Simulation-Based Multiagent Approach for Scheduling Modular Construction

Modular construction is a common practice for building industrial plants, particularly in the oil sands region of Alberta, Canada. Each module is a project with its own activities and constraints. These modules are prefabricated offsite, at a location called the module assembly yard, and then shipped to the site. Effective scheduling of modules of an industrial plant involves developing a realistic schedule that makes best use of limited resources in the yard while satisfying the constraints and uncertainties of the entire project. Scheduling such large-scale, multiunit projects using commercial CPM-based scheduling applications (e.g., Primavera, Microsoft Project) is not effective. In previous work, we have introduced a hybrid framework, called simulation-based auction protocol (SBAP), for effective resource scheduling in large-scale projects. The present study employs the SBAP framework for effective allocation of resources (e.g., space, skilled crew) and for satisfaction of various constraints. This system pulls data from a comprehensive database, runs the simulation model behind the scenes, and generates various graphical reports to aid superintendents and project managers in pertinent project decisions. The developed system is also capable of scheduling the fast-track modular construction projects with limited data available, doing effective resource leveling, and scheduling resources (e.g., space, crew) effectively based on various shifts and calendars. The described case study in this paper demonstrates the capabilities of the developed system for planning the module assembly yards.