Secure Variation along a Robot Path using Discrete Tolerance Allocation Optimization

Paper in proceeding, 2008

Very often industrial robots are used in manufacturing processes when there is a high demand on precision. One important process can, for example, be when a robot is used in an assembly line to place parts in tight areas. This implies that it have to exist a collision free path when placing the part or subassembly including the assembly equipment. However, in real production, all equipment, parts and subassemblies are affected by geometrical variation, resulting in conflicts and on-line adjustments of off-line generated assembly paths. Some of these problems can be adjusted by using calibration, but this can be very time consuming. The focus of this paper is to present optimization methods for securing variation along a robot path. The problem is to choose tolerances in a way such that the variation along the robot path is inside a specified requirement on an allowed variation. This requirement shall be decomposed to tolerances/dimensions contributing to the variation along the path. Different tolerances/dimensions are for examples tolerances on joints and actuators. The optimization method is applied on an industrial robot with a given path. The results illustrate how this method can be used to automatically decompose an allowed variation to tolerances/ dimension in an efficient way.