High resolution bus lane performance evaluation from real time update data

Bus priority measures such as bus lanes are designed to enhance bus performance and increase ridership. Traditionally, benefits have been evaluated at an aggregate level. Newer data sources, however, enable the tracking of micro delays and their relation to detailed bus priority data. Given schedule adjustments for bus priority measures, we anticipate minimal impacts on expected delay at the route-segment level, with the primary benefit being reduced delay variability relative to the schedule. This study analyzes real bus arrival data to examine the impact of stop-to-stop route characteristics on marginal delay. The analysis uses pooled, between-, and within- effects panel regression models to predict average and standard deviation of marginal delay for each stop-to-stop segment within rolling windows of 30 arrivals. Independent variables include priority measures, traffic signals, traffic volumes, scheduled travel time, stop-to-stop link length, scheduled travel speed, cross-traffic turns, precipitation, weekends, holidays, and the COVID stringency index. Findings reveal that bus-taxi lanes and bus-HOV lanes reduce marginal delay by 6–7 s per kilometer. While the direct impact on marginal delay is minimal due to schedule adjustments, these lanes significantly reduce the variability of delay, saving 5–20 s of standard deviation of delay per kilometer. The study also highlights the substantial impact of traffic signals and cross-traffic turns on bus performance reliability. These findings support the effectiveness of bus priority measures in improving bus service reliability. • GTFS-realtime data enable microscopic analysis of bus priority interventions. • Bus lanes reduces both average delay and delay variability. • Cross-traffic turns have a bigger impact on delay and variability than priority lanes.