Robot Emotional Control

The concept of controlling a robot using emotions instead of traditional controls emerged from discussions with my supervisor, Dr. Ehud Sharlin, and my course professor, Dr. Saul Greenberg. We decided to explore the development of a robot that would react accordingly to the way a human feels instead of having to be told what to do. Our idea, which started as a course project, grew into a conference poster presentation, and news media around the world began to notice.

Why is it so new? One main reason. Traditional robots and most computer systems rely on the use buttons, joysticks and other controls to operate. Our prototype works with these too, but it also can be controlled using a headband that reads muscle tension in bioelectrical signals produced directly by the brain. Look ma, no hands! This muscle tension is used to approximate the human emotions of relaxation and stress, so that the robot knows to tune in when the user may want some alone time, move away and look busy.

Related publications

Saulnier, P., Sharlin, E. and Greenberg, S. (2009) Using Bio-electrical Signals to Influence the Social Behaviours of Domesticated Robots. In Adjunct Proc. Human Robot Interaction (Late Breaking Abstracts) - HRI'09. (San Diego, California), 2 pages plus poster, March 11-13. Late-breaking abstract.

Saulnier, P., Sharlin, E. and Greenberg, S. (2009) Using brain-robot interfaces for controlling implicit social patterns. In Workshop on Brain Machine Interfaces for Neuroprostheses and Robot Control, held at the IEEE International Conference on Robotics and Automation. (Kobe, Japan), 3 pages plus poster, May 12. A version of the similar HRI poster/paper above by the same authors.

Press Coverage

View the complete list or click a logo below.

MIT Information Review Engadget Gizmodo Wired UK Citytv CBC Telegraph-Journal