An Iterative and Interactive Approach for Process Planning

Process planning involves producing machining plans for manufacturing mechanical parts. It requires satisfying a variety of hard and soft geometric, kinematic, tolerance based and economic constraints. Most existing approaches to automating process planning involve doing a global search for a process plan that is optimal with respect to a pre-specified objective function. Such approaches suffer from two important limitations. First, the search space for process plans is too large to facilitate an efficient systematic search. Second and perhaps more important, the evaluation metrics for process plans are very much context dependent, and it is rarely the case that an accurate optimality metric is available a priori. To overcome these limitations, in this paper we propose an iterative framework that continuously searches for a better plan based on interaction with the user. The search is modeled as hill-climbing with the objective function changing as a result of interaction. We will discuss how the approch can be implemented within ASUFTB, a modern feature-based manufacturing system. Batchu (batchu@asu.edu) and Kambhampati (rao@asu.edu) are with Computer Science department, Hirode (hirode@enuxsa.eas.asu.edu) and Shah (jshah@asuvax.eas.asu.edu) are with Mechanical Engineering department. This research is supported in part by NSF research initiation award (RIA) IRI-9210997, NSF young investigator award (NYI) IRI-9457634 and ARPA/Rome Laboratory planning initiative grant F30602-93-C-0039 to Kambhampati.