Optimisation of pavement maintenance and rehabilitation activities, timing and work zones for short survey sections and multiple distress types

Existing pavement management systems (PMSs) aggregate short survey sections into long homogeneous sections (typically 1–5 km), aiming to reduce the amount of data and formulate a tractable optimisation problem. However, assuming that sections in a similar condition will undergo a similar deterioration of this condition is incorrect, resulting in imprecise predictions and violations of thresholds. Thus, any maintenance and rehabilitation (M&R) optimisation approach based on long road sections and a single condition indicator (aggregated or not) cannot provide applicable results at the project level. This paper provides a holistic model framework and consistent methodology for a new PMS, abolishing the need for composite condition indices, data aggregation and homogeneous sections in the M&R optimisation process. The proposed bottom-up optimisation is based on short survey sections (25–50m) and solves for optimal treatment type, timing, work-zone length and layout, minimising agency, road user and environmental life cycle costs. Instead of using just one condition indicator, the method models the effects of each treatment on multiple distress types and their combinations. Moreover, the short survey sections are grouped into larger work zones based on purely economic criteria (i.e. economies-of-scale costs). Treatment cost, temporary traffic control cost, and work-zone user cost models are derived as a function of the project size and calibrated for actual prices in Austria. Genetic algorithms are employed as a part of the solution algorithm, being capable of handling an unlimited number of road sections. Furthermore, the approach can be extended to account for budget restrictions, risk and cross-asset allocation of funds for any aging repairable infrastructure system.