Model for the diagnosis of CIM equipment

Abstract This paper presents a model for the diagnosis of CIM (computer integrated manufacturing) equipment. The model uses the deep knowledge of state transitions in the manufacturing process, for fault isolation and determination of the causal agent. This requires the use of valid sensory information. The model utilizes a sensor validation approach which incorporates diagnostic expectation and exclusion techniques. Furthermore a generic mechanism classification is used to reduce the effort in encoding test procedures for each subsystem. The model is developed in details using a workcell with a SCARA robot performing pick-and-place operations. The implementation of the model is done using Intellicorp's Knowledge Engineering Environment on a Symbolics 3640 computer.