AUTOMATIC PASSENGER COUNTING DATA - BETTER SCHEDULES IMPROVE ON-TIME PERFORMANCE

OC Transpo has operated an Automatic Passenger Counting (APC) system for over 10 years. A major benefit of the APC system is the provision of data for the determination of correct running times for future schedules. The study of APC-produced distributions of running times along route sections have revealed the expected overall variation by different times of day but, more importantly, they have also shown wide variation within fairly narrow time-frames. The problem for schedulers is to decide the number of minutes to allocate so that the maximum benefit is achieved. Should the median time be used or should the allocated time ensure that a percentage of all trips have sufficient running times? In order to properly establish correct running times, large volumes of observed data must be collected. The advancement of automated passenger counting and data collection systems has provided a tool to satisfy the transit operator's need for these data. The following questions will be discussed: What are the implications of poorly allocated running time? How should the maximum benefit resulting from the scheduling process be defined. What data is needed to measure maximum benefit? How best can an APC system provide this information? How should APC data be used to set scheduled running times? The paper addresses these issues and their affect on the corporate objectives of on-time performance. It will be seen that the component with the most influence over the quality of service is scheduled running time.