ABSTRACT
Budget constrains forced airport administrations to take preventive control of runway pavements damage. As a result, decision makers now face major challenges: the schedule of interventions for the entire planning period, which segments to repair first, and the timing for each intervention. Airport Pavement Management Systems are tools considerably useful tool for this purpose. However, the long-term maintenance and repair scheduling is a combinatorial problem, therefore, the optimal solution is hard to find. In order to solve this, several heuristics have been studied and, in particular, genetic algorithms have proven to be a technique with great advantage in this topic. This paper explores the use of genetic algorithm technique and its inner methods to support pavement maintenance planning for a twenty year period. In order to test the performance, a smaller instance was tested and the heuristic near-optimal solution was proven to very close to the exact result.
