaboration. “Our team represents four different departments and we’re very proud of that,” he says.

The team hopes that the new model will provide insights into exactly which quarantine or social distancing methods are most effective in stopping the spread of the virus. “Our aspiration is that our model can actually correlate the rate of this growth with various aspects of the policies that are being followed,” Barbastathis adds.

While Barbastathis and his colleagues are hoping to understand the spread of the virus on a national or state level, Lydia Bourouiba, associate professor of civil and environmental engineering with a joint appointment in mechanical engineering at MIT, is trying to understand the spread on a micro level.

Mapping the path of viral particles

Bourouiba has spent her entire career trying to understand how diseases spread from one person to another. After her experience as a graduate student in Canada during the outbreak of SARS-CoV-1, commonly known as SARS, she combined her expertise in fluid dynamics with epidemiology, studying the transmission of a range of influenza viruses as a postdoc and instructor.

When she founded The Fluid Dynamics of Disease Transmission Laboratory at MIT, Bourouiba continued to focus on fundamental fluid dynamics in relation to pathogen transmission, as well as how droplets are exhaled from one person — through sneezing, coughing, or breathing — and spread through the air to another person. This research combines experiments and modeling.

Early this year, Bourouiba became concerned about the patterns she was noticing with the virus that would soon be named SARS-CoV-2, or Covid-19. “I was paying very close attention to the unprecedented efforts of control that were deployed in Wuhan. By the end of January, it was very clear to me that this was going to be a pandemic,” recalls Bourouiba.

She started sounding the alarm to various agencies and organizations while continuing to pursue ongoing efforts in her team’s research. She also focused her teaching in course 2.250 (Fluids and Diseases) on events related to SARS-CoV-2.

In late March, Bourouiba published research in JAMA that continued to discuss the paradigm of disease transmission she had proposed in the past, including during a TEDMED lecture in 2019. In the article, she made a call to challenge and update the current scientific framework that has shaped public health recommendations about the routes of respiratory disease transmission.

Many government and health organizations had used a disease transmission framework developed in the 1930s by William Firth Wells to inform mask policies or social distancing rules, such as staying six feet apart from others. However, based on years of research, Bourouiba found particles exhaled from an individual can travel much farther than previously thought.

The main problem with the outdated model is how exhalations are classified. “The physics of the process of exhalations cannot be categorized into isolated large droplets verses aerosols,” says Bourouiba. “It’s a continuum of droplets moving within a multiphase gaseous cloud, and the cloud is critical to drive the overall flow.”

Bourouiba’s team uses a combination of modeling and optical techniques including high-speed imaging, shadowgraphy, schlieren, and a range of particle detection and imaging, to map the transient flow of various exhalations. They use these technologies to image and quantify a range of exhalations — including coughing and sneezing — and create models of these complex flow exhalations. The resulting gaseous cloud can carry and propel droplets expelled up to 16 feet away from a cough and up to 27 feet away from a sneeze.

The findings and public awareness in Bourouiba’s article helped reshape guidance on wearing face masks in public in various locations. Many, including Bourouiba, felt the substantial delay in issuing guidelines on face masks in some locations did not help with desirable early critical containment of the epidemic.

“The review of the SARS event and the toll it had — although now dwarfed by SARS-CoV-2 — led to one major lesson learned: We cannot wait to have definitive and final scientific answers in the heat of a pandemic, typically involving a new pathogen. The precautionary principle should always be used in combination with continuously evolving knowledge,” she says “In addition, investments in research on prevention and control between pandemics is as critical to allow a strong basis of knowledge to start from in these regularly occurring local or global events.”

Moving forward, Bourouiba will focus on studies that build upon her previous work. This will include multiscale fluid modeling pertaining to the assessment of material efficacy for respiratory protection and collaborations to examine the fluid dynamics effects of the actual Covid-19 virus and other pathogens. She is also focusing on air flow in indoor settings, in particular in educational or health care-related settings, to ensure the safety of occupants, patients, and health care workers.

Another team at MIT has also been focusing on the safety of doctors, nurses, and front-line workers through the mass production of a disposable face shield. Martin Culpepper, Class of 1960 Fellow and professor of mechanical engineering, and his team at MIT Project Manus were one of the first groups of researchers to ramp up manufacturing of a final product in an effort to protect people from the spread of Covid-19.

Protecting essential workers

With the number of infected individuals rising rapidly in cities like New York and Boston, Massachusetts in March, a primary concern in the fight against Covid-19 centered on personal protective equipment, or PPE. N95 masks and other protective equipment were in short supply. Many health care professionals were advised to keep masks on for longer than what is safe, putting both themselves and their patients at risk. Labs across MIT donated masks and gloves to local hospitals to help address the shortage. Meanwhile, well-intentioned people turned to sewing machines and 3D printers to make non-medical-grade solutions.

Culpepper worked with Elazer Edelman, the Edward J. Poitras Professor in Medical Engineering and Science at MIT, director of MIT’s Institute for Medical Engineering and Science, and head the MIT Medical Crisis Outreach Team, to tackle this problem. In addition to being a professor at MIT, Edelman is a practicing cardiologist at Brigham and Women’s Hospital. The pair took a different approach to tackling the PPE shortage.

“People were trying to deal with the mask shortage by making more of them, but we wanted to slow down the rate at which health-care workers need to change their masks,” Culpepper explains.

The solution they landed on was a low-cost disposable face shield that health-care workers could secure around their face and neck — protecting themselves and extending the use of the mask they wore underneath the shield.

Culpepper began working on the initial prototype of the face shield at home in early March. With the help of a laser cutter in his basement and the assistance of his children, he tested materials and made a few prototypes. MIT Project Manus staff then made dozens of the prototypes using a laser cutter in the Metropolis makerspace to iterate the design to a final state. They also used a Zund large-format machine in MIT’s Center for Bits and Atoms to experiment with materials that can’t be processed on a laser cutter. Culpepper collaborated closely with Edelman to test designs in the field.

Edelman worked with his colleagues at the hospital to get feedback on the initial design. “I brought the prototypes into the hospital and showed nurses and physicians how to store, assemble, and use these devices,” says Edelman. “We then asked the nurses and physicians to use them in non-Covid situations to give us feedback on the design.”

Culpepper notes that Edelman’s perspective was vital to the project. “Elazer has 'mens et manus' in his veins,” says Culpepper. “He has an amazing way of taking clinician feedback, combining it with his experience and perspective, and then translating this all into actionable engineering speak. He was a critical link in the chain of successes that made this happen.”

Armed with positive feedback from clinicians, Culpepper and MIT Project Manus looked to mass produce the shields. The shields were specifically designed to be manufactured at scale. Die cutting machines could easily cut the design into thousands of flat sheets per hour. The sheets were made of polycarbonate and polyethylene terephthalate glycol, materials carefully chosen to ensure there wouldn’t be strain on the supply chain.

MIT and the face shield manufacturer, Polymershapes, donated over 100,000 face shields to hospitals, urgent care centers, and first responders in the areas hit hardest by the virus, including Boston and New York. As of October, over 800,000 shields had been produced by Polymershapes.

According to Culpepper, the supply chain stabilized more rapidly than had initially been predicted. “I’m happy the supply chain for face shields is righting itself. It was our job to be the stopgap, to be there when people in an emergency needed something quickly until the supply chain stabilized,” he reflects.

The face shields have helped protect hundreds of thousands of health-care workers and patients who otherwise would have needed to turn to unsafe PPE options as cases rose exponentially.

Over the summer, signs of life slowly returned to campus. More research teams were allowed to go back to their laboratories to resume work on non-Covid related research. A number of undergraduate seniors moved on campus to take classes with in-person components. While many mechanical engineering groups can shift their focus back to other research projects, developing solutions for the new reality the world faces will continue to be a priority.

“We have an obligation to use our diverse set of skills and expertise to help solve the pressing problems we now face in light of the pandemic,” says Wang.

Until a vaccine is administered to enough people to stop the virus in its tracks, mechanical engineers will continue to collaborate with researchers and experts across all disciplines to develop technologies, products, and research that deepens our understanding of the virus and aims to slow its spread across the globe.

de MIT News https://ift.tt/39t3g2h

An escape route for seafloor methane

Methane, the main component of natural gas, is the cleanest-burning of all the fossil fuels, but when emitted into the atmosphere it is a much more potent greenhouse gas than carbon dioxide. By some estimates, seafloor methane contained in frozen formations along the continental margins may equal or exceed the total amount of coal, oil, and gas in all other reservoirs worldwide. Yet, the way methane escapes from these deep formations is poorly understood.

In particular, scientists have been faced with a puzzle. Observations at sites around the world have shown vigorous columns of methane gas bubbling up from these formations in some places, yet the high pressure and low temperature of these deep-sea environments should create a solid frozen layer that would be expected to act as a kind of capstone, preventing gas from escaping. So how does the gas get out?

A new study helps explain how and why columns of the gas can stream out of these formations, known as methane hydrates. Using a combination of deep-sea observations, laboratory experiments, and computer modeling, researchers have found phenomena that explain and predict the way the gas breaks free from the icy grip of a frozen mix of water and methane. The findings are reported today in the journal PNAS, in a paper by Xiaojing (Ruby) Fu SM ’15, PhD ’17, now at the University of California at Berkeley; Professor Ruben Juanes at MIT; and five others in Switzerland, Spain, New Mexico, and California.

Surprisingly, not only does the frozen hydrate formation fail to prevent methane gas from escaping into the ocean column, but in some cases it actually facilitates that escape.

Early on, Fu saw photos and videos showing plumes of methane, taken from a NOAA research ship in the Gulf of Mexico, revealing the process of bubble formation right at the seafloor. It was clear that the bubbles themselves often formed with a frozen crust around them, and would float upward with their icy shells like tiny helium balloons.

Later, Fu used sonar to detect similar bubble plumes from a research ship off the coast of Virginia. “This cruise alone detected thousands of these plumes,” says Fu, who led the research project while a graduate student and postdoc at MIT. “We could follow these methane bubbles encrusted by hydrate shells into the water column,” she says. “That’s when we first knew that hydrate forming on these gas interfaces can be a very common occurrence.”

But exactly what was going on beneath the seafloor to trigger the release of these bubbles remained unknown. Through a series of lab experiments and simulations, the mechanisms at work gradually became apparent.

Seismic studies of the subsurface of the seafloor in these vent regions show a series of relatively narrow conduits, or chimneys, through which the gas escapes. But the presence of chunks of gas hydrate from these same formations made it clear that the solid hydrate and the gaseous methane could co-exist, Fu explains. To simulate the conditions in the lab, the researchers used a small two-dimensional setup, sandwiching a gas bubble in a layer of water between two plates of glass under high pressure.

As a gas tries to rise through the seafloor, Fu says, if it’s forming a hydrate layer when it hits the cold seawater, that should block its progress: “It’s running into a wall. So how would that wall not be preventing it from continuous migration?” Using the microfluidic experiments, they found a previously unknown phenomenon at work, which they dubbed crustal fingering.

If the gas bubble starts to expand, “what we saw is that the expansion of the gas was able to create enough pressure to essentially rupture the hydrate shell. And it’s almost like it’s hatching out of its own shell,” Fu says. But instead of each rupture freezing back over with the reforming hydrate, the hydrate formation takes place along the sides of the rising bubble, creating a kind of tube around the bubble as it moves upward. “It’s almost like the gas bubble is able to chisel out its own path, and that path is walled by the hydrate solid,” she says. This phenomenon they observed at small scale in the lab, their analysis suggests, is also what would also happen at much larger scale in the seafloor.

That observation, she said, “was really the first time we’ve been aware of a phenomenon like this that could explain how hydrate formation will not inhibit gas flow, but rather in this case, it would facilitate it,” by providing a conduit and directing the flow. Without that focusing, the flow of gas would be much more diffuse and spread out.

As the crust of hydrate forms, it slows down the formation of more hydrate because it forms a barrier between the gas and the seawater. The methane below the barrier can therefore persist in its unfrozen, gaseous form for a long time. The combination of these two phenomena — the focusing effect of the hydrate-walled channels and the segregation of the methane gas from the water by a hydrate layer — “goes a long way toward explaining why you can have some of this vigorous venting, thanks to the hydrate formation, rather than being prevented by it,” says Juanes.

A better understanding of the process could help in predicting where and when such methane seeps will be found, and how changes in environmental conditions could affect the distribution and output of these seeps. While there have been suggestions that a warming climate could increase the rate of such venting, Fu says there is little evidence of that so far. She notes that temperatures at the depths where these formations occur — 600 meters (1,900 feet) deep or more — are expected to experience a smaller temperature increase than would be needed to trigger a widespread release of the frozen gas.

Some researchers have suggested that these vast undersea methane formations might someday be harnessed for energy production. Though there would be great technical hurdles to such use, Juanes says, these findings might help in assessing the possibilities.

"The problem of how gas can move through the hydrate stability zone, where we would expect the gas to be immobilized by being converted to hydrate, and instead escape at the seafloor, is still not fully understood," says Hugh Daigle, an associate professor of petroleum and geosystems engineering at the University of Texas at Austin, who was not associated with this research. "This work presents a probable new mechanism that could plausibly allow this process to occur, and nicely integrates previous laboratory observations with modeling at a larger scale."

"In a practical sense, the work here takes a phenomenon at a small scale and allows us to use it in a model that only considers larger scales, and will be very useful for implementing in future work," Daigle says.

The research team included Joaquin Jimenez-Martinez at the Swiss Federal Institute of Aquatic Science and Technology; Than Phon Nguyen, William Carey and Hari Vinaswanathan at Los Alamos National Laboratory; and Luis Cueto-Felgueroso at the Technical University of Madrid. The work was supported by the U.S. Department of Energy.

de MIT News https://ift.tt/3fPScNM

Expressing our immigration stories

The first issue of Rooted: A Sense of Belonging pays tribute to individual immigration stories through a collection of art, poems, and photos. Each piece breathes life into raw and personal accounts of adversity, resilience, and pride of Asian American and Pacific Islander heritage. “Our contributors have poured their dreams, hardships, and success into artistic expression,” says Olivia Yao '20, magazine editor and recent MIT alum. “I encourage everyone to read every single piece in this publication. Embrace their differences. Know that beyond the 25 stories, there are infinitely many more stories to hear.”

Sponsored by MindHandHeart’s Innovation Fund, members of the MIT Asian American Initiative were inspired by Asian American and Pacific Islander (AAPI) Heritage Month to take action and amplify the voices of individuals with immigration stories in the community. The magazine is a celebration of Asian and non-Asian heritage and is a stark juxtaposition to the xenophobic rhetoric that has spread across the nation during the course of the pandemic.

“After we wrote an article about anti-Asian American racism, it made me think about how rampant xenophobia is in this country. If the zine helps one person in the MIT community relate to another student’s immigration story or feel celebrated, that's a win,” says Alana Chandler, zine assistant editor and junior in materials science and engineering.

After a community-wide call for submissions last spring, Yao, with help from Chandler, worked together over the summer to design visual elements to complement the stories and narratives that were submitted. The word art displayed throughout the magazine are phrases pulled from contributor submissions, and is how the zine’s title emerged.

After several weeks of assembling all the elements of the magazine together, the MIT AAI released the digital copy of the zine and printed 200 physical copies to ship nationwide to members in the community. The distribution of physical copies was part of MIT AAI’s fundraising initiative for the Navajo and Hopi COVID-19 Relief Fund. The MIT AAI was able to raise $270 for the fund as a result of the generous contributions from the community.

Although the zine highlights the value in embracing immigration roots, the MIT AAI also recognizes the importance to confront the settler-colonial history of the country, where many individuals live and benefit off of land that belonged to Indigenous people. “Indigenous communities, such as the Navajo Nation, have had the highest rates of Covid-19 in the United States,” says Chandler. “Systemic racism, manifesting itself through food deserts, inequitable health care, lack of federal funding, poor access to running water, among other reasons, has caused Indigenous people to suffer disproportionately at the hands of this pandemic.”

Members of MIT AAI hope the zine serves as a vehicle to reflect and appreciate the wealth of immigrant experiences that the MIT community possesses. Yu Jing Chen, founding member of MIT AAI and a junior in urban studies and planning, wants the zine to be a stark reminder of the importance of reflection and remembering one's own familial history. “In America, it’s very easy to feel like you need to be a certain way to assimilate to be accepted, and we really want to instill that pride in being different and reconnecting with an integral part of each and every one of us,” says Chen.

“Reading the submissions, you learn so much about your peers and it’s really beautiful to see how everyone comes from such different places, but we’re all here in this place called MIT. It’s heartwarming,” adds Chandler.

The MIT Asian American Initiative is a new student-run organization for Asian American advocacy, alliance, and civic engagement. The group serves as a platform to explore, unpack, and educate students on issues within the Asian American community at MIT and beyond related to mental health, race, equity, and inclusion of Asian Americans. 

de MIT News https://ift.tt/2VqDfs1

Sensor can detect scarred or fatty liver tissue

About 25 percent of the U.S. population suffers from fatty liver disease, a condition that can lead to fibrosis of the liver and, eventually, liver failure.

Currently there is no easy way to diagnose either fatty liver disease or liver fibrosis. However, MIT engineers have now developed a diagnostic tool, based on nuclear magnetic resonance (NMR), that could be used to detect both of those conditions.

“Since it’s a noninvasive test, you could screen people even before they have obvious symptoms of compromised liver, and you would be able to say which of these patients had fibrosis,” says Michael Cima, the David H. Koch Professor of Engineering in MIT’s Department of Materials Science and Engineering, a member of MIT’s Koch Institute for Integrative Cancer Research, and the senior author of the study.

The device, which is small enough to fit on a table, uses NMR to measure how water diffuses through tissue, which can reveal how much fat is present in the tissue. This kind of diagnostic, which has thus far been tested on mice, could help doctors catch fatty liver disease before it progresses to fibrosis, the researchers say.

MIT PhD recipient Ashvin Bashyam and graduate student Chris Frangieh are the lead authors of the paper, which appears today in Nature Biomedical Engineering.

Tissue analysis

Fatty liver disease occurs when liver cells store too much fat. This leads to inflammation and eventually fibrosis, a buildup of scar tissue that can cause jaundice and liver cirrhosis, and eventually liver failure. Fibrosis is usually not diagnosed until the patient begins to experience symptoms that include not only jaundice but also fatigue and abdominal swelling. A biopsy is needed to confirm the diagnosis, but this is an invasive procedure and may not be accurate if the biopsy sample is taken from a part of the liver that is not fibrotic.

To create an easier way to check for this kind of liver disease, Cima and his colleagues had the idea of adapting a detector that they had previously developed to measure hydration levels before and after patients undergo dialysis. That detector measures fluid volume in patients’ skeletal muscle by using NMR to track changes in the magnetic properties of hydrogen atoms of water in the muscle tissue.

The researchers thought that a similar detector could be used for identifying liver disease because water diffuses more slowly when it encounters fatty tissue or fibrosis. Tracking how water moves through tissue over time can reveal how much fatty or scarred tissue is present.

“If you watch how the magnetization changes, you can model how fast the protons are moving,” Cima says. “Those cases where the magnetization doesn't go away very fast would be ones where the diffusivity was low, and they would be the most fibrotic.”

In a study of mice, the researchers showed that their detector could identify fibrosis with 86 percent accuracy, and fatty liver disease with 92 percent accuracy. It takes about 10 minutes to obtain the results, but the researchers are now working on improving the signal-to-noise ratio of the detector, which could help to reduce the amount of time it takes.

Early detection

The current version of the sensor can scan to a depth of about 6 millimeters below the skin, which is enough to monitor the mouse liver or human skeletal muscle. The researchers are now working on designing a new version that can penetrate deeper below the tissue, to allow them to test the liver diagnosis application in human patients.

If this type of NMR sensor could be developed for use in patients, it could help to identify people in danger of developing fibrosis, or in the early stages of fibrosis, so they could be treated earlier, Cima says. Fibrosis can’t be reversed, but it can be halted or slowed down through dietary changes and exercise. Having this type of diagnostic available could also aid in drug development efforts, because it could allow doctors to more easily identify patients with fibrosis and monitor their response to potential new treatments, Cima says.

Another potential application for this kind of sensor is to evaluate human livers for transplant. In this study, the researchers tested the monitor on human liver tissue and found that it could detect fibrosis with 93 percent accuracy.

The research was funded by the Koch Institute Support (core) Grant from the National Cancer Institute, the National Institutes of Health, a Fannie and John Hertz Foundation Graduate Fellowship, and a National Science Foundation Graduate Fellowship.

de MIT News https://ift.tt/2VkoOpu

Computer-aided creativity in robot design

So, you need a robot that climbs stairs. What shape should that robot be? Should it have two legs, like a person? Or six, like an ant?

Choosing the right shape will be vital for your robot’s ability to traverse a particular terrain. And it’s impossible to build and test every potential form. But now an MIT-developed system makes it possible to simulate them and determine which design works best.

You start by telling the system, called RoboGrammar, which robot parts are lying around your shop — wheels, joints, etc. You also tell it what terrain your robot will need to navigate. And RoboGrammar does the rest, generating an optimized structure and control program for your robot.

The advance could inject a dose of computer-aided creativity into the field. “Robot design is still a very manual process,” says Allan Zhao, the paper’s lead author and a PhD student in the MIT Computer Science and Artificial Intelligence Laboratory (CSAIL). He describes RoboGrammar as “a way to come up with new, more inventive robot designs that could potentially be more effective.”

Zhao is the lead author of the paper, which he will present at this month’s SIGGRAPH Asia conference. Co-authors include PhD student Jie Xu, postdoc Mina Konaković-Luković, postdoc Josephine Hughes, PhD student Andrew Spielberg, and professors Daniela Rus and Wojciech Matusik, all of MIT.

Ground rules

Robots are built for a near-endless variety of tasks, yet “they all tend to be very similar in their overall shape and design,” says Zhao. For example, “when you think of building a robot that needs to cross various terrains, you immediately jump to a quadruped,” he adds, referring to a four-legged animal like a dog. “We were wondering if that’s really the optimal design.”

Zhao’s team speculated that more innovative design could improve functionality. So they built a computer model for the task — a system that wasn’t unduly influenced by prior convention. And while inventiveness was the goal, Zhao did have to set some ground rules.

The universe of possible robot forms is “primarily composed of nonsensical designs,” Zhao writes in the paper. “If you can just connect the parts in arbitrary ways, you end up with a jumble,” he says. To avoid that, his team developed a “graph grammar” — a set of constraints on the arrangement of a robot’s components. For example, adjoining leg segments should be connected with a joint, not with another leg segment. Such rules ensure each computer-generated design works, at least at a rudimentary level.

Zhao says the rules of his graph grammar were inspired not by other robots but by animals — arthropods in particular. These invertebrates include insects, spiders, and lobsters. As a group, arthropods are an evolutionary success story, accounting for more than 80 percent of known animal species. “They’re characterized by having a central body with a variable number of segments. Some segments may have legs attached,” says Zhao. “And we noticed that that’s enough to describe not only arthropods but more familiar forms as well,” including quadrupeds. Zhao adopted the arthropod-inspired rules thanks in part to this flexibility, though he did add some mechanical flourishes. For example, he allowed the computer to conjure wheels instead of legs.

A phalanx of robots

Using Zhao’s graph grammar, RoboGrammar operates in three sequential steps: defining the problem, drawing up possible robotic solutions, then selecting the optimal ones. Problem definition largely falls to the human user, who inputs the set of available robotic components, like motors, legs, and connecting segments. “That’s key to making sure the final robots can actually be built in the real world,” says Zhao. The user also specifies the variety of terrain to be traversed, which can include combinations of elements like steps, flat areas, or slippery surfaces.

With these inputs, RoboGrammar then uses the rules of the graph grammar to design hundreds of thousands of potential robot structures. Some look vaguely like a racecar. Others look like a spider, or a person doing a push-up. “It was pretty inspiring for us to see the variety of designs,” says Zhao. “It definitely shows the expressiveness of the grammar.” But while the grammar can crank out quantity, its designs aren’t always of optimal quality.

Choosing the best robot design requires controlling each robot’s movements and evaluating its function. “Up until now, these robots are just structures,” says Zhao. The controller is the set of instructions that brings those structures to life, governing the movement sequence of the robot’s various motors. The team developed a controller for each robot with an algorithm called Model Predictive Control, which prioritizes rapid forward movement.

“The shape and the controller of the robot are deeply intertwined,” says Zhao, “which is why we have to optimize a controller for every given robot individually.” Once each simulated robot is free to move about, the researchers seek high-performing robots with a “graph heuristic search.” This neural network algorithm iteratively samples and evaluates sets of robots, and it learns which designs tend to work better for a given task. “The heuristic function improves over time,” says Zhao, “and the search converges to the optimal robot.”

This all happens before the human designer ever picks up a screw.

“This work is a crowning achievement in the a 25-year quest to automatically design the morphology and control of robots,” says Hod Lipson, a mechanical engineer and computer scientist at Columbia University, who was not involved in the project. “The idea of using shape-grammars has been around for a while, but nowhere has this idea been executed as beautifully as in this work. Once we can get machines to design, make and program robots automatically, all bets are off.”

Zhao intends the system as a spark for human creativity. He describes RoboGrammar as a “tool for robot designers to expand the space of robot structures they draw upon.” To show its feasibility, his team plans to build and test some of RoboGrammar’s optimal robots in the real world. Zhao adds that the system could be adapted to pursue robotic goals beyond terrain traversing. And he says RoboGrammar could help populate virtual worlds. “Let’s say in a video game you wanted to generate lots of kinds of robots, without an artist having to create each one,” says Zhao. “RoboGrammar would work for that almost immediately.”

One surprising outcome of the project? “Most designs did end up being four-legged in the end,” says Zhao. Perhaps manual robot designers were right to gravitate toward quadrupeds all along. “Maybe there really is something to it.”

de MIT News https://ift.tt/2HOIbnt

3 Questions: Using fabric to “listen” to space dust

Earlier this month a team of MIT researchers sent samples of various high-tech fabrics, some with embedded sensors or electronics, to the International Space Station. The samples (unpowered for now) will be exposed to the space environment for a year in order to determine a baseline for how well these materials survive the harsh environment of low Earth orbit.

The hope is that this work could lead to thermal blankets for spacecraft, that could act as sensitive detectors for impacting micrometeoroids and space debris. Ultimately, another goal is new smart fabrics that allow astronauts to feel touch right through their pressurized suits.

Three members of MIT’s multidisciplinary team, graduate students Juliana Cherston of the Media Lab, Yuchen Sun of the Department of Chemistry, and Wei Yan of the Research Laboratory of Electronics and the Department of Materials Science and Engineering, discussed the experiment’s ambitious aims with MIT News.

Q:​ Can you describe the fabric samples that you sent to the International Space Station, and what kinds of information you are hoping to get from them after their exposure in space?

Cherston: The white color of the International Space Station is actually a protective fabric material called Beta cloth, which is a Teflon-impregnated fiberglass designed to shield spacecraft and spacesuits from the harsh elements of low Earth orbit. For decades, these fabrics have remained electrically passive, despite offering large-area real estate on the exterior of space assets.

We imagine turning this spacecraft skin into an enormous space debris and micrometeoroid impact sensor. The samples that we worked with JAXA, the Japanese space agency, and Space BD to send to the International Space Station incorporate materials like charge-sensitive synthetic fur — an early concept — and vibration-sensitive fiber sensors — our project’s focus — into space-resilient fabrics. The resulting fabric may be useful for detecting cosmic dust of scientific interest, and for damage detection on spacecraft. 

It’s easy to assume that since we’re already sending these materials to space, the technology must be very mature. In reality, we are leveraging the space environment  to complement our important ground-testing efforts. All of these fabric sensors will remain unpowered for this first in-space test, and the quilt of samples occupies a total area of 10 by 10 centimeters on the exterior walls of the station.

Our focus is on baselining their resiliency to the space environment. In one year, these samples will return to Earth for postflight analysis. We’ll be able to measure any erosion from atomic oxygen, discoloration from UV radiation, and any changes to fiber sensor performance after one year of thermal cycling. There is some chance that we will also find hints of micron-scale micrometeoroids. We’re also already preparing for an electrically powered deployment currently scheduled for late 2021 or early 2022 (recently awarded to the project by the ISS National Lab). At that point we’ll apply an additional protective coating to the fibers and actually operate them in space.

Yan: The fabric samples contain thermally drawn “acoustic” fibers developed with ISN funding that are capable of converting mechanical vibration energy into electric energy (via the piezoelectric effect). When micrometeoroids or space debris hit the fabric, the fabric vibrates, and the “acoustic” fiber generates an electrical signal. Thermally drawn multimaterial fibers have been developed by our research group at MIT for more than 20 years; what makes these acoustic fibers special is their exquisite sensitivity to mechanical vibrations. The fabric has been shown in ground facilities to detect and measure impact regardless of where the space dust impacted the surface of the fabric.

Q: What is the ultimate goal of the project? What kinds of uses do you foresee for advanced fabrics in the space environment?

Cherston: I am particularly keen to demonstrate that instrumentation useful for fundamental scientific inquiry can be incorporated directly into the fabric skin of persistent spacecraft, which to date is unused and very precious real estate. In particular, I am beginning to evaluate whether these skins are sensitive enough to detect cosmic dust produced in million-year-old supernova explosions tens or hundreds of light-years away from Earth. Just last year, an isotopic signature for this type of interstellar dust was discovered in fresh Antarctic snow, so we believe that some of this dust is still whizzing around the solar system, holding clues about the dynamics of supernova explosions. In-situ characterization of their distribution and kinematics is currently my most ambitious scientific goal.

More generally, I’d love to see advanced fibers and fabrics tackle other questions of fundamental physical interest in space, maybe by leveraging optical fibers or radiation sensitive materials to create large aperture sensors.

Some students in my group have also developed a conceptual prototype in which sensory data on the exterior skin of a pressurized spacesuit armband is mapped to haptic actuators on the wearer’s biological skin. Using this system, astronauts will be able to feel texture and touch right through their spacesuits! This direct experience of a new environment is very central to humanity’s drive to explore.

An impact-sensitive skin can also be used for damage detection on persistent space craft. In practice, the fabric’s ability to localize damage from space debris and micrometeoroids is how we will really sell the concept to aerospace engineers.

Yan: Although the space age began 63 years ago when Soviet Union’s Sputnik 1 was launched into an elliptical low Earth orbit, many unanswered questions remain regarding the effect of the space environment on humans, as well as the safety of astronauts as they operate in the space environment. While our project’s main focus has been on augmenting fabrics used on the exterior of spacecraft, I also envision that future spacesuits will be electrically active and highly multifunctional.

Textiles buried within the suit will be able monitor the health condition of astronauts in real time by interrogating physiological signals over large areas. Fabrics may also serve as  localized heating and cooling systems, radiation dosimeters, and efficient communications infrastructure (via fabric optics and acoustics). They may harvest solar energy as well as small amounts of energy from vibration, and store this energy in fiber batteries or supercapacitors, which would allow the system to be self-powered. Fabrics might even serve as part of an exoskeleton that assists astronauts in maneuvering on planetary bodies and in microgravity. One broad vision at play is to pack an enormous amount of function into space resilient textiles, creating an analogue of “Moore’s law” for space fabrics.

​​​Q: What got you interested in this subject, and what has this experience been like for you in getting the materials ready to be sent into space?

Yan: Space is definitely a new frontier for our research, while lots of terrestrial applications have been envisioned in ambient conditions and even under water. From low Earth orbit to planetary bodies, space is a unique environment with atomic oxygen, radiation, high speed impactors, and extreme temperature cycling. How will the fibers and fabrics perform there and what changes will be induced in the fiber materials? How should electronic fabrics be designed in order to meet demands of aerospace applications? There are so many scientific and technological questions.

Sun: Our group [with professor of chemistry Keith Nelson] strives to push the limits of what is experimentally achievable for impact testing, and we are always excited by a new challenge. Recently, we have been venturing into the area of high-speed mechanics, testing novel materials spanning polymers, thin films, and nanoarchitected materials using a laser accelerator facility designed by our lab to impinge tiny particles on target surfaces at speeds exceeding 1 kilometer per second.

When the idea emerged to test a material capable of detecting impact signatures in low Earth orbit and beyond, there was immediate interest on our side since it is fundamentally different from our previous research focus. These experiments are certainly more difficult and complex than what we are used to, with many more active parts to maintain. I think we were all quite pleasantly surprised when our preliminary impact experiments were successful and encouraging.

Cherston: While space launches are exciting, in reality some of our most convincing data to date has come from impact testing on the ground. Initially, it was not at all obvious that a fabric sensor with sparsely integrated sensing elements could actually detect such small and fast particles. There were a really great few minutes at our first serious impact testing campaign during which Yuchen gradually increased the number of particles accelerated onto our sensor, while holding all other aspects of the experiment constant. The growing signal was a smoking gun indication that we were seeing a true impact signature. 

On a personal level, I’m really fascinated by the idea of leveraging very unconventional technology like fabric for questions of scientific significance. And I think the idea of feeling right through a pressurized spacesuit is delightful!

de MIT News https://ift.tt/2KuSM83

Second annual MIT Science Bowl Invitational takes virtual format

Researchers in the Baker group at MIT are studying the mechanisms of proteins that catalyze protein unfolding, which often involves large conformational changes. Which technique can be used to actively monitor these protein conformational changes?

A) Förster Resonance Energy Transfer

B) Fluorescence In Situ Hybridization

C) Fluorescence Recovery After Photobleaching

D) Mass Spectrometry

If you chose A, you might have stood a chance at the second annual MIT Science Bowl Invitational. On Nov. 7, 192 high school students logged in to their computers for the event, where they competed to answer trivia questions like this one.

Amidst the Covid-19 pandemic, the event took on a completely virtual format, enabling 40 high school teams — almost twice as many teams as attended last year — to join in on the fun. Teams registered from locations as far as California and Hawaii. For the first time, the organizers spent six weeks writing all of the competition’s questions, including many about current research at MIT.

“Our volunteers and teams are as excited and motivated as ever to make these competitions happen,” says senior and economics major Paolo Adajar, the president of MIT Science Bowl. “And even though things are virtual, interest in science is not something that will go away. People will inevitably be excited about science.”

Going virtual

Ten MIT students organized and planned this fall’s high school invitational.

Last year’s inaugural invitational welcomed 24 teams, mostly from the Northeast, to MIT’s campus. Adajar expected about the same number to register this year. However, the virtual format led to an explosion of interest from across the country. “We had almost 80 teams register,” Adajar says. “We wondered, ‘How do we accommodate this many teams?’ So we expanded our recruitment for volunteers as much as we could.”

Thirty MIT students volunteered on the day of the event, helping teams get logged into the correct Zoom links at the correct times, and serving as scorekeepers and judges. The event also required an additional 50 non-MIT affiliated volunteers, including Science Bowl alumni and friends of MIT students. Organizers and volunteers communicated throughout the day via Slack.

The event, sponsored by the School of Science, began at 9 a.m. with a morning full of fast-paced, round-robin-style competitions. Teams broke out into Zoom rooms to compete against one another for a spot in the afternoon competition, which was double-elimination style.

“Once it was far enough into the competition and we got to the point where it was one match happening at a time, we livestreamed those matches to all competitors to recreate the feeling of sitting in an auditorium and watching those matches,” Adajar says. “Just some amount of watching it all happen is pretty fun. It’s nice even to just watch the live chat go on and see all of these dozens of high school students, who are very excited about science, discussing answers among themselves.”

Asking the tough questions

Because last year was the first invitational the club hosted, the group had been mostly focused on learning the ropes of hosting a large high school event. They used a list of Science Bowl questions curated by the U.S. Department of Energy to quiz participants. However, this year, the MIT Science Bowl Club decided to write the questions themselves.

“It’s very difficult to write good questions at a very high level of difficulty,” Adajar says. “But with the incredible people who are students here at MIT, and the number of Science Bowl alum that go to this school, if there's any place in the world that could write questions of high difficulty and with a very high caliber, it's going to be us.”

Junior and mathematics major Mihir Singhal, treasurer of the MIT Science Bowl, worked with a group of around 15 MIT students to write more than 1,000 questions. The questions spanned six disciplines: math, biology, chemistry, physics, Earth and space science, and an additional category that touched on MIT-based research.

“We decided to write the questions this year because we thought it would be cool to integrate MIT and also to be able to use new questions that haven't been used before,” Singhal says. Both Adajar and Singhal participated in Science Bowl competitions in middle and high school.

A new format brings new challenges

While the remote competition allowed for more volunteers and students to participate remotely, the new format also posed a few unexpected logistical challenges. For one, at an in-person Science Bowl, competitors press buzzers to indicate that they have an answer to a question.

Initially, while planning this year’s event, Adajar wondered how it would be possible for two teams thousands of miles apart to buzz against each other in real-time.

But with a bit of problem-solving, the Science Bowl organizers discovered a platform called the Conventional Online Buzzer Application, which allowed participants to buzz in remotely without flaw. At the beginning of each round, all participants tested their buzzers.

In addition to overcoming technological challenges, Adajar, Singhal, and the organizing team had to get creative when thinking of ways to cultivate a fun, social atmosphere. Adajar said a special part of most Science Bowl competitions is the in-person bonding and camaraderie that takes place between students. Normally, in the weeks leading up to the event, the organizing team is busy ordering T-shirts and pizza for 100 students. This year, social planning for the event looked different.

“We definitely did design opportunities for teams to meet each other in more low-stakes settings than just the competition,” Adajar says. “We shared some contact information with teams in the afternoon so they could scrimmage each other just for fun.”

Ultimately, North Hollywood High School, located in Los Angeles, California, was this year’s winner.

The MIT Science Bowl Club also organizes a regional competition for middle schoolers every spring. Part of the U.S. Department of Energy’s national competition, winners of the Northeast Regional Science Bowl proceed to the annual National Science Bowl competition. While the middle school science bowl is coming up on its sixth year at MIT, last year was the first time the club hosted their informal 24-team high school invitational event on campus.

de MIT News https://ift.tt/3nUfTr2

How humans use objects in novel ways to solve problems

Human beings are naturally creative tool users. When we need to drive in a nail but don’t have a hammer, we easily realize that we can use a heavy, flat object like a rock in its place. When our table is shaky, we quickly find that we can put a stack of paper under the table leg to stabilize it. But while these actions seem so natural to us, they are believed to be a hallmark of great intelligence — only a few other species use objects in novel ways to solve their problems, and none can do so as flexibly as people. What provides us with these powerful capabilities for using objects in this way?

In a new paper published in the Proceedings of the National Academy of Sciences describing work conducted at MIT’s Center for Brains, Minds and Machines, researchers Kelsey Allen, Kevin Smith, and Joshua Tenenbaum study the cognitive components that underlie this sort of improvised tool use. They designed a novel task, the Virtual Tools game, that taps into tool-use abilities: People must select one object from a set of “tools” that they can place in a two-dimensional, computerized scene to accomplish a goal, such as getting a ball into a certain container. Solving the puzzles in this game requires reasoning about a number of physical principles, including launching, blocking, or supporting objects.

The team hypothesized that there are three capabilities that people rely on to solve these puzzles: a prior belief that guides people’s actions toward those that will make a difference in the scene, the ability to imagine the effect of their actions, and a mechanism to quickly update their beliefs about what actions are likely to provide a solution. They built a model that instantiated these principles, called the “Sample, Simulate, Update,” or “SSUP,” model, and had it play the same game as people. They found that SSUP solved each puzzle at similar rates and in similar ways as people did. On the other hand, a popular deep learning model that could play Atari games well but did not have the same object and physical structures was unable to generalize its knowledge to puzzles it was not directly trained on.

This research provides a new framework for studying and formalizing the cognition that supports human tool use. The team hopes to extend this framework to not just study tool use, but also how people can create innovative new tools for new problems, and how humans transmit this information to build from simple physical tools to complex objects like computers or airplanes that are now part of our daily lives.

Kelsey Allen, a PhD student in the Computational Cognitive Science Lab at MIT, is excited about how the Virtual Tools game might support other cognitive scientists interested in tool use: “There is just so much more to explore in this domain. We have already started collaborating with researchers across multiple different institutions on projects ranging from studying what it means for games to be fun, to studying how embodiment affects disembodied physical reasoning. I hope that others in the cognitive science community will use the game as a tool to better understand how physical models interact with decision-making and planning.”

Joshua Tenenbaum, professor of computational cognitive science at MIT, sees this work as a step toward understanding not only an important aspect of human cognition and culture, but also how to build more human-like forms of intelligence in machines. “Artificial Intelligence researchers have been very excited about the potential for reinforcement learning (RL) algorithms to learn from trial-and-error experience, as humans do, but the real trial-and-error learning that humans benefit from unfolds over just a handful of trials — not millions or billions of experiences, as in today’s RL systems,” Tenenbaum says. “The Virtual Tools game allows us to study this very rapid and much more natural form of trial-and-error learning in humans, and the fact that the SSUP model is able to capture the fast learning dynamics we see in humans suggests it may also point the way towards new AI approaches to RL that can learn from their successes, their failures, and their near misses as quickly and as flexibly as people do.” 

de MIT News https://ift.tt/2IVLmKT

Center to advance predictive simulation research established at MIT Schwarzman College of Computing

Understanding the degradation of materials in extreme environments is a scientific problem with major technological applications, ranging from spaceflight to industrial and nuclear safety. Yet it presents an intrinsic challenge: Researchers cannot easily reproduce these environments in the laboratory or observe essential degradation processes in real-time. Computational modeling and simulation have consequently become indispensable tools in helping to predict the behavior of complex materials across a range of strenuous conditions

At MIT, a new research effort aims to advance the state-of-the-art in predictive simulation as well as shape new interdisciplinary graduate education programs at the intersection of computational science and computer science.

Strengthening engagement with the sciences

The Center for Exascale Simulation of Materials in Extreme Environments (CESMIX) — based at the Center for Computational Science and Engineering (CCSE) within the MIT Stephen A. Schwarzman College of Computing — will bring together researchers in numerical algorithms and scientific computing, quantum chemistry, materials science, and computer science to connect quantum and molecular simulations of materials with advanced programming languages, compiler technologies, and software performance engineering tools, underpinned by rigorous approaches to statistical inference and uncertainty quantification.

“One of the goals of CESMIX is to build a substantive link between computer science and computational science and engineering, something that historically has been hard to do, but is sorely needed,” says Daniel Huttenlocher, dean of the MIT Schwarzman College of Computing. “The center will also provide opportunities for faculty, researchers, and students across MIT to interact intellectually and create a new synthesis of different disciplines, which is central to the mission of the college.”

Leading the project as principal investigator is Youssef Marzouk, professor of aeronautics and astronautics and co-director of CCSE, which was renamed from the Center of Computational Engineering in January to reflect its strengthening engagement with the sciences at MIT. Marzouk, who is also a member of the Statistics and Data Science Center, notes that “CESMIX is trying to do two things simultaneously. On the one hand, we want to solve an incredibly challenging multiscale simulation problem, harnessing quantum mechanical models of complex materials to achieve unprecedented accuracy at the engineering scale. On the other hand, we want to create tools that make development and holistic performance engineering of the associated software stack as easy as possible, to achieve top performance on the coming generation of exascale computational hardware.”

The project involves participation from an interdisciplinary cohort of eight faculty members, serving as co-PIs, and research scientists spanning multiple labs and departments at MIT. The full list of participants includes:

• Youssef Marzouk, PI, professor of aeronautics and astronautics and co-director of CCSE;
• Saman Amarasinghe, co-PI, professor of computer science and engineering;
• Alan Edelman, co-PI, professor of applied mathematics;
• Nicolas Hadjiconstantinou, co-PI, professor of mechanical engineering and co-director of CCSE;
• Asegun Henry, co-PI, associate professor of mechanical engineering;
• Heather Kulik, co-PI, associate professor of chemical engineering;
• Charles Leiserson, co-PI, the Edwin Sibley Webster Professor of Electrical Engineering;
• Jaime Peraire, co-PI, the H.N. Slater Professor of Aeronautics and Astronautics;
• Cuong Nguyen, principal research scientist of aeronautics and astronautics;
• Tao B. Schardl, research scientist in the Computer Science and Artificial Intelligence Laboratory; and
• Mehdi Pishahang, research scientist of mechanical engineering.

MIT was among a total of nine universities selected as part of the Predictive Science Academic Alliance Program (PSAAP) III to form a new center to support science-based modeling and simulation and exascale computing technologies. This is the third time that PSAAP centers have been awarded by the U.S. Department of Energy’s National Nuclear Security Administration (DoE/NNSA) since the program launched in 2008 and is the first time that the Institute has ever been selected. MIT is one of just two institutions nationwide chosen to establish a Single-Discipline Center in this round and will receive up to $9.5 million in funding through a cooperative agreement over five years.

Advancing predictive simulation

CESMIX will focus on exascale simulation of materials in hypersonic flow environments. It will also drive the development of new predictive simulation paradigms and computer science tools for the exascale. Researchers will specifically aim to predict the degradation of complex (disordered and multi-component) materials under extreme loading inaccessible to direct experimental observation — an application representing a technology domain of intense current interest, and one that ­­­exemplifies an important class of scientific problems involving material interfaces in extreme environments.

“A big challenge here is in being able to predict what reactions will occur and what new molecules will form under these conditions. While quantum mechanical modeling will enable us to predict these events, we also need to be able to address the times and length scales of these processes,” says Kulik, who is also a faculty member of CCSE. “Our efforts will be focused on developing the needed software and machine learning tools that tell us when more affordable physical models can address the length scale challenge and when we need quantum mechanics to address the accuracy challenge.”

CESMIX researchers plan on disseminating their results via multiple open-source software projects, engaging their developer and user communities. The project will also support the work of postdocs, graduate students, and research scientists at MIT with the overarching goal of creating new paradigms of practice for the next generation of computational scientists.

de MIT News https://ift.tt/3kXuoZf

Regulating the regulators

MicroRNAs are short RNA sequences that maintain a tight control on which genes are expressed and when. They do this by regulating which messenger RNA (mRNA) transcripts — the single-stranded templates for proteins — are actually read by the cell. But what controls these cellular controllers? 

In a study published Nov. 12 in Science, researchers in David Bartel’s lab at the Whitehead Institute for Biomedical Research show that mRNAs and other RNAs often turn the tables on their microRNA regulators — and show that the path to microRNA degradation is not what scientists expected it to be.

“A lot of people know that microRNAs repress mRNAs — that’s textbook,” says Charlie Shi, a graduate student in Bartel’s lab and first author on the paper. “But in certain cases, this logic is reversed. And I think that's really interesting and weird, this idea that often the tables are turned.” 

When transcripts attack

MicroRNAs typically control gene expression by binding to mRNA transcripts, and then working together with a protein called Argonaute to “silence” those transcripts by causing them to be more rapidly degraded. Because microRNAs are held cozily inside of the Argonaute protein, they are shielded from destructive enzymes in the cell, and are thus fairly long-lived by cellular standards. They can persist for up to a week, causing the destruction of many mRNA molecules over that time.

Sometimes, however, a microRNA binds to a special target site on an mRNA transcript that leads to premature destruction of the microRNA. This phenomenon — called target-directed microRNA degradation, or TDMD — happens naturally in cells, and is a way to control how much of certain microRNAs are allowed to persist at any given time. 

Bartel’s lab began studying this form of degradation after researchers in the lab discovered that an RNA called CYRANO, which doesn’t code for any proteins, leads to the degradation of a specific microRNA called miR-7. This interaction was interesting to the researchers because the mechanism did not seem to line up with the current theories about TDMD. 

Previous models of TDMD suggested that special target sites, like the one in CYRANO, cause one end of the microRNA to stick out of Argonaute and become vulnerable to the addition and subtraction of nucleotides by cytoplasmic enzymes. This process, called tailing and trimming, was thought to be a key step in the path to degradation of the microRNA.

“But when you knock out the enzyme that causes tailing of miR-7, it has no impact on the degradation,” Shi says. “So that's curious, right? So how can we really perturb this supposedly responsible system and have no impact?”

A new model 

In order to further probe the mechanism of TDMD, the researchers focused in on this interaction between the CYRANO noncoding RNA and miR-7. Shi designed a CRISPR screen to identify genes essential for the microRNA’s degradation when it encountered a CYRANO transcript.

The screen yielded one gene that was essential to the microRNA’s degradation, called ZSWIM8. When they looked up the gene’s function, the researchers found that it codes for a component of a ubiquitin ligase. Ubiquitin — so named because it is found in virtually all types of cells — serves as a flag to mark proteins for degradation in a cellular garbage disposal called the proteasome. 

The finding of the ZSWIM8 ubiquitin ligase implied that CYRANO-mediated microRNA degradation involves destruction of the Argonaute protein. In this new molecular model for TDMD, the regulating RNA, CYRANO, binds to the microRNA, mir-7, encased in its protective Argonaute protein, and then recruits the ZSWIM8 ubiquitin ligase. This ligase then sticks a few ubiquitin molecules onto the microRNA’s Argonaute, leading Argonaute to be degraded, and thereby exposing its microRNA cargo to be destroyed by enzymes in the cell. Importantly, this process does not require any trimming and tailing of the microRNA. 

“The discovery of this unanticipated pathway for TDMD illustrates the power of CRISPR screens, which can simultaneously query essentially every protein in the cell, including those that you never dreamed would be involved,” says Bartel, who is also a professor of biology at MIT and  an investigator of the Howard Hughes Medical Institute.

A multitude of microRNAs

When the researchers looked at other known examples of TDMD, they found the ZSWIM8 was essential in all of them. Having identified this key part of the degradation pathway allowed them to seek out more microRNAs that are subject to this regulation. 

“When we started this project, there were only around four examples in nature of endogenous RNAs that are encoded by the cell that can perform TDMD,” Shi says. “We had a feeling that there would be many more, and so by finding a factor that was required for TDMD in a general way — ZSWIM8 — we were then able to ask and answer the question, ‘how widespread this phenomenon?’”

As it turns out, TDMD is fairly common in multicellular organisms. The researchers looked for evidence of the microRNA degradation mechanism in different cell types — two from mice, and one from fruit flies — and found that in any given cell, up to 20 different microRNAs were regulated by TDMD out of a couple hundred total microRNAs in the cell. 

The researchers also observed this mechanism in human cells and nematodes, suggesting that TDMD as a method for regulating microRNAs dates back to the common ancestor of these disparate species. “That definitely creates a lot of questions for us,” Shi says. “Each one of these microRNAs is a story.”

de MIT News https://ift.tt/2IWz1pO

Six MIT faculty elected 2020 AAAS Fellows

Six MIT faculty members have been elected as fellows of the American Association for the Advancement of Science (AAAS).

The new fellows are among a group of 489 AAAS members elected by their peers in recognition of their scientifically or socially distinguished efforts to advance science.

A virtual induction ceremony for the new fellows will be held on Feb. 13, 2021. 

Nazli Choucri is a professor of political science, a senior faculty member at the Center of International Studies (CIS), and a faculty affiliate at the Institute for Data, Science, and Society (IDSS). She works in the areas of international relations, conflict and violence, and the international political economy, with a focus on cyberspace and the global environment. Her current research is on cyberpolitics in international relations, focusing on linking integrating cyberspace into the fabric of international relations.

Catherine Drennan is a professor in the departments of Biology and Chemistry. Her research group seeks to understand how nature harnesses and redirects the reactivity of enzyme metallocenters in order to perform challenging reactions. By combining X-ray crystallography with other biophysical methods, the researchers’ goal is to “visualize” molecular processes by obtaining snapshots of enzymes in action.

Peter Fisher is a professor in the Department of Physics and currently serves as department head. He carries out research in particle physics in the areas of dark matter detection and the development of new kinds of particle detectors. He is also interested in compact energy supplies and wireless energy transmission.

Neil Gershenfeld is the director of MIT's Center for Bits and Atoms, which works to break down boundaries between the digital and physical worlds, from pioneering quantum computing to digital fabrication to the “internet of things.” He is the founder of a global network of over 1,000 fab labs, chairs the Fab Foundation, and leads the Fab Academy.

Ju Li is the Battelle Energy Alliance Professor of Nuclear Science and Engineering and a professor of materials science and engineering. He studies how atoms and electrons behave and interact, to inform the design new materials from the atomic level on up. His research areas include overcoming timescale challenges in atomistic simulations, energy storage and conversion, and materials in extreme environments and far from equilibrium.

Daniela Rus is the Andrew and Erna Viterbi Professor of Electrical Engineering and Computer Science and director of the Computer Science and Artificial Intelligence Laboratory (CSAIL) at MIT. Her research interests include robotics, mobile computing, and data science. Rus is a Class of 2002 MacArthur Fellow, a fellow of ACM, AAAI and IEEE, and a member of the National Academy of Engineering, and the American Academy for Arts and Science.

This year’s fellows will be formally announced in the AAAS News and Notes section of Science on Nov. 27.

de MIT News https://ift.tt/3pWPwTk

Meghan Davis named 2022 Mitchell Scholar

MIT senior Meghan Davis has been named one of the 12 winners of the George J. Mitchell Scholarship’s Class of 2022. After graduating next spring with dual majors in biological engineering and urban planning, she will pursue a master’s in global health at Trinity College in Dublin.

Mitchell Scholars are selected on the basis of academic achievement, leadership, and dedication to public service. The scholarship is named in honor of U.S. Senator Mitchell’s contributions to the Northern Ireland peace process. This year, over 450 American students applied for the prestigious fellowship, which is sponsored by the U.S.-Ireland Alliance and funds a year of graduate studies in Ireland.

Davis, who is the third MIT student to receive this award, was born and raised in Greensboro, North Carolina, and moved to Prosper, Texas, in high school. Her goal is to become a physician-scientist.

An interdisciplinary researcher, Davis is committed to tackling health inequities faced by vulnerable and marginalized communities. Recently, she pursued a mixed-methods approach to understanding the cardiovascular disease disparities in urban Black women and interventions that can be implemented to reduce these disparities. In her current research on cardio-oncology, she is investigating the cellular mechanism of doxorubicin treatment in the laboratory of Professor Laurie Boyer in the Department of Biology. The social side of her research, conducted in the Department of Urban Studies and Planning, focuses on breast cancer in Black women and is being done in collaboration with local community health organizations centered on empowering Black women.

Davis is currently the senior director of BoSTEM Scholars Academy, a program that aims to bridge the racial and socioeconomic gap in STEM education through a five-week summer program for underrepresented Boston area high school students. Davis is also an educator and executive board member of PLEASURE: Peers Leading Education About Sexuality and Standing Up for Relationship Empowerment, an organization that acts to prevent gender-based violence on campus. She was honored earlier this year with the MIT Martin Luther King Jr. Leadership Award for service to the community. She also received MIT's Bridge Builder award for her "strong commitment to and passion for diversity education and cultural celebration."

Professor Boyer states, “Meghan exemplifies leadership in every sense of the word: service, commitment, sacrifice, and motivation. By bringing together her talent for research and passion for working on problems that have an impact on human health particularly relevant to underserved populations, I fully expect that Meghan will light the way for change in science and medicine.  We have all benefited from having Meghan in the Boyer lab and we wish her the very best during her year as a Mitchell Scholar.”

“Meghan embodies so much of what we teach and what we practice in the Department of Urban Studies and Planning; specifically, the importance of asking the right questions, gathering knowledge, and working alongside our marginalized communities,” says Cherie Miot Abbanat, lecturer in international development. “What is so impressive about Meghan’s work is not just her intellect, her persistence, and her depth of questioning, but her passion to make a difference in the lives of Black women, and by extension all of our underrepresented communities. Meghan is dedicated to using science as her weapon to interrogate and root out injustices in our health care systems.”

Davis was supported in the application process by MIT’s Distinguished Fellowships team in Career Advising and Professional Development, and the Presidential Committee on Distinguished Fellowships. “We are proud that Meghan will be representing MIT on the Mitchell Scholarship,” says Kim Benard, assistant dean of distinguished fellowships. “Our entire committee was impressed with her dedication to serving others and examining racial inequities in health care. She has already done impressive work with BoSTEM as a mentor and teacher, as well as research funded by the Eloranta Fellowship to study cardiovascular disease risk in urban Black women. Her time in Ireland will provide her with further knowledge of public health, and we cannot wait to see what the future holds for her.”

de MIT News https://ift.tt/3l1wXd5