Genetic Algorithm-based Clustering Methodology for Maintenance Scheduling in Healthcare Facilities

Maintenance scheduling is considered a fundamental process within facility management frameworks particularly in high-risk buildings like healthcare facilities. Moreover, a proper agglomeration of maintenance actions is crucial for spaces necessitating a smooth continuous operation like Intensive Care Units and Operation Rooms. Consequently, this paper proposes an optimization-based clustering framework that first selects the most suitable actions to be applied for each hospital asset according to the cost, improvement and downtime associated with each maintenance action. This is performed on a multi-objective Genetic Algorithm basis. The output is further fed into a hierarchical clustering algorithm that combines the relevant closely located actions together. This aids in efficiently allocating available resources as well as minimizing disruptions in critical locations within healthcare facilities. The preliminary results of applying the proposed framework on a case study hospital surpassed the current practice by 29% by offering a highly improved performance of hospital assets within the lowest possible operational downtime.