Automated Process Planning for CNC Machining

A large portion of today’s industrial manufacturing relies on subtractive machining, a process in which a fast-spinning tool successively removes material from raw stock, for example, a block of aluminum, in order to arrive at a part geometry as specified by a design engineer in a computer-aided design (CAD) file. Planning for this process typically involves identifying the sequence of orientations in which the work piece needs to be fixed, identifying the sequence of tools to be used in each orientation, identifying the part of the volume to be removed in each step of each orientation, and identifying the machine to use for each step of each orientation. Until now, this planning process was not automated but done by humans. The Defense Advanced Research Projects Agency (DARPA) recognized that this lack of automation was a source of delays in the design and production of new vehicles and requested proposals to address this issue and to provide automated manufacturing feedback back to designers. At Palo Alto Research Center (PARC), researchers recognized the potential business value to designers as well as manufacturers, and this value proposition was validated during project execution by presenting early prototypes of the software to potential users. The objective of PARC’s uFab project hence was to create a software tool that, given just a CAD file and a representation of available machines and tools, generates a process plan in real time. While work in this area had been done in the 1980s under the name computer-aided process planning (CAPP) (Alting and Zhang 1989), none of the approaches that were pursued then resulted in a fully automated solution. A major shortcoming of these systems was their reliance on features, recognizable configurations of faces on a part such as pockets, slots, and holes, in order to represent states and actions. In these approaches, planning amounted merely to selecting Automated Process Planning for CNC Machining