OPTIMAL MAINTENANCE DECISIONS UNDER SOME UNCERTAIN AUTOMOBILE EQUIPMENT DETERIORATION

Two critical components (brainbox and axle) of a commercial bus fleet have been selected for study towards obtaining optimal maintenance decisions, under some uncertain automobile equipment deterioration. An interview-assisted questionnaire was used to obtain data for the project. Three prominent probability distributions were studied, namely the exponential, weibull and lognormal functions.The weibull and lognormal functions were optimal for the brainbox and axle components, respectively. Models obtained were y=0.06t1'43e't2'43/41 for brainbox and y=0.08fle(lm-|7-36)2/sl for axle, interchangeable, at same / value, with below 20% error margin. Equipment reliability of up to 70% was achievable. Bathtubs of the components exhibit similarity and this was supported by the Kolmogorov-Smirnov test (0.194), failure ratios (0.20 and 0.24) and correlation coefficients (0.823&0.816) for brainbox and axle respectively. The general linear models are 3.93810+7.85E-0.3t, for brainbox and 13.3079+0.1980691, for axle scatter plots. It is postulated that optimal (preventive) maintenance (drying, trouble-shooting and cleaning) of a brainbox every 7.5x104km and relightening ,greasing and change of lubricating oil in the axle, every 2.85xl05km, would reduce reactive, corrective (replacement) maintenance to less than 30%.