Applied Intuition and the Police Science Institute at Korean National Police University (PSI, Korean: 경찰대학 치안정책연구소) are partnering to create simulation scenarios and validate autonomous vehicle (AV) performance in a virtual environment.
PSI is the sole comprehensive security research institute in Korea that integrates social sciences with science and technology. In addition to conducting research on humanities and social fields pertaining to security policy, PSI actively engages in research on security science and technology, such as artificial intelligence (AI) and big data, alongside cutting-edge security research and development (R&D) in response to evolving scientific and technological advancements and changes in the security landscape. PSI generates research outcomes that span a wide array of disciplines.
As part of its Korean government research and development (R&D) project, ‘Development of technology for validating the autonomous driving services in perspective of laws and regulations,’ PSI is conducting a project to create simulation scenarios that help address challenges in the Korean legal system. The project includes:
The project’s goal is to assess the feasibility of validating an AV stack’s performance in a virtual simulation environment. PSI derives the scenarios from the Korean Road Traffic Act and carefully curates representative cases for virtual verification.
PSI decided to use Applied’s simulation product, Simian, because of its versatility in scenario creation and seamless integration with any AV stack. The team also chose Applied’s verification and validation (V&V) platform, Basis, to manage scenario requirements and conduct comprehensive analyses of simulation results for individual scenarios.
PSI uses Simian to generate engaging scenarios and conduct object-level simulations. Additionally, PSI leverages the Simian scenario language to establish regulations based on various provisions in the Korean Road Traffic Act. By employing advanced functionalities such as parameter sweeps, PSI can automatically generate numerous similar scenarios from a single representative scenario.
Basis helps PSI manage previously generated scenarios and ensure they adhere to the defined requirements. PSI also uses Basis to analyze simulation outcomes, evaluate their success or failure based on established rules, and track an AV stack’s overall performance over time.
“All road vehicles must adhere to traffic laws, which are currently designed for human drivers,” said Min-je Cho, Researcher at the Police Science Institute. “Interpreting these laws in the context of autonomous driving is crucial for the integration of autonomous vehicles on the road. Applied Intuition’s products demonstrated exceptional performance in implementing autonomous driving scenarios and evaluating compliance with traffic laws for the Police Science Institute. The project’s findings will serve as valuable data for the establishment of future legal and institutional guidelines for autonomous driving.”