A Reactive Behavioral System for the Intelligent Room

Traditional computing interfaces have drawn users into a world of windows, icons and pointers. Pervasive computing believes that human-computer interaction (HCI) should be more natural: computers should be brought into our world of human discourse. The Intelligent Room project shares this vision. We are building an Intelligent Environment (IE) to allow for more natural forms of HCI. We believe that to move in this direction, we need our IE to respond to more than just direct commands; it also needs to respond to implicit user commands, such as body language, behavior, and context, just as another human would. This thesis presents ReBa, a context-aware system to provide the IE with this type of complex behavior, in reaction to user activity. Thesis Supervisor: Dr. Howard Shrobe Title: Principal Research Scientist Thesis Supervisor: Prof. Brian Williams Title: Associate Professor