When the locomotive system breaks down, it is hard to control the quality of the locomotive overhaul process using traditional network diagram, owing to the complexity of fault phenomena and fault reasons. Moreover, conventional method may lead to damage to some fault-free components easily. Ant colony algorithm (ACA) was adopted to improve the quality in the locomotive overhaul system. In view of the overhaul difficulty, occurrence probability, human resources and time resources in the traditional network diagram, the overhaul plan was optimized based on ant colony algorithm and resource allocation. The experiment results show the ACA-based system is effective for improving the efficiency and quality of the locomotive overhaul process