Diagnostic expert systems for PLC controlled manufacturing equipment

Although manufacturing productivity could benefit from the implementation of diagnostic expert systems, the costs of developing these systems prohibit their widespread application. To alleviate this problem, a method that automatically generates rules from a PLC program and facts from a PLC data table is explored. The resulting prototype system supports the diagnostics of equipment hardware failures by allowing a technician to traverse an AND/OR tree derived from a problem hypothesis. The tree is pruned on the basis of the relevance of a branch to the problem diagnostics.