Game theory to help robots interact safely with humans

Researchers have for the first time used game theory to programme a robot that can understand its human user's behaviour in order to better anticipate their movements and respond to them.

Game theory is commonly used to understand how economic agents decide and interact with each other in order to maximise their own gain.

While game theory has had important impacts in economics during the last century, "to apply it for human-robot interaction, it was necessary to understand how the robot can identify the human user's control goals simultaneously to smoothly interacting with them", said Etienne Burdet, Professor at the Imperial College, London.

The study showed that the reactive robotic programming system enables a robot to continuously learn the human user's control and adapt its own control correspondingly.

It is also able to understand the human user's action and then respond to and assist them to perform tasks successfully with minimal effort.

The findings, published in the journal Nature Machine Intelligence, could help robots complementing humans for sport training, physical rehabilitation or shared driving, the researchers said.

"It is still very early days in the development of robots and at present, those that are used in a working capacity are not intuitive enough to work closely and safely with human users," said lead author Yanan Li, Lecturer at the University of Sussex in the UK.

"By enabling the robot to identify human users' behaviour and exploiting game theory to let the robot optimally react to them, we have developed a system where robots can work along humans as humans do," Li added.

For the study, the team used adaptive control and Nash equilibrium game theory and outlined the way they adapted game theory for the physical interaction of a robot with a human.

They also analysed how this could be used to help an impaired stroke survivor retrain their motor control.