Guangming Zhao

To guarantee the safety and reliability of Autonomous Vehicle (AV) systems, corner cases play a crucial role in exploring the system’s behavior under rare and challenging conditions within simulation environments. However, current approaches often fall short in meeting diverse testing needs and struggle to generalize to novel, high-risk scenarios that closely mirror real-world conditions. To tackle this challenge, we present AutoScenario, a multimodal Large Language Model (LLM)-based framework for realistic corner case generation. It converts safety-critical real-world data from multiple sources into textual representations, enabling the generalization of key risk factors while leveraging the extensive world knowledge and advanced reasoning capabilities of LLMs. Furthermore, it integrates tools from the Simulation of Urban Mobility (SUMO) and Car Learning to Act (CARLA) simulators to automatically interpret and execute the scenario code by LLMs. Our experiments demonstrate that AutoScenario can generate realistic and challenging test scenarios, precisely tailored to specific testing requirements or textual descriptions. Additionally, we validated its ability to produce diverse and novel scenarios derived from multimodal real-world data involving risky situations, harnessing the powerful generalization capabilities of LLMs to effectively simulate a wide range of corner cases. The implementation is available at https://github.com/THU-AI-Testing/AutoScenario.

Traffic regulations provide a solid foundation for the safety of all road users; however, the ambiguous provisions and unclear safety thresholds within these regulations pose significant challenges to compliance, particularly concerning the safe operation of autonomous vehicles. To address this issue, this paper conducts an in-depth analysis of vehicle emergency braking behavior based on the Aerial Dataset for China Congested Highway and Expressway (AD4CHE). The extraction method for the emergency braking risk scenario of natural driving data is proposed, and the correlation between safe distance, safe speed, and driving safety under the scenario of a slightly congested expressway is elaborated in detail. The safety threshold of ambiguous traffic rules obtained can be used for the digitalization of traffic rules that can support the functional development and traffic safety testing of automated driving systems.

In unanticipated obstacle scenarios at intersections, the safety and mobility of autonomous vehicles (AVs) are negatively impacted due to the conflict between traffic law compliance and obstacle avoidance. To solve this problem, an obstacle avoidance motion planning algorithm based on artificial potential field (APF) is proposed. An APF-switching logic is utilized to design the motion planning framework. Collision risk and travel delay are quantified as the switching triggers. The intersection traffic laws are digitalized and classified to construct compliance-oriented potential fields. A potential violation cost index (PVCI) is designed according to theories of autonomous driving ethics. The compliance-oriented potential fields are reconfigured according to the PVCI, forming violation cost potential fields. A cost function is designed based on compliance-oriented and violation cost potential fields, integrated with model predictive control (MPC) for trajectory optimization and tracking. The effectiveness of the proposed algorithm is verified through simulation experiments comparing diverse traffic law constraint strategies. The results indicate that the algorithm can help AVs avoid obstacles safely in unanticipated obstacle scenarios at intersections.

Compliance with traffic laws is a fundamental requirement for human drivers on the road, and autonomous vehicles must adhere to traffic laws as well. However, current autonomous vehicles prioritize safety and collision avoidance primarily in their decision-making and planning, which will lead to misunderstandings and distrust from human drivers and may even result in accidents in mixed traffic flow. Therefore, ensuring the compliance of the autonomous driving decision-making system is essential for ensuring the safety of autonomous driving and promoting the widespread adoption of autonomous driving technology. To this end, the paper proposes a trigger-based layered compliance decision-making framework. This framework utilizes the decision intent at the highest level as a signal to activate an online violation monitor that identifies the type of violation committed by the vehicle. Then, a four-layer architecture for compliance decision-making is employed to generate compliantly trajectories. Using this system, autonomous vehicles can detect and correct potential violations in real-time, thereby enhancing safety and building public confidence in autonomous driving technology. Finally, the proposed method is evaluated on the DJI AD4CHE highway dataset under four typical highway scenarios: speed limit, following distance, overtaking, and lane-changing. The results indicate that the proposed method increases the vehicle's overall compliance rate from 13.85% to 84.46%, while reducing the proportion of active violations to 0%, demonstrating its effectiveness.