Autonomous mobile robots in flexible manufacturing systems

The authors are exploring and developing techniques to enable mobile robots to work in a less constrained world than that of an AGV (autonomous guided vehicle). By having the robot maintain an a priori world-map, supplementing it as necessary, with sensor-based world perceptions (especially ultrasound and vision), its utility and mobility are enhanced. By maintaining an explicit representation of spatial uncertainty and revision positional estimates using incoming sensor data, the robot does not rely on wire or paint-stripe tethers to reality to navigate. Research is being performed using a Denning research vehicle in a flexible manufacturing system (FMS) environment. AI techniques in knowledge representation are used to embed the knowledge necessary for a robot's achievement of successful goal-oriented behaviors. Relevant behaviors for a mobile robot in an (FMS) environment are being defined.<<ETX>>