Using k-NN Artificial Intelligence for Predictive Maintenance in Facility Management

Abstract

This article presents a study on the application of the k-Nearest Neighbor (k-NN) machine learning algorithm for predictive maintenance in facility management. The implementation of predictive maintenance is crucial for the elimination of unforeseen machine breakdowns, optimization of operational efficiency, and reduction of costs. The k-NN algorithm was employed on a dataset comprising diverse operational factors to predict the probability of a machine's malfunction. The findings of our case study demonstrate that the k-NN algorithm possesses favorable qualities for application in predictive maintenance scenarios, owing to its straightforward implementation and versatility in generating accurate outcomes. Nevertheless, supplementary measures beyond the selection and implementation of models are necessary to actualize the potential of predictive maintenance fully. The procedures encompass the creation of a dependable data framework, the continual surveillance and refinement of models, and the assessment of more intricate modelling methodologies. The study's results indicate that the k-NN algorithm exhibits promise as a valuable tool for predictive maintenance, thereby offering significant benefits to facility management strategies in terms of efficiency and effectiveness.