Scheduling interrelated activities in complex projects under high-order rework: A DSM-based approach

Complex projects, such as product development projects, usually involve myriads of interrelated activities. Identifying an appropriate sequence of these activities is a challenge to project managers because of the existence of rework iterations. Researchers have developed a series of methods to find a schedule that minimizes the first-order rework of interrelated activities using the design structure matrix (DSM). By contrast, this study presents a more accurate approach for determining a sequence that minimizes the high-order rework. First, a new objective function is proposed to describe the first-order and second-order rework time (FSRT) of complex projects. Second, a parallel branch-and-prune algorithm and two local search heuristics are proposed, which can be conveniently used to reduce the FSRT. Experimental results show that we can reduce more rework of complex projects through minimizing FSRT than using other objective functions. Finally, the efficiency of the proposed methods is validated using random experiments.