Investigation of Self-localization using Side Camera

Yoshiki Sakadani･Toshihiko Hashiba･Masaya Endo (Mitsubishi Electric)

When locating a vehicle's position on a High-definition map, if the accuracy of GNSS decreases, there is a method to locate the vehicle's position by obtaining the lane shape with onboard sensors. However, there are scenes where it is difficult to locate the vehicle's position because the lane shape cannot be obtained with sensors depending on the road environment. In such scenes, we considered a method to estimate the lane shape and enable the location of the vehicle's position.

Technological Development and Establishment of Simulation Environment for Lane Merging Support

Nobuhiro Araki･Kenta Shintoku (Kozo Keikaku Engineering)･Koichi Miyashita･Satomi Aiko (Mitsubishi Research Institute)

This study evaluated the improvement and feasibility of merging processes using the traffic flow simulation environment with two merging assistant systems through the infrastructure: Day 2 system which continuously acquire information on the main lane and provide to the vehicle on the merging lane, and Day 3 system which coordinate gaps between vehicles on the main lane. Result showed that the Day 2 system effectively improved merging and met feasibility criteria. While the Day 3 system is also effective, but the additional support in the Day 3 system did not show significant improvement with the parameters adopted in this study.

Analysis of the Effect of Road Alignment on Visibility Delay for Curved Roads

Hiroshi Kuniyuki･Seiya Tanaka･Fumitaka Fukuzawa (Suwa University of Science)･Masashi Makita (Teikyo University)

This study analyzed the effects of road alignment on visibility delay for curves by experiments using a driving simulator. The results showed that the visibility behavior for curves was first recognition of the existence of the curve, followed by recognition of the curvature of the curve. Furthermore, it was found that the curve visibility delay in those phases was influenced by the left-right difference and the curvature difference.

Proposal of Virtual Platform by Utilizing Multi-agent Traffic Simulation for Sophisticated Automated Driving System Development

Sou Kitajima･Nobuyuki Uchida (JARI)･Naoki Suganuma (Kanazawa University)･Tadashi Okuno (OS Planning)･Jun Tajima (Misaki Design)

This paper describes an effective virtual platform to develop more sophisticated Automated Driving System by generating both complicated and perplexing scenarios. Because it is difficult to find and accumulate these scenarios through field operational test, an enhanced function is developed and implemented our multi agent traffic simulation. The results indicates that this function enable not only to reinforce significant variation of extracted rare scenarios but also to realize decision making improvement of Automated Driving System.