Urban Travel Time Estimation Using Real Time Bus Tracking Data

In this paper, the concept of using bus probes for urban street travel time estimation in the advanced traveler information system (ATIS) application is investigated through both a thorough literature review and a case study in downtown Chicago using the Chicago Transit Authority (CTA)'s buses. First, existing bus probes studies are reviewed and compared with respect to the research objectives, facility type, data source, modeling technique, and key research findings. The literature supports the concept of bus probes to ATIS applications. However, the past studies have focused on highways (freeways and principal arterials) and used what is called archived automatic vehicle location (AVL) bus data, which are not suitable for real time forecasting (online updating). The case study presented in the second part of the paper is the first of its kind that uses real-time bus tracking data for urban signalized street travel time estimation. Multivariate time series state-space modeling techniques are applied. The case study finds significant interrelation between bus and car speeds. In particular, stronger influence of car operations (in terms of speed) is found on bus operations in the traffic stream than buses on cars. These findings indicate that AVL buses are plausible probes for urban street ATIS. This conclusion, however, must be understood within the study limitations discussed at the end of the paper.