Empirical Study of Scooter–Vehicle Mixed Traffic Propagation on Urban Arterials

Abstract

Despite the popularity of scooters as a primary transportation mode in many developing countries, design guidelines and software for arterial signals for accommodating heavy scooter–vehicle mixed flows are not yet available. Because of insufficient research on the fundamental properties of such mixed flow, traffic professionals often must apply existing signal models or simulation programs, developed mainly for nonscooter traffic flows, to intersections plagued by scooters’ complex lane-changing and filtering maneuvers. This study conducts field observations of the complex interactions between scooters and vehicle flows from discharging during the green phase to the formation of stop queues at the downstream intersection. By using statistical analysis results, the study develops a series of formulations to describe the behavior of mixed traffic flows, including estimation of discharging rates, lane choice decision, speed evolution, propagation to join stop queues, and the filtering process of scooters. The study uses field data to evaluate the collective results of the calibrated models for two key signal design parameters: discharging rates for scooters and passenger cars in various mixed flow rates and the time-varying, cumulative queue length at the intersection stop line. Empirical comparison results confirm the promising properties of the developed models, which can serve as the basis for design control plans for arterials with heavy scooter–vehicle mixed flows.