A Survey: Modelling Strategies for Predictive Maintenance

Abstract

This paper deals with various modelling Strategies to achieve predictive maintenance. Predictive Maintenance (PdM) is introduced into various industries and sectors. It helps to predict the condition of equipment of machines that are already in use this tells whether the maintenance is required or not. This technique ensures that cost-saving has done as compared to regular maintenance where unnecessary replacements have been done without proper utilization of resources. And this can be achieve using various modelling techniques as per the needs or requirements. The paper has four basic principles or strategies or ways to implement PdM as per the type of data available for business requirements. We have also covered the type of data that one should accumulate as per there business or user requirements. We have also tried to cover some of the basic questions that one commonly faces during the implementation of predictive modelling. The big importance of this technology is to predetermine the failure so, that proper timeframe is given for maintenance which simply means its cost-effective, lower the downtime, and better user experience. This approach is widely used in industries like oil, semiconductors, production, etc. (Susto et al. in IEEE Trans Ind Inf, [1]).