Automatic assembly sequence planning from STEP CAD files

Abstract

In recent years, many research studies related to theoretical and applied assembly sequence planning have been conducted. However, with most prior proposed assembly planning methods, users need to manually enter all or part of the geometrical information needed to find valid assembly sequence plans for a given product design. In this paper, we propose a fully automated assembly sequence planner, which directly extracts geometrical information from an input CAD file and then finds assembly sequences, which require a minimum or near-minimum number of assembly direction reorientations, for a given product design. To make the method compatible with different CAD software packages, the proposed assembly planner takes, as input, a CAD STEP (STandard for the Exchange of Product model data) file. The planner uses a powerful and efficient genetic algorithm to find assembly sequences with a minimum or near-minimum number of reorientations.