Congestion Quantification under Resource Constraints:
A Study to Support it through Taxi-Probe Data of Tokyo

Road traffic congestion is quantified in many ways however, travel time measure is considered to have wider use and users. Under resource constraints, Test Vehicle Method can be used to collect travel time information however, the number of samples required for certain accuracy varies in literature and practice. Beside other factors, the sample size called Test Runs also depends on the variation in travel time. One source of studying this variation is to make use of advanced data collection techniques that are rich in spatial-temporal coverage of traffic behavior. This research utilizes passive probe vehicle as one. In this case, probe vehicles are 20 taxis, running for entirely commercial purposes in Tokyo and instrumented to store real time geo-referenced time and speed record.

The scope of this study is prioritized to principal arterials, stratified by a higher signal density of more than four signals per km. The time period is taken as morning peak period during weekdays for studying recurrent congestion. An analogy is proposed wherein probe data is filtered for obtaining test vehicle samples. The analysis is performed for checking coverage, accuracy, variation and determining sample sizes. It is seen that probe database can be studied to establish useful indicators for conducting test vehicle surveys.