An Algorithm to Measure Daily Bus Passenger Miles Using Electronic Farebox Data for National Transit Database (NTD) Section 15 Reporting

摘要

New York City Transit (NYCT) implemented an automated algorithm to estimate daily busunlinked trips, infer passenger-miles, and compute average trip lengths by route using transactiondata from an entry-only Automated Fare Collection (AFC) system. Total onboard miles areinferred by taking advantage of symmetries in bus passengers' daily activity patterns. NYCT'salgorithm utilizes rigourously-tested engineering assumptions to detect common data errors frommechanical failures, imperfect driver-farebox interactions, and operational reality, applyingstatistically measured adjustment factors to correct or interpolate for missing passengers fromnon-AFC boardings and malfunctions. Surveys revealed that under typical operating conditions,non-AFC passengers and farebox data transmission errors accounted for 12% and 5½% ofmissing ridership, respectively. The fault-tolerant algorithm uses non-geographic transactiondata from an AFC system without Automated Vehicle Locator (AVL) functionality, directlycomputing aggregate passenger-miles by inferring origin locations from transaction timestampsusing scheduled average speed assumptions, and without assigning each passenger's precisedestination. NYCT focused on fully automatic, production-ready algorithms by rejectingalternatives requiring excessive coding effort, processor time, difficult-to-obtain data, or manualintervention in favour of logical inference, statistical estimation, and symmetry. Meticulousparallel testing demonstrated that resulting average trip lengths are stable across days andcorrelate well with manually collected stop-by-stop ridership data. Annual passenger-miles arewithin –1% to 4% of the National Transit Database (NTD) ±10% sample data and were approvedby Federal Transit Administration (FTA) for NTD Section 15 submission.