Building Origin–Destination Matrices in Bus Networks from Smartphone App Call Detail Records: Evidence from Italy

Origin–destination (OD) matrices of passenger flows are essential for planning, designing, evaluating, operating, and monitoring public transport services. On the one hand, traditional approaches, such as surveys, ride-check counts, and/or smartcard-based methods, could be applied to determine OD matrices. On the other hand, the widespread diffusion of portable devices has opened new ways to estimate OD matrices in public transport. In this paper, a new framework is developed for inferring OD matrices on transit networks from smartphone app call detail records (SACDRs). Real data provided by an Italian bus operator are considered as case study to evaluate the viability of the framework. User-friendly map-based reports are built to represent the OD flows clearly. A validation performed against a benchmark OD matrix demonstrates that the new framework estimates the number of trips both in origin and destination with accuracy. Hence, it can be adopted by public transport companies for an accurate replanning of routes along their networks and service enhancements.