Humans teach robots by thought
A research team at Boston University and MIT have developed a brain-computer interface that lets people correct robots’ mistakes by thought in real-time. The system uses electroencephalography (EEG) to monitor a person’s brain signals as they watch a robot work. When it detects a signal suggesting the person has spotted an error, it can correct the robot’s action right away. When we witness a mistake, we generate brain signals called “error potentials”. In a test on five volunteers, the system correctly recognised these signals in 70 percent of cases.
This system gives human collaborators a fast, natural way to interact with robots. For example, passengers of an autonomous car could notify an anomaly that the car sensors may have missed. Communicating using brain activity is so intuitive that it doesn’t need the human to consciously formulate its intention. Humans could then become a passive go-between in the interaction.
Spotted in 2018
- United States