Evaluating the Need for Traffic Signal Retiming Using Connected Vehicle Data

The emergence of automated traffic signal performance measures (ATSPMs), aided by recent technological advancements, allows for continuous traffic performance monitoring, and supports traffic agencies in taking proactive measures. High-resolution trajectory data from connected vehicles (CVs) has surfaced as a cost-effective method for assessing ATSPMs. Although various metrics have been developed to measure traffic signal performance, none have been specifically designed to estimate the benefits of signal retiming. This study devises a novel methodology to estimate the potential reduction in overall intersection delay resulting from signal retiming using only CV data. This methodology results in a new metric, the traffic signal suboptimality index, that uniquely estimates the potential avoidable delay rather than simply the observed signal delay. This new metric could enable traffic agencies to predict the benefits of potential signal retiming without the need for conducting costly traffic surveys and help these agencies prioritize locations and times of day for signal retiming. This study employs the VISSIM microsimulation software to implement and evaluate the methodology under various traffic scenarios and CV market penetration rates. In our experiments, the proposed methodology successfully detected signal retiming needs in situations involving an imbalanced degree of saturation, traffic demand fluctuations on competing movements, and changes in traffic direction, even with CV penetration rates as low as 10%. Furthermore, the sensitivity analysis reveals that the temporal aggregation period can be increased to further compensate for low CV penetration rates.