公道自動運転レベル4を目指した路車連携技術の実証
トンネル壁面への高反射塗料の塗布による自車位置補正の有効性検証
人口減少が進展する中，公共交通のドライバー不足や，過疎地域での路線維持困難が深刻化しており，これらの交通課題に対し自動運転に寄せられる期待は大きい。筆者らは自動運転サービスの社会実装に向けて，東京都公募事業に参画し，公道における実証実験を実施した。その中で，本報では道路等のインフラと連携して走行する路車連携の一つとして，トンネル壁面に高反射塗料を塗布し，自車位置補正に用いる実証について報告する。既往の自車位置補正方法は，走行する周囲の環境条件により精度が低下する等の課題があり，特にトンネル区間は形状が単調であり長距離となるため，位置誤差が大きくなる。実証の結果，検出精度は98.3％であり，最大で100cm誤差となる箇所を±10cm程度まで補正可能であることを確認した。今後は，検出精度の更なる向上や，メンテナンス方法を確立していき，自動運転とそれを支えるインフラの次世代化への一つの道筋として貢献していきたい。
キーワード：自動運転，自己位置推定，LiDAR，路車連携，路車間通信
＊１ 技術センター イノベーション戦略部 技術開発戦略室
＊２ 都市開発本部 新事業推進部
＊３ （株）ティアフォー
＊４ 大成ロテック（株）
Demonstration of Vehicle-Infrastructure Cooperation Technology Aiming at Public Road Autonomous Driving Level 4
Verification of Effectiveness of Vehicle Position Self-correction by Applying Highly Reflective Paint to Tunnel Wall Surfaces
As the population declines, the shortage of drivers for public transportation and the difficulty of maintaining routes in depopulated areas are becoming more serious, and there are high expectations for autonomous driving to address these traffic issues. The authors participated in a public call for proposals project by Tokyo Metropolitan Government, and conducted demonstration experiments on public roads for social implementation of autonomous driving services. This report describes one of the experiments to demonstrate the effect of applying highly reflective paint to tunnel wall surfaces and using it for vehicle position self-correction, as one aspect of road-vehicle cooperation in which vehicles run in cooperation with infrastructure such as roads. The existing vehicle position self-correction method has problems such as a decrease in accuracy due to the environmental conditions around the vehicle. In particular, tunnel sections have a monotonous shape and a long distance, so the positional error becomes large. From the demonstration results it was confirmed that the detection accuracy was 98.3%, and it was possible to correct locations where the maximum error was 100 cm to about ±10 cm. In the future, it is intended to further improve the detection accuracy and establish maintenance methods, contributing to the next generation of autonomous driving and the infrastructure that supports it.
Keywords: autonomous driving, localization, LiDAR, vehicle-Infrastructure cooperation, vehicle-to-infrastructure(V2I)
*1 Research and Development Strategy Section, Innovation Strategy Department, Taisei Advanced Center of Technology
*2 New Business Development Department, Urban Development Division
*3 TIER IV, Inc.
*4 Taisei Rotec Corporation