Transit services connect people to jobs and opportunities, fostering vibrant communities and multimodal travel along service corridors. A transit right-of-way (ROW) can help buses bypass congestion and stay on schedule. Many studies have proved that transit ROWs effectively improve service reliability and reduce user costs. However, these studies often focus on one or two service corridors, limiting comprehensive impact assessment. This project addresses this gap by investigating service reliability for all route segments across a transit system. We derived reliability metrics at the route segment level using high-resolution automatic vehicle location (AVL) and automatic passenger count (APC) data collected in the Twin Cities metropolitan area.
We then collected and integrated data from various sources via spatial-temporal computing to capture service characteristics, operating environments, traffic conditions, and land-use features along route segments. We applied the Gradient Boosting Model (GBM) to examine nonlinear relationships between these factors and bus travel time reliability. Lastly, we used the trained model to estimate potential improvements in reliability with dedicated ROWs. Through these steps, we worked with members of the Technical Advisory Panel (TAP) to illustrate our methodology and demonstrate its utility for transit agencies. Specifically, the results proved that the ratio of bus lanes and busways was associated with more reliable travel time along route segments. We also found that route segments along a few service corridors with unreliable services can greatly benefit from implementing a dedicated ROW.
