An Optimal Scheduling for Medical Equipment Preventive Maintenance Over a Finite Planning Horizon Using Ant Colony Algorithm Neven Saleh, PhD, Samanta Rosati, PhD, Amr Sharawi, PhD, Manal Abdel Wahed, PhD, and Gabriella Balestra, PhD The importance of preventive maintenance management has been gradually recognized specially with the great attention to the role of health technology management. Finding the optimal schedule to perform preventive maintenance for medical equipment is rarely considered in the literature. This research suggests using ant colony optimization method to solve the problem of finding the optimal preventive maintenance schedule. We developed 2 versions of the algorithm, both starting from a prioritized medical equipment list and differing in the heuristic function. The experimental results indicate the effectiveness of the ant colony optimization algorithm for this kind of problems. Preventive maintenance (PM) is a core function of clin- ical engineering, having as objectives the assurance of ongoing safety and performance of medical devices and the preservation of the investment in the equipment through improved longevity. 1 Preventive maintenance consists of a set of activities that aims at improving the overall reliability and avail- ability of a system. 2 The main goal is to keep equipment in a specified condition such as safety and quality. Because PM is expensive, it is important to decide when it should be performed. One of the most used concept is periodical PM, 3,4 which specifies the interventions following to be performed at equal time. Another approach is the so-called sequential PM, characterized by search of optimal number of maintenance actions to be performed during a given period. 3 Although the first concept is more convenient, sequential PM results are more realistic be- cause it usually complies better with budget constraints. Optimizing PM is an old problem, which is very im- portant in different fields, so it has been discussed ex- tensively in the literature. In general, PM sequence can be classified into single-unit and multiunit systems. 5 The literature is rich of different optimization tech- niques for PM scheduling for single-unit and multiunit systems. Examples in the studies by Khan and Haddara, 6 Kim and Ozturkoglu, 7 and Leou 8 present different tech- niques for optimum single-unit PM interval, whereas the studies of Joo, 9 Saleh and Balestra, 10 and Sitayeb et al 11 show different cases for multiunit systems. In case of medical equipment (ME), several models were developed to optimize maintenance interval, con- sidering both costs and reliability. An adaptive PM pro- tocol based on inspections database was designed in Arslan and Ulgen. 12 It uses a risk-based approach to deter- mine the optimal PM interval. In the study of Joseph and Madhukumar, 13 a PM index was developed for every device in the inventory to assign an optimal PM interval based on a risk level coefficient of the device. Risk level coefficient was calculated through 5 different classified factors related to the ME electrical risk. A mathematical model is developed in Khalaf et al’s study 14 using a mixed integerYbased approach for maintenance operation sched- ules for ME. Field data are used to get the parameters of the model by nonlinear least square regression. A greedy algo- rithm is proposed to give an initial solution for the model. To reduce costs, it is important to optimize the inter- vention sequence decreasing the time spent in going through the departments. Even if there are different models that deal with optimal PM policy or strategy for ME, the literature review has shown that no empirical approach has been presented to find the optimum sequential list of ME to perform PM in an efficient sequence. In other words, if we have a set of equipment that should undergo PM, what is the best sequence of devices that minimizes time, labor, and, consequently, costs? Feature Article 142 www.jcejournal.com Volume 42 & Number 3 & July/September 2017 Corresponding author: Neven Saleh, PhD, is an assistant professor at the Faculty of Engineering, Systems and Biomedical Department, Cairo University, Oula, Giza, Egypt, and Electronics and Telecommunication Department, Duca degli Abruzzi, 24, 10129, Politecnico di Torino, Italy. She can be reached at nevensaleh76@gmail.com. Samanta Rosati, PhD, is a researcher at the Electronics and Tele- communication Department, Politecnico di Torino, Italy. Amr Sharawi, PhD, is an associate professor at the Faculty of Engineering, Systems and Biomedical Department, Cairo University, Oula, Giza, Egypt. Manal Abdel Wahed, PhD, is a professor at the Faculty of Engineering, Systems and Biomedical Department, Cairo University, Oula, Giza, Egypt. Gabriella Balestra, PhD, is an associate professor at the Electronics and Telecommunication Department, Politecnico di Torino, Italy. The authors declare no conflicts of interest. Copyright * 2017 Wolters Kluwer Health, Inc. All rights reserved. DOI: 10.1097/JCE.0000000000000227 Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.