Edinburgh Napier University

  A major challenge faced by road authorities is to manage road networks within budget constraints and the uncertainty associated in predicting future performance of pavements; this is particularly true for the low volume road network. Road authorities need to balance spending expenditure on new capital expenditure with the effective management of existing networks. Pavement management systems (PMS) and maintenance management practices commonly adopted by road authorities are complex, data hungry and are generally developed for major freeways. Such complex PMS do not meet the overall requirements for low volume road network and thus there is a need to develop an efficient model which is able to accommodate all of these challenges. This paper outlines a study that has investigated the effectiveness of pavement maintenance on low volume roads (LVR). A performance prediction model for LVR using Markov chain theory has been developed. This model has the capacity of optimizing using a stochastic technique based on the Markov decision process to assess long term maintenance options. An example using this model demonstrates its applicability for road authorities. Critical assumptions provided in the model consider performance and budget constraints and incorporates the relationship between maintenance actions, costs, and performance. It is suggested that this new approach as a methodology to bridge between performance prediction and optimization theories based on stochastic models may assist road authorities in managing their LVR networks