Self-driving Simulator Test Scenario Framework with Event-triggered Functionality

• Item type: Symposium(1)
• 公開日: 2022-01-28
• タイトル: Self-driving Simulator Test Scenario Framework with Event-triggered Functionality
• タイトル / 言語: en
• タイトル: Self-driving Simulator Test Scenario Framework with Event-triggered Functionality
• 言語: eng
• 資源タイプ / 資源タイプ識別子: http://purl.org/coar/resource_type/c_5794
• 資源タイプ: conference paper
• 著者所属: Saitama University
• 著者所属: Saitama University
• 著者所属: Saitama University
• 著者所属: Saitama University
• 著者所属(英): en
• 著者所属(英): Saitama University
• 著者所属(英): en
• 著者所属(英): Saitama University
• 著者所属(英): en
• 著者所属(英): Saitama University
• 著者所属(英): en
• 著者所属(英): Saitama University
• 著者名: Takuma, Yabe; Yuto, Koyanagi; Keita, Miura; Takuya, Azumi
• 著者名(英): Takuma, Yabe; Yuto, Koyanagi; Keita, Miura; Takuya, Azumi
• 論文抄録 / 内容記述タイプ: Other
• 論文抄録 / 内容記述: This paper proposes a testing framework with an event-triggered functionality to validate flexible test scenarios of self-driving systems in a virtual environment. Autonomous vehicles have been developed and tested worldwide, and virtual environments testing using simulators has become the mainstream driving test method for self-driving systems. However, the virtual environment testing is difficult to set up complex movements of vehicles and pedestrians, and to determine whether an ego vehicle is in dangerous states is difficult. The proposed framework provides the event-triggered functionality to verify the movements of vehicles and pedestrians flexibly. With the event-triggered functionality, the vehicle and pedestrian movement can be controlled by developers. The proposed framework consists of SVL simulator, and the simulator can be linked to Autoware (open-source self-driving software) and configured to move vehicles and pedestrians according to the position and speed of an ego vehicle. This relationship is used to obtain the position and speed values of the ego vehicle. Experimental results demonstrate the degree to which the event-triggered functionality affects simulation runtime.
• 論文抄録(英) / 内容記述タイプ: Other
• 論文抄録(英) / 内容記述: This paper proposes a testing framework with an event-triggered functionality to validate flexible test scenarios of self-driving systems in a virtual environment. Autonomous vehicles have been developed and tested worldwide, and virtual environments testing using simulators has become the mainstream driving test method for self-driving systems. However, the virtual environment testing is difficult to set up complex movements of vehicles and pedestrians, and to determine whether an ego vehicle is in dangerous states is difficult. The proposed framework provides the event-triggered functionality to verify the movements of vehicles and pedestrians flexibly. With the event-triggered functionality, the vehicle and pedestrian movement can be controlled by developers. The proposed framework consists of SVL simulator, and the simulator can be linked to Autoware (open-source self-driving software) and configured to move vehicles and pedestrians according to the position and speed of an ego vehicle. This relationship is used to obtain the position and speed values of the ego vehicle. Experimental results demonstrate the degree to which the event-triggered functionality affects simulation runtime.
• 書誌情報: Proceedings of Asia Pacific Conference on Robot IoT System Development and Platform; 巻 2021, p. 8-15, 発行日 2022-01-28
• 出版者 / 言語: ja
• 出版者: 情報処理学会