3D Model-Based Non-interference Assembly Sequence Generation for Products with a Large Number of Parts

Current approaches to generating product assembly sequences with given assembly constraints require manual input. To reduce human intervention in the process, in this paper, we propose an automated sequence planning method using only a STEP file, a kind of 3D CAD file. The automated process starts by investigating geometric information about the assembled product, disassembling it into individual parts, and examining interferences between the parts that occur in the assembly process. To find the assembly sequence of many parts with the least number of direction changes, we use a genetic algorithm and provide a method to select the initial gene based on the interferences rather than merely generating randomized initial genes. The proposed method can generate an assembly sequence even for an assembled product with 32 parts.