Abstract:
School bus routing and scheduling in North Carolina are completed by using the transportation information management system (TIMS), according to state legislative mandate. TIMS uses an optimization algorithm to generate acceptable route alternatives to best serve the transported students and to minimize the required number of buses. This algorithm uses link travel times and speeds encoded in the geocode (digitized map) to determine the routes for each school bus in the state. The speeds and travel times used by TIMS have no mathematical basis and have not been calibrated for conditions in North Carolina. The encoded speeds are simply reductions of the posted speed limit to account for student stops and the bus's general inability to reach the posted speed. There is a need for models that can accurately estimate link travel times and student loading times so that TIMS can provide better estimates of optimal routes. On-board mobile Global Positioning System equipment was used on several school bus routes throughout the state to collect the data necessary for model development and validation. Two student loading time models were developed: a general model that can be used in an urban, a suburban, or a rural area, and an urban-specific model, which involves fewer predictor variables. Link travel-time models were developed according to the two most pertinent factors: link length and traffic control device type at the link's end. Model validation was successful, and now more-efficient school bus routes can be developed because of more-accurate geocode speed and travel-time estimates.
