Abstract : A major part of our research effort on unmanned autonomous vehicles is the development and fabrication of an aerial platform capable of supporting research on a number of topics, including multi-agent hybrid systems involving sensor fusion, discrete decision making under uncertainty, coordinated mission planning, and distributed control. The foundation of an experimental system on these topics is a dependable autonomous aerial platform that is responsive to requests for basic flight maneuvers such as takeoff, landing, hover, and waypoint navigation. The autonomous aerial platform developed on ARO Grant DAAG55-98-1-0094 at the University of California, Berkeley, consists of reliable aerial vehicles, integrated position and attitude sensors, embedded real-time flight controller and auxiliary computing systems, communication packages, and vision capabilities.