Optimization of Traffic Signal Maintenance Performance: A Case Study of Hsinchu City Using DEA and Tobit Regression

Abstract

This study applies a two-stage meta-frontier DEA and Tobit regression to assess traffic signal maintenance efficiency in Hsinchu City. Results reveal that the East District leads in performance, while spatial constraints affect Xiangshan's efficiency. Key factors such as peak-hour operations, dispatch frequency, and vehicle-task mismatch significantly influence outcomes. Policy recommendations include establishing a decentralized operation center and optimizing dispatch strategies. Findings provide practical insights for improving resource allocation and urban traffic safety.