A case study on maintenance of bus engines using the sequential method

Abstract

This paper reports on the application of the sequential method, presented in a previous paper by Lai et al. in 2001, to determine optimal strategies of when to carry out a preventive maintenance action for an engine and when to replace an engine in use. The sequential method was run with real data provided by the Kowloon Motor Bus Company Limited in Hong Kong. First, both the maximum‐likelihood density estimation procedure and the nearest‐neighbour density estimation procedure were applied to assess the model parameters, and the goodnesses of fit of the distribution was assessed. Second, some values were assigned to the corrective maintenance indicator and the preventive maintenance indicator. Third, the other input values for the method were assessed. Finally, the optimal preventive maintenance and replacement strategies based on minimising loss were determined. This case study shows that the sequential method can be used to solve a maintenance and replacement problem efficiently, and also shows that the method has advantages over the non‐homogeneous Poisson process model in comparison with the results obtained by the latter.