The proposed technology offers new approaches towards knowledge capture and analysis that enables generation of optimal CNC tool path sequences. The CNC path sequences generated by our proposed expert knowledge base will be as optimal as those produced by human experts, but in significantly less time. The proposed approach is based on use of a process meta-language and visual tools for capturing a descriptive and accurate process model from one or more CNC experts. Using the captured process model, we will create a computer model of the expert and use it to generate a highly pruned game tree. The generated tree will represent all the acceptable tool selections and path strategies deemed 'good' by the expert. Application of a min-max optimization algorithm to this game tree will then produce a near optimal CNC tool selection and path strategy. The tree must be highly pruned and each node should represent a particular tool and its associated path constraints. To determine the actual tree structure, variables, and pruning rules, we will rely on the process model extracted from the CNC expert.
