Tuesday, November 19th, 2019

Time	Event
12:00a	MIT Energy Initiative report charts pathways for sustainable personal transportation In our daily lives, we all make choices about how we travel and what type of vehicle we own or use. We consider these choices within the constraints of our current transportation system and weigh concerns including costs, convenience, and — increasingly — carbon emissions. "Insights into Future Mobility," a multidisciplinary report released today by the MIT Energy Initiative (MITEI), explores how individual travel decisions will be shaped by complex interactions between technologies, markets, business models, government policies, and consumer preferences — and the potential consequences as personal mobility undergoes tremendous changes in the years ahead. The report is the culmination of MITEI’s three-year Mobility of the Future study, which is part of MIT’s Plan for Action on Climate Change. The report highlights the importance of near-term action to ensure the long-term sustainability of personal mobility. The researchers ultimately find that continued technological innovation is necessary and must be accompanied by cross-sector policies and changes to consumer behavior in order to meet Paris Agreement targets for greenhouse gas emissions reductions. “Understanding the future of personal mobility requires an integrated analysis of technology, infrastructure, consumer choice, and government policy,” says MITEI Director Robert C. Armstrong, a professor of chemical engineering at MIT. “The study team has examined how these different dimensions will develop and interact, and the report offers possible pathways toward achieving a more sustainable personal transportation system.” The study team of MIT faculty, researchers, and students focused on five main areas of inquiry. They investigated the potential impact of global climate policies on fleet composition and fuel consumption, and the outlook for vehicle ownership and travel, with a focus on the U.S. and China. They also researched characteristics and future market share of alternative fuel vehicles, including plug-in electric and hydrogen fuel cell vehicles, and infrastructure considerations for charging and fueling, particularly as they affect future demand. Another main area of focus was the future of urban mobility, especially the potentially disruptive role of ride-hailing services and autonomous vehicles. The researchers find that there is considerable opportunity for reducing emissions from personal mobility by improving powertrain efficiency and deploying alternative fuel vehicles in the coming decades. These changes must be accompanied by decarbonization of the production of the fuels and electricity that power these vehicles in order to reach global emissions mitigation targets and achieve cleaner air and other environmental and human health benefits. “Our analysis shows that reducing the carbon intensity of the light-duty vehicle fleet contributes to climate change mitigation goals, as part of the larger solution,” says Sergey Paltsev, deputy director of the MIT Joint Program on the Science and Policy of Global Change and senior research scientist at MITEI. “If we are to reach international goals for limiting temperature rise and other climate change-related impacts, we will need comprehensive climate policies that promote the adoption of alternative fuel vehicles in the transportation sector and simultaneously decarbonize the electricity sector.” Several factors influence an individual’s decision to adopt an alternative fuel vehicle, such as a battery electric vehicle. The researchers found that the most important, interrelated factors that impact alternative vehicle adoption include cost, driving range, and charging convenience. They conclude that as production volumes increase, battery costs and the purchase price of electric vehicles will decrease, which will in turn drive sales. Improved batteries would extend the vehicle range, reinforcing the attractiveness of alternative fuel vehicles to consumers. Greater deployment of electric vehicles creates a larger market for publicly available charging infrastructure, which is critical for supporting charging convenience. Early government support for alternative fuel vehicles and charging and fueling infrastructure can help launch a self-reinforcing trajectory of adoption — and has already contributed to an increase in alternative fuel vehicle deployment. “We found that substantial uptake of battery electric vehicles is likely and that the extent and speed of this transition to electrification is sensitive to evolving battery costs, availability of charging infrastructure, and policy support,” says William H. Green, a professor of chemical engineering at MIT and the study chair. This large-scale deployment of battery electric vehicles is expected to help them reach total cost-of-ownership parity with internal combustion engine vehicles in approximately 10 years in the U.S. It should also lead to new business opportunities, including solutions for developing cost-effective methods of recycling batteries on an industrial scale. The researchers also examined the role of consumer attitudes toward car ownership and use in both established and emerging economies. In the U.S., the researchers analyzed trends in population and socioeconomic factors to estimate future demand for vehicles and vehicle travel. While many have argued that lower car ownership and use among millennials may lead to a reduced personal vehicle fleet in coming decades, the study team found that generational differences could be completely explained by differences in socioeconomics — meaning that there is no significant difference in preferences for vehicle ownership or use between millennials and previous generations. Therefore, the stock of light-duty vehicles and number of vehicle-miles traveled will likely increase by approximately 30 percent by 2050 in the U.S. In addition, the analysis indicates that “car pride” — the attribution of social status and personal image to owning and using a car — has an effect on car ownership as strong as that of income. An analysis of car pride across countries revealed that car pride is higher in emerging vehicle markets; among established markets, car pride is highest in the U.S. The adoption of new technologies and business models for personal mobility at scale will require major shifts in consumer perceptions and behaviors, notes Joanna Moody, research program manager of MITEI’s Mobility Systems Center and a coordinating author of the report. “Symbolic and emotional attachments to car ownership and use, particularly among individuals in emerging economies, could pose a significant barrier to the widespread adoption of more sustainable alternatives to privately owned vehicles powered by petroleum-based fuels,” Moody says. “We will need proactive efforts through public policy to establish new social norms to break down these barriers.” The researchers also looked at China, the largest market for new vehicle sales, to analyze how cities form transportation policies and to estimate how those local-level policies might impact the future size of China’s vehicle stock. To date, six major Chinese cities and one province have implemented car ownership restriction policies in response to severe congestion and air pollution. Our researchers found that if the six megacities continue with these restrictions, the country’s light-duty vehicle fleet could be 4 percent (12 million vehicles) smaller by 2030 than it would be without these restrictions. If the policies are adopted in more of China’s cities facing congestion and air pollution challenges, the fleet could be up to 10 percent (32 million vehicles) smaller in 2030 than it would be without those restrictions. Finally, the team explored how the introduction of low-cost, door-to-door autonomous vehicle (AV) mobility services will interact with existing modes of transportation in dense cities with incumbent public transit systems. They find that introducing this low-cost mobility service without restrictions can lead to increased congestion, travel times, and vehicle miles traveled — as well as reduced public transit ridership. However, these negative impacts can be mitigated if low-cost mobility services are introduced alongside policies such as “first/last mile” policies (using AVs to transport riders to and from public transit stations) or policies that reduce private vehicle ownership. The findings apply even to cities with vastly different levels of public transit service. Building on the research started under the Mobility of the Future study, MITEI has now launched a new Low-Carbon Energy Center, the Mobility Systems Center. Approaching mobility from a sociotechnical perspective, the center identifies key challenges, investigates current and potential future trends, and analyzes the societal and environmental impacts of emerging solutions for global passenger and freight mobility. The Mobility of the Future study received support from an external consortium of international companies with expertise in various aspects of the transportation sector, including energy, vehicle manufacturing, and infrastructure. The report, its findings, and analyses are solely the work of the MIT researchers. For more information or to access the "Insights into Future Mobility" report, visit energy.mit.edu/insightsintofuturemobility. (Comment on this)
12:53p	Technique identifies T cells primed for certain allergies or infections When your immune system is exposed to a vaccine, an allergen, or an infectious microbe, subsets of T cells that can recognize a foreign intruder leap into action. Some of these T cells are primed to kill infected cells, while others serve as memory cells that circulate throughout the body, keeping watch in case the invader reappears. MIT researchers have now devised a way to identify T cells that share a particular target, as part of a process called high-throughput single-cell RNA sequencing. This kind of profiling can reveal the unique functions of those T cells by determining which genes they turn on at a given time. In a new study, the researchers used this technique to identify T cells that produce the inflammation seen in patients with peanut allergies. In work that is now underway, the researchers are using this method to study how patients’ T cells respond to oral immunotherapy for peanut allergies, which could help them determine whether the therapy will work for a particular patient. Such studies could also help guide researchers in developing and testing new treatments. “Food allergies affect about 5 percent of the population, and there’s not really a clear clinical intervention other than avoidance, which can cause a lot of stress for families and for the patients themselves,” says J. Christopher Love, the Raymond A. and Helen E. St. Laurent Professor of Chemical Engineering and a member of MIT’s Koch Institute for Integrative Cancer Research. “Understanding the underlying biology of what drives these reactions is still a really critical question.” Love and Alex K. Shalek, who is the Pfizer-Laubach Career Development Associate Professor at MIT, an associate professor of chemistry, a core member of MIT’s Institute for Medical Engineering and Science (IMES), and an extramural member of the Koch Institute, are the senior authors of the study, which appears today in Nature Immunology. The lead authors of the paper are graduate student Ang Andy Tu and former postdoc Todd Gierahn. Extracting information The researchers’ new method builds on their previous work developing techniques for rapidly performing single-cell RNA sequencing on large populations of cells. By sequencing messenger RNA, scientists can discover which genes are being expressed at a given time, giving them insight into individual cells’ functions. Performing RNA sequencing on immune cells, such as T cells, is of great interest because T cells have so many different roles in the immune response. However, previous sequencing studies could not identify populations of T cells that respond to a particular target, or antigen, which is determined by the sequence of the T cell receptor (TCR). That’s because single-cell RNA sequencing usually tags and sequences only one end of each RNA molecule, and most of the variation in T cell receptor genes is found at the opposite end of the molecule, which doesn’t get sequenced.  “For a long time, people have been describing T cells and their transcriptome with this method, but without information about what kind of T cell receptor the cells actually have,” Tu says. “When this project started, we were thinking about how we could try to recover that information from these libraries in a way that doesn’t obscure the single-cell resolution of these datasets, and doesn’t require us to dramatically change our sequencing workflow and platform.” In a single T cell, RNA that encodes T cell receptors makes up less than 1 percent of the cell’s total RNA, so the MIT team came up with a way to amplify those specific RNA molecules and then pull them out of the total sample so that they could be fully sequenced. Each RNA molecule is tagged with a barcode to reveal which cell it came from, so the researchers could match up the T cells’ targets with their patterns of RNA expression. This allows them to determine which genes are active in populations of T cells that target specific antigens. “To put the function of T cells into context, you have to understand what it is they’re trying to recognize,” Shalek says. “This method lets you take existing single-cell RNA sequencing libraries and pull out relevant sequences you might want to characterize. At its core, the approach is a straightforward strategy for extracting some of the information that’s hidden inside of genome-wide expression profiling data.” Another advantage of this technique is that it doesn’t require expensive chemicals, relies on equipment that many labs already have, and can be applied to many previously processed samples, the researchers say. Analyzing allergies In the Nature Immunology paper, the researchers demonstrated that they could use this technique to pick out mouse T cells that were active against human papilloma virus, after the mice had been vaccinated against the virus. They found that even though all of these T cells reacted to the virus, the cells had different TCRs and appeared to be in different stages of development — some were very activated for killing infected cells, while others were focused on growing and dividing. The researchers then analyzed T cells taken from four patients with peanut allergies. After exposing the cells to peanut allergens, they were able to identify T cells that were active against those allergens. They also showed which subsets of T cells were the most active, and found some that were producing the inflammatory cytokines that are usually associated with allergic reactions. “We can now start to stratify the data to reveal what are the most important cells, which we were not able to identify before with RNA sequencing alone,” Tu says. Love’s lab is now working with researchers at Massachusetts General Hospital to use this technique to track the immune responses of people undergoing oral immunotherapy for peanut allergies — a technique that involves consuming small amounts of the allergen, allowing the immune system to build up tolerance to it. In clinical trials, this technique has been shown to work in some but not all patients. The MIT/MGH team hopes that their study will help identify factors that could be used to predict which patients will respond best to the treatment. “One would certainly like to have a better sense of whether an intervention is going to be successful or not, as early as possible,” Love says. This strategy could also be used to help develop and monitor immunotherapy treatments for cancer, such as CAR-T cell therapy, which involves programming a patient’s own T cells to target a tumor. Shalek’s lab is also actively applying this technique with collaborators at the Ragon Institute of MGH, MIT and Harvard to identify T cells that are involved in fighting infections such as HIV and tuberculosis. The research was funded by the Koch Institute Support (core) Grant from the National Institutes of Health, the Koch Institute Dana-Farber/Harvard Cancer Center Bridge Project, the Food Allergy Science Initiative at the Broad Institute of MIT and Harvard, the Arnold and Mabel Beckman Foundation, a Searle Scholar Award, a Sloan Research Fellowship in Chemistry, the Pew-Stewart Scholars program, and the National Institutes of Health. (Comment on this)
11:59p	Bot can beat humans in multiplayer hidden-role games MIT researchers have developed a bot equipped with artificial intelligence that can beat human players in tricky online multiplayer games where player roles and motives are kept secret. Many gaming bots have been built to keep up with human players. Earlier this year, a team from Carnegie Mellon University developed the world’s first bot that can beat professionals in multiplayer poker. DeepMind’s AlphaGo made headlines in 2016 for besting a professional Go player. Several bots have also been built to beat professional chess players or join forces in cooperative games such as online capture the flag. In these games, however, the bot knows its opponents and teammates from the start. At the Conference on Neural Information Processing Systems next month, the researchers will present DeepRole, the first gaming bot that can win online multiplayer games in which the participants’ team allegiances are initially unclear. The bot is designed with novel “deductive reasoning” added into an AI algorithm commonly used for playing poker. This helps it reason about partially observable actions, to determine the probability that a given player is a teammate or opponent. In doing so, it quickly learns whom to ally with and which actions to take to ensure its team’s victory. The researchers pitted DeepRole against human players in more than 4,000 rounds of the online game “The Resistance: Avalon.” In this game, players try to deduce their peers’ secret roles as the game progresses, while simultaneously hiding their own roles. As both a teammate and an opponent, DeepRole consistently outperformed human players. “If you replace a human teammate with a bot, you can expect a higher win rate for your team. Bots are better partners,” says first author Jack Serrino ’18, who majored in electrical engineering and computer science at MIT and is an avid online “Avalon” player. The work is part of a broader project to better model how humans make socially informed decisions. Doing so could help build robots that better understand, learn from, and work with humans. “Humans learn from and cooperate with others, and that enables us to achieve together things that none of us can achieve alone,” says co-author Max Kleiman-Weiner, a postdoc in the Center for Brains, Minds and Machines and the Department of Brain and Cognitive Sciences at MIT, and at Harvard University. “Games like ‘Avalon’ better mimic the dynamic social settings humans experience in everyday life. You have to figure out who’s on your team and will work with you, whether it’s your first day of kindergarten or another day in your office.” Joining Serrino and Kleiman-Weiner on the paper are David C. Parkes of Harvard and Joshua B. Tenenbaum, a professor of computational cognitive science and a member of MIT’s Computer Science and Artificial Intelligence Laboratory and the Center for Brains, Minds and Machines. Deductive bot In “Avalon,” three players are randomly and secretly assigned to a “resistance” team and two players to a “spy” team. Both spy players know all players’ roles. During each round, one player proposes a subset of two or three players to execute a mission. All players simultaneously and publicly vote to approve or disapprove the subset. If a majority approve, the subset secretly determines whether the mission will succeed or fail. If two “succeeds” are chosen, the mission succeeds; if one “fail” is selected, the mission fails. Resistance players must always choose to succeed, but spy players may choose either outcome. The resistance team wins after three successful missions; the spy team wins after three failed missions. Winning the game basically comes down to deducing who is resistance or spy, and voting for your collaborators. But that’s actually more computationally complex than playing chess and poker. “It’s a game of imperfect information,” Kleiman-Weiner says. “You’re not even sure who you’re against when you start, so there’s an additional discovery phase of finding whom to cooperate with.” DeepRole uses a game-planning algorithm called “counterfactual regret minimization” (CFR) — which learns to play a game by repeatedly playing against itself — augmented with deductive reasoning. At each point in a game, CFR looks ahead to create a decision “game tree” of lines and nodes describing the potential future actions of each player. Game trees represent all possible actions (lines) each player can take at each future decision point. In playing out potentially billions of game simulations, CFR notes which actions had increased or decreased its chances of winning, and iteratively revises its strategy to include more good decisions. Eventually, it plans an optimal strategy that, at worst, ties against any opponent. CFR works well for games like poker, with public actions — such as betting money and folding a hand — but it struggles when actions are secret. The researchers’ CFR combines public actions and consequences of private actions to determine if players are resistance or spy. The bot is trained by playing against itself as both resistance and spy. When playing an online game, it uses its game tree to estimate what each player is going to do. The game tree represents a strategy that gives each player the highest likelihood to win as an assigned role. The tree’s nodes contain “counterfactual values,” which are basically estimates for a payoff that player receives if they play that given strategy. At each mission, the bot looks at how each person played in comparison to the game tree. If, throughout the game, a player makes enough decisions that are inconsistent with the bot’s expectations, then the player is probably playing as the other role. Eventually, the bot assigns a high probability for each player’s role. These probabilities are used to update the bot’s strategy to increase its chances of victory. Simultaneously, it uses this same technique to estimate how a third-person observer might interpret its own actions. This helps it estimate how other players may react, helping it make more intelligent decisions. “If it’s on a two-player mission that fails, the other players know one player is a spy. The bot probably won’t propose the same team on future missions, since it knows the other players think it’s bad,” Serrino says. Language: The next frontier Interestingly, the bot did not need to communicate with other players, which is usually a key component of the game. “Avalon” enables players to chat on a text module during the game. “But it turns out our bot was able to work well with a team of other humans while only observing player actions,” Kleiman-Weiner says. “This is interesting, because one might think games like this require complicated communication strategies.” Next, the researchers may enable the bot to communicate during games with simple text, such as saying a player is good or bad. That would involve assigning text to the correlated probability that a player is resistance or spy, which the bot already uses to make its decisions. Beyond that, a future bot might be equipped with more complex communication capabilities, enabling it to play language-heavy social-deduction games — such as a popular game “Werewolf” —which involve several minutes of arguing and persuading other players about who’s on the good and bad teams. “Language is definitely the next frontier,” Serrino says. “But there are many challenges to attack in those games, where communication is so key.” (Comment on this)

<< Previous Day

2019/11/19 [Calendar]

Next Day >>