Applying cluster identification algorithm and simulation to generate probabilistic network schedules for design projects

Scheduling of a design project is complex because design activities often have information dependencies between each other. This study proposes a network-based model to schedule design projects and generate probabilistic project durations. The proposed model applies a modified cluster identification algorithm to evaluate information dependencies between design activities to facilitate the establishment of a schedule network (and regroup activities to support the assignment of design work); it also uses a simulation approach to incorporate the effect on duration of the uncertain number of design iterations. The model is implemented in four stages, which are breaking down the design work; evaluating the dependencies; identifying concurrent activities; and estimating the durations of activities and simulating the duration of design project. The advantages of the proposed model are demonstrated through its application to an example project, which was reviewed by industrial practitioners. Practitioners felt that the generated detailed scheduling data could help them to control their design work more precisely than a bar chart. Additionally, the simulated probabilistic project duration provided them with an awareness of the risk involved in meeting the contractual deadline.