This study proposes a quantitative method for determining the scope of system changes in response to requirement modifications in manufacturing. Current design change processes rely heavily on tacit knowledge, raising concerns about skill succession amid labor shortages. We introduce a "Constraint-based SAM (System Architecture Map)" model that represents system interactions through a bipartite graph of attribute and constraint nodes, enabling the analysis of non-independent influence propagation. The method incorporates the concept of "offset capability," where appropriate setting of certain attributes can avoid changes to others. Using constrained NSGA-III algorithm, we derive Pareto-optimal design change processes based on five evaluation metrics: change cost, coordination difficulty, constraint deviation, conflict degree, and loop degree. Validation through egg cooking and front pillar assembly systems demonstrates the method's effectiveness in identifying trade-offs between evaluation metrics and analyzing the impact of fixed requirements and initial change points. This approach enables systematic decision-making in complex system design changes.
SAKATA et al. (Wed,) studied this question.