Data-Driven Maintenance Prioritisation and Scheduling for Industrial Equipment

Abstract

Effective maintenance planning is essential for sustaining productivity, improving equipment reliability, and maintaining cost-efficiency in modern manufacturing environments. As production systems grow in complexity, the reliance on data has become more crucial for informed, timely, and scalable maintenance decisions. Traditional rule-based approaches often fail to account for the dynamic nature of operational data such as technician availability, machine utilisation, failure history, and cost trends, which limits their effectiveness in real-world industrial settings.. This thesis responds to these challenges by developing a tailored decision support system, integrating a hybrid multi-criteria decision-making model with constraint programming. The proposed Decision Support System combines the Analytic Hierarchy Process and the Technique for Order Preference by Similarity to Ideal Solution to prioritise maintenance tasks based on costs, estimated downtime and risk priorities. These maintenance tasks are ranked and subsequently fed into a Constraint Programming model that generates an optimised maintenance schedule that accounts for technician availability, shift structures, and other production constraints. The complete system is implemented within an interactive dashboard, replacing traditional manual planning methods with a scalable, data-driven solution. This research demonstrates how hybrid decision-making techniques, when coupled with constraint-aware optimisation, can bridge the gap between expert-driven maintenance strategies and real-time operational planning. The resulting approach provides a replicable and adaptable methodology for proactive, optimised maintenance scheduling in industrial settings. Unlike the existing literature addressing Multi Criteria Decision Making and optimisation techniques individually, this thesis addresses combines a hybrid framework, through AHP and TOPSIS, with Constraint Programming into a unified and deployable framework designed to handle real world constraints in a dynamic manufacturing environment.