The scientific goals of the project include determining the power of vision as a sensing mode for a robot, how to integrate ongoing sensing, and how much knowledge and reasoning is needed for planning actions. In general, we seek to delve into the interaction between processing, perception, and action in a dynamic environment.
Other goals include learning about capabilities of situated/embedded vision, integration of technology, and exploration of dynamic environments. We wish to develop a flexible computing environment that can handle multiple visual tasks, processing modes, and cameras to support a visually guided robot that can operate outdoors.