One of the most critical phases in product manufacturing is the design of the assembly process of the product. It affects not only the manufacturing costs (including parts and assembly line), but also the maintenance costs. Often, the assembly process design is led by very experienced engineers. In its early stage, the design is largely a heuristic process. The design is then refined, sometimes manually (again based on experience), and sometimes by employing some computing tools. This question will naturally arise: Can we do something in the refinement so that we have an “optimal” assembly process? This paper attempts to answer this question.

The optimization of an assembly process requires a mathematical model that describes the assembly process in an analytical way. The author forms the mathematical model with a conflict matrix, a connection matrix, and a Berge graph. The optimization is then conducted using this mathematical model. The author describes in detail how the weight of the graph nodes is assigned, and illustrates the flow chart of the computer program, the graphical user interface (GUI) design, and the interaction of the program with modern computer-aided design (CAD) systems. At the end of the paper, the author also discusses the feasibility of integrating the whole optimization process into a virtual reality (VR) system.

This interesting paper is a good reference for CAD software engineers and researchers. While the paper focuses on the optimization of the refinement stage of an assembly design process, it could be extended to cover the early stage (the sketch stage) by discussing how the weight change of the graph nodes would affect the design.