A Practical Approach for Resource-Constrained Project Scheduling

Abstract

This work considers project scheduling and planning for the decision support of organizations, that continuously require the implementation of many projects for their successful operation. The efficient scheduling and planning of these projects is essential for the timely completion of all projects, utilizing the appropriate resources. To this end, this work presents a novel integer linear program (ILP) formulation, that takes into account the project requirements, the involved teams, their interdependencies, as well as other constraints, so as to provide an optimal scheduling plan. Moreover, additional constraints are considered to address practical challenges such as complexity and uncertainties. Finally, techniques are introduced in this work to address scalability issues, as well as dynamic changes that may occur when the obtained schedule is currently being implemented. All aforementioned techniques present a number of advantages for an organization, as they reduce considerably the person-hours required by the management team to perform the scheduling, they produce scheduling plans that span large planning horizons, and they decrease the project completion times, thus reducing the cost that the organization incurs for implementing the projects. Realistic scenarios are considered, where real data on projects and teams are taken into account. From the results obtained, it is evident that the proposed ILP can obtain the optimal solution in terms of minimizing the duration for the completion of all projects, while the proposed practical (heuristic) approaches, can obtain solutions close to the optimal in terms of planning horizon and objective score with significant reduction in computation time, from hours to seconds/minutes. Moreover, it is shown that the scheduling plan can be adapted in the event of miscalculations related to the effort required for implementing the projects or new projects can be added within the existing scheduling plan.