A robot can do almost anything with the right program, and new research is developing technology to control robots through simple thought.
Instead of giving a robot an explicit command for a specific task, researchers at MIT and Boston University have found a way for robots to perform a task that someone is thinking about.
The researchers used electroencephalography (EEG) to monitor human brain activity and detect when the human noticed an error by the robot. The robot then corrected the error.
For the study, robots performed sorting tasks with only two choices, but the discovery has the potential for more complex robotic control.
Instead of a person having to look at something corresponding with the robot's task, the person only has to think the task, the paper's senior author and director of MIT's Computer Science and Artificial Intelligence Laboratory Daniela Rus said.
"Imagine being able to instantaneously tell a robot to do a certain action, without needing to type a command, push a button or even say a word," Rus said in a release.
Rus's team partnered with Frank Guenther's speech neuroscience lab at Boston University for this project.
Guenther's lab had previously looked into brain signals in monkeys when a mistake was going to be made, and they wanted investigate these signals in humans and how it could interact with robots.
"The act of noticing an error can be a very complicated thing that a human does automatically," Guenther told CBC News. "It can be very difficult to program. So we can capitalize on the fact the human brain is watching the whole scene and detects an error."
For practical applications, Guenther said this could help with supervision of robots in factories or for driverless cars because the robot could be programmed to react when the human notices a mistake.
"As you watch the robot, all you have to do is mentally agree or disagree with what it is doing," Rus's statement said. "You don't have to train yourself to think in a certain way — the machine adapts to you, and not the other way around."
Guenther said the natural action of noticing a mistake also makes this technology easy to use, which is important for brain-computer interaction.
In order for more practical applications to be possible, Guenther said the programs will need to be sophisticated enough to manoeuvre through more than two choices.
Another possibility for this technology, Guenther said, is to help people who are severely paralyzed and can't communicate. The available technology is error prone and can be cumbersome to use, he said.
"It's very slow and very effortful and requires a lot of concentration on the user's part," Guenther said. "If we could make it more automatic, that's a great improvement."