This page describes the brain-computer interface project at the Neural Systems Lab at the University of Washington. We are working towards enabling sophisticated control and interaction with the environment, using very simple selection-based interfaces that can be operated using brain waves alone. The following people are working on this project:

• Christian Bell
• Rawichote Chalodhorn
• Ravi Kiran
• Pradeep Shenoy
• Prof. Rajesh Rao (advisor)

Figure 1: The robot’s environment has tables and objects that the robot can interact with. The robot uses computer vision to locate the tables and identify the objects on the tables. The picture shows that the robot has identified two objects on the table in front of it.

Figure 2: The discovered objects are sent to the Brain-Computer interface, where the subject focuses on the object they want the robot to pick up. The interface works by flashing a tiny border around each image in succession. When the border around the attended-to image flashes, the user’s brain signals register this event. The brain signals are measured noninvasively from the user's scalp, using EEG electrodes.

Figure 3. The essence of the Brain-computer interface. The user is attending to the green object shown, and we consider two scenarios: when a border around the green object (target) is flashed, and when the border around a different object (non-target) is flashed. The second picture shows how, around 300 milliseconds later, the subject’s brain responds characteristically to the target flash, but not to a non-target flash.

Figure 4: The robot has sophisticated capabilities, for example the ability to pick objects and carry them to a different destination table, also selected by the user via the brain-computer interface.