A simulation study of platooning AV fleet service in shared urban environments with uncertainties

This paper conducts a simulation study of low-speed Autonomous Vehicles (AVs), referred to as “pods”, platooning in shared urban environments. The proposed on-demand transport service can help solve the “last mile” challenge and improve mobility for non-drivers, elderly, and disabled people. To help the industry understands the dynamic system for deployment, this paper provides a practical prospect for pod platooning without prior planning. We designed a simulation system for the new transport. Non-homogenous Poisson processes were adopted to simulate arrivals of user requests. A three-density-zone map was designed from real-world city layouts. Practical time and location patterns of user demand were used. The supervision cost reduction as a benefit of platooning was estimated. System performance and user experience were evaluated. We identified how deployment can influence platooning. We found that platooning can reduce supervision cost by approximately 20% and 14%, during peak times on weekdays and midday on weekends respectively.