Analysis and Modeling of Cyclist's Decision for Left-Turn Vehicle at Uncontrolled Intersection

Ryo Wakisaka･Kazunori Ban (Toyota Technical Development)･Takuma Yamaguchi･Hiroyuki Okuda･Tatsuya Suzuki (Nagoya University)

In order to evaluate automated driving systems through simulation, it is necessary to model traffic participants. In particular, cyclists are an important factor in terms of safety because of their erratic movements due to ambiguity in rule recognition and other factors. In this study, we analyzed and modeled cyclist behavior for a left-turn vehicle at an unsignalized intersection based on measured data from a cycling simulator.

Analysis of Pedestrian and Vehicle Behavior using Machine Learning for Risk Assessment of Crosswalks

Miki Hayashima･Yuji Matsuki (Fukuoka Institute of Technology)

To clarify the causes of accidents at crosswalks, an analysis of the movements of pedestrians and vehicles is essential. In this study, we developed a system that employs a machine learning model to perform this analysis. This model utilizes images captured by multiple general-purpose cameras as input, allowing for an accurate analysis of the dynamics at crosswalks. Furthermore, we applied this system to quantitatively assess the risk associated with pedestrian crossings.

Effect of Duration and Frequency of Distraction on the Number of Rear-End Accidents

Hodaka Kita･Yuki Arai･Juan C. Gonzalez Palencia (Gunma University)･Noriaki Takenoue (GSEC)･Kenji Amagai･Mikiya Araki (Gunma University)

Understanding the effect of the duration and frequency of distraction on the number of rear-end accidents can contribute to the development of ADAS. In this research, distracted driving was simulated by stopping visual information input at arbitrary durations and frequencies for traveling vehicles using a multi-agent model. Result show that the number of accidents begins to increase when the duration of the distraction is higher than approximately 1 s.

Analysis on Operation Behavior for Emergency Stop Switch while Driver's Sudden Illness

Hiroshi Kuniyuki･Toshiaki Tanaka･Shuhei Tazawa (Suwa University of Science)･Daisuke Ito (Kansai University)

The Emergency Driving Stop System (EDSS), which can stop a vehicle when a driver's sudden illness occurs while driving a vehicle by pressing the switch, is beginning to be adopted for large commercial bus etc. This study analyzed the operation behavior of pressing the emergency stop switch using a subject experiment, and examined the operation method and switch placement that would facilitate pressing the switch while driver's sudden illness.