Estimating Reduction in Travel Time Based on Large Scale Driving Data from Connected Vehicles

Yuta Fukasawa, Kota Yamada, Yoshinao Ishii, Takeyuki Sasai, Shintaro Fukushima

Abstract: This study estimates the potential reduction in travel time owing to the selection of detour routes with shorter travel times. This is a central problem in intelligent transportation systems. However, the number of vehicles and drivers employed in the previous studies was severely limited. Consequently, to strengthen the evidence of a reduction in travel time, we quantitatively estimate the reduction by leveraging largescale vehicle driving data collected from connected vehicles on metropolitan expressways in Japan. The scale of our dataset is overwhelming in its size, that is, $\mathbf{1 5 3, 4 8 0}$ trips extracted from 611,480 passenger vehicles over one month, compared with one in previous studies, up to 200 passenger vehicles. The experimental results yield the following outcomes. (i) The reduction rate in travel time exceeded 5 % for approximately $\mathbf{1 5. 7 \%}$ of all trips. (ii) A consistent trend of significant travel time reduction was observed across consecutive years. The findings of this study provide an important foundation for the further consideration of transportation measures for datadriven urban traffic management.