• Session No.136 Safety of Autonomous Driving
  • October 17Kitakyushu International Conference Center 1113:10-15:50
  • Chair: TBD
For presentations that will not be available video streaming after congress, a “✕” is displayed in the “Video” column, so please check.
No. Video Title・Author (Affiliation)
1

Impact of Roadside Sensor’s Sensing Performance on Automated Driving at Intersections Using Cooperative Systems

Hiroshi Yoshitake・Jiang Wu (Institute of Science Tokyo)・Wataru Kugimiya (The University of Tokyo)・Motoki Shino (Institute of Science Tokyo)

To realize safe and efficient automated driving in mixed traffic environments, our previous study demonstrated the effectiveness of cooperative systems utilizing roadside sensors. In this study, we hypothesize that the sensing performance of roadside sensors—such as detection rate and accuracy—can significantly affect the safety and efficiency of automated driving. To examine this impact, we conducted numerical simulations focusing on an automated bus traveling through an intersection.

2

Extraction of Contributing Factors for Safe Speed Calculation of Automated Buses Traveling Straight Through Intersections

Taichi Sawanobori・Hiroshi Yoshitake (Institute of Science Tokyo)・Yui Matsuura・Masaya Segawa (Advanced Smart Mobility)・Motoki Shino (Institute of Science Tokyo)

Currently, the speed of automated buses is subjectively determined based on the driver’s experience. However, for social implementation, a logical basis for speed setting is required. This study focuses on the difference between the safe speed that allows for collision avoidance with anticipated road users and the actual set speed. We identified the factors that should be considered when driving straight through intersections and examined a method for calculating safe speed that accounts for these factors.

3

Development of ultrasonic transmission simulation to predict the effect of sonar false detection

Motoyasu Ukai・Takaaki Nakamura (Aisin)

High-precision sensing technology is essential to realize autonomous driving and autonomous parking. AEB (Automatic Emergency Braking System) detects obstacles using ultrasonic sonar, but false detections occur due to the influence of surrounding vehicle parts. In this study, we report on our efforts to develop CAE technology that can predict sensor signals when installed in a vehicle and judge whether they are good or bad.

4

Effects of Cooperative Assistance Systems on Pedestrians by Using Mixed Reality

Yuki Sakamura・Ryohei Homma・Takashi Wakasugi・Genya Abe (JARI)・Motoki Shino・Hiroshi Yoshitake (Institute of Science Tokyo)・Yuji Takagi (Honda R&D)・Takashi Misumi・Tadafumi Shima (MLIT)

We investigated the effects of cooperative assistance using networks on pedestrians crossing roads in avoiding accidents, and the minimum timing of assistance notification that should be satisfied. An experiment using mixed reality technology were conducted on two actual cases in which pedestrian accidents have been occurring frequently in Japan. The effectiveness rate of the assistance was confirmed from the experiment data, and appropriate assistance timing was examined.

5

Camera Exposure Time for Reproducing Human Perceived Visibility in Snowstorms

Toshimitsu Sakurai・Hirotaka Takechi・Ikku Koshikuni・Masaru Matsuzawa (Civil Engineering Research Institute for Cold Region, PWRI)

Road administrators refer to CCTV camera footage to assess snowstorm conditions on-site and determine whether to dispatch road patrols or snow removal operations. However, since human vision and CCTV camera capabilities differ, the snowstorm conditions captured by cameras do not always correspond to the visibility experienced by individuals. This study aims to reproduce human-perceived snowstorm conditions using cameras. The results indicate that setting the camera’s exposure time to approximately 15 milliseconds enables recognition of visibility comparable to human perception. Furthermore, reducing the exposure time enhances long-distance visibility even in severe weather conditions, such as snowstorms.

6

2D quantitative measurement of friction coefficient μ with icy road surface using a NIR camera

Tomoki Kawahara・Akihiro Kido (Tohoku Gakuin University)

There are many technical challenges in measuring the slipperiness of roads in snowy and cold regions, and it is currently limited to qualitative measurement. In this study, we focused on the fact that ice has an absorption band in the near-infrared range, and attempted non-contact quantitative measurement using a near-infrared camera. As a result, two-dimensional quantitative measurement of the friction coefficient μ became possible by the processing the captured images.

Back to Top