Dynamic ride-sharing impacts of greater trip demand and aggregation at stops in shared autonomous vehicle systems

Sharing vehicles and rides may become the norm with public use of fully-automated self-driving in near future, assuming pandemic-related health concerns fade. Dynamic ride-sharing (DRS) or ride-pooling trips can significantly improve system performance by lowering empty vehicle-miles traveled (eVMT) increasing average vehicle occupancy (AVO). With several cities looking to promote efficient curb space use, especially pickup drop-off locations (PUDOs), this study explores impacts PUDOs on DRS rates AVO values. Various PUDO spacings trip-demand densities were studied, across Bloomington, Illinois region, using agent-based simulator POLARIS. Results reveal that greater spacing distances between stops higher levels SAV trip demand increase (by up 20% per 4-seater SAV, average) decrease VMT 27%) compared door-to-door fleet operations without PUDOs. A quarter-mile is recommended downtown regions, similar current transit stop spacing, keep walking short relatively high. At 0.25 mi (thoughtfully placed, origin destination clusters), travelers walked less than 5 min at either end, average, while 0.5 led another 1 (approximately) walking. More evenly distributed save 39% total from stops. It also important prepare for queuing areas settings high density, limit curbside congestion.