Reducing preventive maintenance cost error caused by uncertainty Available to Purchase

Maintenance managers must often make preventive maintenance (PM) decisions based on limited historical failure data. These small data sets introduce significant uncertainties which can lead to selecting suboptimal (more costly) PM intervals. Examines the performance of two commonly used estimation techniques (MLE and rank regression), and develops a third hybrid technique which improves the PM interval accuracy for small data sets. In addition, characterizes the risks involved in making PM decisions based on small data sets for all three methods, to provide useful guidelines for the practising maintenance engineer.