Anthony Cuthbertson of Newsweek reports that PhD student Claudia Pérez D’Arpino has developed a system that allows robots to learn a skill and teach it to another robot. Armed with knowledge of how to perform a task, a 3-D interface demonstrates the tasks “allowing [the robot] to understand the motions it is being taught in the real world,” explains Cuthbertson.
This CNN video highlights a new system developed by CSAIL researchers that allows noncoders to teach robots to perform a task after a single demonstration. The new programming method also enables robots to learn from other robots, which could enable “a variety of robots to perform similar tasks.”
Wired reporter Matt Simon writes that CSAIL researchers have developed a new system that allows noncoders to be able to teach robots a wide range of tasks, and enables robots to transfer new skills to other robots. Simon notes that the development is a “glimpse into a future where, more and more, robots communicate without humans at all.”
NuTonomy, an MIT startup, will soon start testing self-driving cars in Boston’s Seaport District and Fort Point areas, writes Doug Newcomb for Forbes.
Using an algorithm and a 3-D scan of a human body, Prof. Carlo Ratti designed a foldable stool that can conform to different shapes, writes Margaret Rhodes for Wired. Allowing algorithms to determine the form of each hinge, “allows you to create shapes and functions that you otherwise wouldn’t be able to,” Ratti explains.
BBC News reporter Zoe Kleinman writes that graduate student Joy Buolamwini has developed an initiative aimed at tackling algorithmic bias. "If we are limited when it comes to being inclusive that's going to be reflected in the robots we develop or the tech that's incorporated within the robots,” says Buolamwini.
CSAIL researchers have developed software that allows users to design and virtually test drones, writes Alex Davies for Wired. Graduate student Tao Du explains that the software can help users “explore and try different shapes and different controllers.”
Prof. Daniela Rus speaks to the BBC’s Gareth Mitchell about the robots developed by CSAIL that can modify their behavior based on brain waves detected by a human operator. “We imagine operating prosthetic devices, a wheelchair, even autonomous vehicles,” says Prof. Rus.
A study co-authored by Prof. Carlo Ratti finds ride-sharing is feasible in different cities around the world, reports Eleanor Cummins for Popular Science. “The key thing is how this can transform our cities. Every car you remove from the road, you are removing pollution and traffic,” explains Ratti.
CSAIL researchers have developed a system that allows robots to correct their mistakes based on input from the brainwaves of human operators, reports Wired’s Matt Simon. “It’s a new way of controlling the robot,” explains Prof. Daniela Rus, “in the sense that we aim to have the robot adapt to what the human would like to do.”
Anthony Cuthbertson of Newsweek writes that CSAIL researchers have developed a system that allows robots to change their actions based on feedback from the brain waves of a human operator. “Imagine robots or smartphones that could immediately correct themselves when you realize they’re making a mistake,” says PhD candidate Joseph DelPreto.
A feedback system developed by CSAIL researchers allows humans to correct a robot’s mistakes using brain signals, writes Janet Burns for Forbes. The system could be used as a “communication method for those who can't use verbal means, such as immobilized or even 'locked in' victims of paralysis,” explains Burns.
MIT researchers have developed a device that allows humans to guide robots using brainwaves, reports Clive Cookson for the Financial Times. The prototype brain-computer interface “enables a human observer to transmit an immediate error message to a robot, telling it to fix a mistake when it does something wrong.”
New Scientist reporter Matt Reynolds writes that MIT researchers have developed a new brain-computer interface that enables people to correct robots’ mistakes using brain signals. “We’re taking baby steps towards having machines learn about us, and having them adjust to what we think,” explains Prof. Daniela Rus, director of CSAIL.
In this video, TechCrunch explores how MIT researchers are designing new techniques to make it easier for robots to grasp and manipulate objects. Graduate student Nikhil Chavan-Dafle explains that he and his colleagues developed a model that allows robots to “predict how an object is going to move in the grasp.”