Ah yes, the beauty of robots. Whether it be making our lives a thousand times easier or their potential to take over the world, humans have always been fascinated by these machines we have created. Currently, the only way we can communicate with robots and tell them what to do is by programming them or inputting certain commands, which is good, but not great if you want them to do what you tell them to do as soon as possible. That’s what a team of researchers at the Massachusetts Institute of Technology Computer Science and Artificial Intelligence Laboratory (CSAIL) are working on: Robots that we can control with our mind (sort of).
The researchers did this by creating a feedback system that allows robots to correct their mistakes. This system monitors your brain activity via an EEG monitor, in which the machine continuously scans for error-related potentials (ErrPs): brain signals that are generated in our brain whenever we notice a mistake. The robot will then sense these signals and correct their actions.
Source: Massachusetts Institute of Technology Computer Science and Artificial Intelligence Laboratory
To test this, the researchers used a multipurpose robot to sort out a bunch of items. They were either wires or cans of paint, and it was the robot’s job to put each item in its respective box. If a human connected to the EEG monitor notices that the robot is putting a paint can in the box where the wires are supposed to go, the monitor immediately picks up an ErrP signal. Once the robot detects this signal, it immediately corrects its actions and puts the paint can in the paint box instead.
Source: Youtube, MITCSAIL
What’s even better is that this works in real-time, all thanks to the system the team created, which detects these brain signals in just 10-20 milliseconds. “As you watch the robot, all you have to do is mentally agree or disagree with what it is doing” says Daniela Rus, MIT-CSAIL director and senior author of the team’s research paper. “You don’t have to train yourself to think in a certain way — the machine adapts to you, and not the other way around.”
This research was so successful, it was accepted to enter the IEEE International Conference on Robotics and Automation (ICRA), in which they will present their paper on May.
While this is still a long way from fully controlling robots with our minds, this is a very promising start. “Given how difficult it can be to translate human language into a meaningful signal for robots, work in this area could have a truly profound impact on the future of human-robot collaboration.” Comments Wolfram Burgard, a Computer Science Professor at the University of Washington. This technology indeed brings us one step closer from achieving that goal, and who knows where it’ll go next?