Algorithms

MIT scientists have discovered a new method to maximize wind farm output, reports Gwen Egan for Boston.com. “While there are pros and cons to this strategy, it’s possible that it could allow for smaller wind farms that take up less land to produce more energy,” says Prof. Michael Howland. “It’s critically important we do this now, as we embark on building much more offshore wind. We need to ensure that our future wind farms maximize efficiency to increase the pace of decarbonization.”

MIT scientists have found that changing the angle of turbine blades on wind farms could increase energy output, reports David Abel for The Boston Globe. “Given the scale of wind deployment needed to achieve state and federal climate goals, we need optimal wind farm performance to ensure efficient, rapid decarbonization,” says Prof. Michael Howland. “Our method resulted in significant energy gains over standard industry operations, and, importantly, it can be instituted with minimal cost.”

The MIT mini cheetah broke a speed record after learning to adapt to difficult terrain and upping its speed, reports Rienk De Beer for Popular Mechanics.

MIT researchers have developed a new system that enabled the mini robotic cheetah to learn to run, reports John Koetsier for Forbes. ““Traditionally, the process that people have been using [to train robots] requires you to study the actual system and manually design models,” explains Prof. Pulkit Agrawal. “This process is good, it’s well established, but it’s not very scalable. “But we are removing the human from designing the specific behaviors.”

MIT researchers have developed FuseBot, a new system that combines RFID tagging with a robotic arm to retrieve hidden objects from a pile, reports Brian Heater for TechCrunch. “As long as some objects within the pile are tagged, the system can determine where its subject is most likely located and the most efficient way to retrieve it,” writes Heater.

A study co-authored by MIT researchers finds that algorithms based on clinical medical notes can predict the self-identified race of a patient, reports Katie Palmer for STAT. “We’re not ready for AI — no sector really is ready for AI — until they’ve figured out that the computers are learning things that they’re not supposed to learn,” says Principal Research Scientist Leo Anthony Celi.

CSAIL graduate student Yunzhu Li and his colleagues have trained a robot to use two metal grippers to mold letters out of play dough, reports Jeremy Hsu for New Scientist. "Li and his colleagues trained a robot to use two metal grippers to mould the approximate shapes of the letters B, R, T, X and A out of Play-Doh," explains Hsu. "The training involved just 10 minutes of randomly manipulating a block of the modelling clay beforehand, without requiring any human demonstrations."

TechCrunch reporter Brian Heater spotlights multiple MIT research projects, including MIT Space Exploration Initiative’s TESSERAE, CSAIL’s Robocraft and the recent development of miniature flying robotic drones.

Researchers at MIT and Harvard Medical School have created an artificial intelligence program that can accurately identify a patient’s race based off medical images, reports Tony Ho Tran for The Daily Beast. “The reason we decided to release this paper is to draw attention to the importance of evaluating, auditing, and regulating medical AI,” explains Principal Research Scientist Leo Anthony Celi.

CSAIL researchers have developed a robotic arm equipped with a sensorized soft brush that can untangle hair, reports Douglas Belkin for The Wall Street Journal. “The laboratory brush is outfitted with sensors that detect tension," writes Belkin. “That tension reads as pain and is used to determine whether to use long strokes or shorter ones.”

TechCrunch reporter Kyle Wiggers spotlights how MIT researchers have developed a new computer vision algorithm that can identify images down to the individual pixel. The new algorithm is a “vast improvement over the conventional method of ‘teaching’ an algorithm to spot and classify objects in pictures and videos,” writes Wiggers.

TechCrunch reporter Brian Heater spotlights new MIT robotics research, including a team of CSAIL researchers “working on a system that utilizes a robotic arm to help people get dressed.” Heater notes that the “issue is one of robotic vision — specifically finding a method to give the system a better view of the human arm it’s working to dress.”

MIT researchers have developed reconfigurable, self-assembling robotic cubes embedded with electromagnets that allow the robots to easily change shape, reports John Koetsier for Forbes. “If each of those cubes can pivot with respect to their neighbors you can actually reconfigure your first 3D structure into any other arbitrary 3D structure,” explains graduate student Martin Nisser.

MIT researchers have utilized a new reinforcement learning technique to successfully train their mini cheetah robot into hitting its fastest speed ever, reports Matt Simon for Wired. “Rather than a human prescribing exactly how the robot should walk, the robot learns from a simulator and experience to essentially achieve the ability to run both forward and backward, and turn – very, very quickly,” says PhD student Gabriel Margolis.

MIT researchers have created a new computer algorithm that has allowed the mini cheetah to maximize its speed across varying types of terrain, reports Shi En Kim for Popular Science. “What we are interested in is, given the robotic hardware, how fast can [a robot] go?” says Prof. Pulkit Agrawal. “We didn’t want to constrain the robot in arbitrary ways.”