Simulation based Scheduling of multiple change propagations in multistage product development processes

Multiple engineering changes can simultaneously occur in different product development domains or stages, which can cause product delivery delays or development failure. To carefully manage multiple change propagations in multistage product development processes, a digraph-based model, combined with input and output logics, is adopted to represent entities obtained in different product development stages. Dependencies and mapping relationships are used to model inter- and intra- domain entity connections, along which engineering changes can propagate. Mathematical models are developed to compute change impacts and required efforts for resolving changes. Simulation algorithms are present to explore possible change evolution paths across different stages. Genetic algorithm (GA) is employed to find the optimal change propagation paths. A multistage development model representing functional design, structural design and manufacturing process of an air-conditioner is used to test the change scheduling method. Change durations and propagation traces for one, two and three emergent changes are compared to find the best strategy for coping with multiple change requests in multistage product development processes.