Process knowledge involves the shape producing capability and technological constraints for each of the available machining processes. A variety of knowledge representation techniques are used to model process knowledge, with production rules and frames among the most popular. Production rules involve the notion of condition-action sets, and are often expressed in the from of IF-THEN rules. Examples of such systems are XCUT [152] and AMPS [153]. Frames permit representation of both procedural and declarative information in terms of attributes, hierarchical relations with other frames, constraints, and procedures. SIPS [154] and NEXT-Cut [63] use frames to represent process knowledge.
The process selection task is performed by examining the shape and tolerance requirements of an individual feature and selecting a process that is capable of meeting the requirements. Quite often, a feature needs a roughing operation followed by one or more finishing operations. Backward planning strategies have been successfully used to select the multiple operations needed for certain features. A number of process planing systems, among them AMPS [153], SIPS [154], use this technique to perform process selection.