This study investigates a complex scheduling problem involving parallel machines with sequence-dependent setup times, job release dates, and worker-based human resource constraints. Unlike classical scheduling models that consider only machines as limited resources, this problem reflects real industrial settings where both machines and workers must be coordinated simultaneously. Each job can only start after its release date and requires a fixed number of workers from a limited workforce. To ensure workers are not assigned to overlapping jobs, explicit disjunctive constraints are incorporated into the model. Sequence-dependent setup times are also considered, meaning that setup durations vary according to job order. This significantly increases problem complexity while accurately representing industries such as manufacturing and healthcare. The integration of release dates, workforce requirements, and setup-dependent sequencing makes the problem NP-hard. To solve the problem, a mixed-integer linear programming (MILP) model is proposed. The model assigns jobs to machines, defines sequencing relations, and integrates processing, setup, and workforce constraints. The objective minimizes a weighted combination of makespan and total completion time, balancing operational efficiency and flow performance. The commercial solver Gurobi is used to obtain optimal solutions for small and medium-sized instances. For larger problems, a General Variable Neighborhood Search (GVNS) heuristic is developed. The proposed GVNS effectively explores multiple neighborhoods and incorporates release date and workforce constraints. Computational results show that GVNS produces near-optimal solutions with significantly lower computational times, making it suitable for large-scale industrial scheduling applications.

Keywords

Parallel machine scheduling, Sequence-dependent setup times, Workforce allocation, General Variable Neighborhood Search