A schema-based representation for execution monitoring and error recovery in assembly systems

A schema-based approach for representing an assembly plan knowledge is presented. Such an approach can lead to a more flexible way of monitoring the execution of the plan as well as of recovering from an error that may occur during the execution. the high-level description of the plan is based on a causal reasoning model. Using the causal reasoning model it is possible to describe the process plan in terms of its activities and states in the form of a causal net. The schemata are used to represent the activities and states in the net. Using the schema-based approach, the procedure for representing the hierarchical structures of the activities and states inherent in many assembly plans is relatively simple. Furthermore, since validating a state schema and activating an activity schema is done using the object-oriented message passing methodology, the modularity of the assembly plan knowledge can be maintained. The procedure for validating a state schema and activating an activity schema is presented with a simple example