Optimal Inspection and Replacement Policy using Stochastic Method for Deterioration Prediction

This paper introduces a Weibull hazard model formulated by the aggregated Markov method to forecast the expected life length of highway facilities and equipments and the inspection/replacement processes of tunnel lighting systems. First, the paper focuses on presenting a methodology to overcome the estimation bias issues caused by the incompleteness of the data set due to the lack of full life length information. The incompleteness of the data structure varies in different monitoring schemes. Hence, the necessity to develop an estimation methodology of the deterioration hazard models considering the available incomplete life length data from the observed samples arises. The paper also investigates the optimal inspection and replacement policy of tunnel lighting systems using a deterioration hazard model. The inspection/replacement intervals and the maximum life length of the light bulbs are considered as the major management policy variables, and the impacts of these management variables upon the life cycle costs and the fault probability of the light bulbs are investigated. Subsequently, the optimal inspection/replacement model to find the optimal policy that minimizes the life cycle costs given the levels of the fault risks is presented. The applicability of the methodology presented in this paper is examined against real world data concerning the facilities and equipments on the highways.