A touch-based probe of a coordinate measurement machine (CMM) is generally used to validate the dimensions of the artifacts and associated features which come out of the manufacturing line against its dimensions to ensure to be within prescribed GD&T [Wilson 14] limits. Although there could be other applications using a CMM like reverse engineering, however the stated inspection process is quite crucial for maintaining quality factor and to regulate merits of the manufacturing process especially with the continuous wear-tear of the involved tooling. However this inspection process, which is considered as auxiliary process, needs to have minimum cycle time in order to effectively have more quality units to be produced.
In this paper we describe Computer Aided Process Planning (CAPP) in order to complete Computer Aided Inspection (CAI) process using CMM on the work-piece which is created through Computer Aided Design (CAD) and Manufacturing (CAM). This leads to Computer Integrated Manufacturing (CIM) process. As a result, the features, which need to be inspected, could be recognized from its corresponding CAD file and appropriate information could be culled out by the Dassault Software System to calculate the approach and retract points to inspect the features [Choudhury 03].
The main objective of this research paper is to determine the shortest collision free trajectory from the retract point of one feature being inspected to the approach point of another feature being inspected so that the sequencing for the set of trajectories could be calculated through existing Traveling Sales Person (TSP) algorithm in order to shorten the total distance travelled by the CMM probe, which in turn would linearly reflect in reducing the cycle inspection time. Further since different features in the work-piece might need to be inspected by different probes, an overhead time and displacement for probe change, needs to consider in order to shorten the cycle time. These include clustering of similar type of features, which can be inspected by the same probe. Also clustering of similar type of features need to be assessed against spatial distance of separation between the clusters that the probe has to travel. Since the CMM travels at a constant low speed and the different probes are of similar shape, size and comparatively very small weight, the motion dynamics of the CMM does not influence the total distance travelled for the inspection process.
Our main contribution is in the development of a spatial algorithm which not only reduces the inter-feature distance but also does that by avoiding the potential for any collision with the feature or the artifact without the use of any computer vision or collision avoidance sensor information. The resulting spatial algorithm has implicit embedded information of feature clustering, which when fed into an optimization algorithm generates a path planning which would eventually shorten the cycle time. The illustration is done with two separate simulations.