Scheduling delays in synchronous transportation networks

This paper investigates the performance of proposed network management policies responsible for routing and scheduling vehicles in automated transportation networks Performance is quantified in probabilistic terms of expected delays inherent to the scheduling process. Improved performance resulting from extensions of synchronous management to include considerations of multiple feasible routes and scheduled slot slipping is quantified and compared to quasi-synchronous management. This comparison suggests a bound on the minimum number of feasible routes and slot slipping capability for synchronous management to rival quasi-synchronous. The implications of network management on queue formation in stations and, consequently, on station design and layout are presented. It is found that synchronous-type management policies require parallel or lateral berthing in stations, whereas quasi-synchronous management imposes no such restriction on station design.