Hybrid AHP - TOPSIS and Constraint Programming for Maintenance Prioritisation and Scheduling in Production Systems

Abstract Efficient maintenance planning is essential for sustaining productivity and minimizing downtime in industrial production. It ensures equipment reliability, cost efficiency, and stable production flows within modern manufacturing systems. Traditional scheduling approaches often depend on manual expertise and static rules, which limits their scalability in dynamic industrial environments. At the same time, modern manufacturing practices are becoming increasingly complex, as maintenance planning, equipment, production schedules, and resource management are tightly interlinked. Maintaining reliability while reducing costs and downtime poses a significant challenge, especially under Industry 4.0, where digital integration enables more reliable and data-driven decision-making. Existing studies overlook practical constraints like technician availability, cost variations across shifts, and the alignment of maintenance with production schedules and resources. This gap often results in decision-making processes that are either theoretically sound but operationally impractical, or operationally feasible but lacking robust prioritization logic. This paper proposes a hybrid decision support system that integrates the Analytical Hierarchy Process (AHP), the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS), and Constraint Programming to provide a structured approach to maintenance prioritization and planning. The system was implemented as an interactive dashboard and validated through a real-world manufacturing case study. Results show that the AHP-TOPSIS model enables transparent prioritization across risk, cost, and downtime dimensions, while Constraint Programming generates conflict-free schedules that consider technical skills, production cycles, and resource constraints. This integrated approach bridges the gap between expert-driven prioritization and real-time operational planning, offering a scalable and replicable methodology for effective maintenance management.