Fault detection and diagnosis of automated manufacturing systems

The authors develop a controller methodology for fault detection and diagnosis using Petri nets and fault trees in automated manufacturing systems. The controller has two levels. At the first level there are dedicated diagnostic systems for each of the subsystems, such as machine centers, robots, conveyers, etc. At the second level there is an intelligent controller monitoring the part flow and coordinating the local diagnostic systems and controllers. The authors assume that local controller and diagnostic systems exist for subsystem-level fault detection and diagnosis, and they present a Petri-net-based intelligent controller for system-level fault detection and diagnosis. The authors also describe fault-free-based diagnostics.<<ETX>>