Navigation of an Intelligent Vision Based Mobile Robot

Period of Performance: 06/30/1993 - 06/30/1995

$488K

Phase 2 SBIR

Recipient Firm

Advanced Technology & Research Corp.
6650 Eli Whitney Drive, Suite 400
Columbia, MD 21046
Principal Investigator

Research Topics

Abstract

The major effort of this research is to create a generic sensory interactive control architecture for autonomous vehicles. There are two primary components in real-time control. One component collects data from the environment through sensors, then processes it into a useful format; the other component analyzes this processed sensory data in the context of the task to control timing and amount of actuator movement. This proposed concept offers a control system solution that is an innovative coordination between a knowledge based component for autonomous vehicle control with multiple sensory devices that report on a complex, non-linear, real world environment. This control can be efficiently implemented on common, low-cost commercial processors. The first element of the system is the Real-time Control System (RCS). RCS will handle task decomposition sequencing, aspects of world modeling, planning and task knowledge based reasoning and control. The sensory processing element will be developed using artificial intelligence algorithms and learning algorithms associated with knowledge based sensory processing and neural networks. Coordinated interaction between vehicle control and sensory processing elements where the executing control task sets the context for the appropriate level of sensory processing at the prsent instant to allow for a robust, incremental approach to the goal state is the unique aspect of this project.