Applying of K-NN algorithm in the assignment of corrective maintenance work orders for biomedical equipment

Abstract

The maintenance management of biomedical equipment in a health care institution is important to determine different management indicators, including the time of availability of the technology. However, on many occasions, this availability of the technology is not adequate, mainly due to the time it is out of service in the corrective maintenance process. These factors depend exclusively on the personnel or companies that carry out the equipment interventions. The increasing of availability is essential to improve aspects like effectiveness, efficiency, and productivity; the reduction of equipment intervention time will ensure the provision of health services and good quality. Thus, the optimization of this process is an important task in hospital technology management. Whenever it is possible to improve this indicator, in this context, this paper presents the classification using the K-NN algorithm to determine which personnel/companies are more efficient to perform specific corrective maintenance. The training is performed on the failure history of the Fundación Hospital San Pedro (FHSP), Pasto city, Colombia, for reducing the time of equipment intervention and the possibility of performing updates in training for personnel of specific areas of maintenance. The results obtained indicate that the K-NN algorithm is appropriate, to perform an adequate classification to determine the personnel/companies to make the corrective maintenance.