An automated vehicle arrival notification system for paratransit customers at Texas A&M University

Abstract

For paratransit passengers, not knowing when their ride (the paratransit vehicle) will be arriving to pick them up is a major concern. The objectives of this research included designing, implementing, and evaluating a system for automatically notifying paratransit customers in advance of the upcoming arrival of their ride. The core of the Vehicle Arrival Notification System (VANS) was an algorithm that estimated the time remaining until the vehicle's arrival at future passenger pick-up points, based on GPS-based Automatic Vehicle Location (AVL) and the vehicle's scheduled itinerary of stops. A twenty-two (22) day demonstration of the notification system was conducted with Texas A&M University's paratransit service. At the time of the demonstration, there were nine (9) regular users of the service; of these, eight (8) chose to participate in the demonstration and were supplied with personal alphanumeric pagers. Equipment-related problems during the demonstration prevented the AVL-equipped vehicle from transmitting data. In the absence of AVL information, VANS simply sent messages to passengers a few minutes before the scheduled arrival time of their ride. The accuracy of the new system's arrival-time prediction with AVL data was evaluated via computer simulations. The reliability of the page-sending process was between forty-four (44) and sixty-four (64) percent during the demonstration; all but five (5) percent of this unreliability was due to known and preventable causes such as the computer hardware being unplugged. The computer simulations predicted that, in future applications, VANS could predict arrival times accurate to within plus or minus five (5) minutes for at least ninety-five (95) percent of trips. Passengers using the system generally favored future notification programs. The demonstration's effects, if any, on the paratransit agency's operational characteristics were uncertain. The objectives of this research were accomplished, as a viable notification system was designed, implemented, and evaluated. Future studies may involve improving the VANS software, investigating funding alternatives for notification, developing transportation networks tailored to fit paratransit operations, and measuring the benefits and other effects of implementing a notification system, from the perspective of both the paratransit provider and the paratransit customer.