Flexible project scheduling with diminishing productivity: a data‐driven study in maintenance environments

Abstract Timely completion of maintenance projects is essential for operational efficiency and equipment availability, making schedule reliability a critical concern in maintenance environments. Conventional scheduling models often assume constant resource productivity, yet in practice, productivity decreases as additional workers are assigned to the same task due to crowding and coordination losses. This study develops and empirically validates a practical scheduling approach that explicitly incorporates diminishing productivity into workforce allocation decisions. An optimization‐based formulation and a computationally efficient heuristic are applied to data from 100 real maintenance operations. Results show that conventional scheduling often leads to infeasible or delayed outcomes, whereas, relative to the current‐practice schedule of 148 time units, the proposed diminishing‐productivity heuristic reduces the makespan to 135 time units, corresponding to an approximately 9% reduction. The heuristic solution yields an average cost that is only 0.9% higher than that obtained by the optimal solution. The approach demonstrates practical applicability, achieving near‐optimal performance and producing feasible schedules under real resource constraints. The findings provide a validated decision‐support framework that aids maintenance planners in workforce allocation, coordination management, and schedule feasibility in complex operational environments.