This research has afforded efficient and robust navigational techniques that are being explored in a diversity of domains: manufacturing environments, aerospace and undersea applications, campus settings, military scenarios, nuclear waste management, etc. The emphasis is on generalizable, flexible methods for intelligent robotic control. Modularization of behaviors and perceptual strategies affords computationally efficient solutions to navigation in complex and unpredictable domains. A high-level goal is to produce survivable robotic systems capable of fitting into a particular ecological niche and successfully competing and cooperating with other environmental agents.
Some areas of recent research activity include: cooperation, communication, and mission specification in reactive multiagent robotic systems; ecological robotic systems; unmanned aerial vehicles; usable autonomous agents; human-robot interaction and robot ethics; coordinated control of a mobile manipulator using a hybrid reactive/deliberative architecture; and motor behavior learning using genetic algorithms, case-based reasoning, and adaptive on-line methods.