The mandatory national standard "Safety Requirements for Automated Driving Systems of Intelligent Connected Vehicles" has been officially released.

The "Safety Requirements for Autonomous Driving Systems of Intelligent Connected Vehicles" applies to M-class and N-class vehicles equipped with Level 3 and Level 4 systems, excluding automated parking systems. This standard is based on the actual needs of industry development and regulatory requirements in China, while taking into account technological feasibility, industry adaptability, and practical implementation. It establishes a comprehensive safety requirement system with clear criteria adapted to national conditions. Firstly, it enhances the safety guarantee mechanism for the entire lifecycle of enterprises to solidify the safety foundation. The standard requires vehicle manufacturers to establish safety capabilities covering product design, development, production, and the entire post-deployment process from four dimensions: safety policy, risk management, safety assurance, and safety enhancement, thereby reducing potential safety risks of autonomous driving systems from the source. At the same time, it specifies that vehicle manufacturers should organize corresponding simulation, site, and road tests during the research and validation of autonomous driving systems and ensure that the relevant testing resources and conditions meet the research and validation requirements. Secondly, it strengthens the system's capability to perform dynamic driving tasks to ensure safe usage. The standard stipulates that the safety level of an autonomous driving system should at least match that of a qualified, attentive human driver engaged in dynamic driving tasks, and it must not pose unreasonable safety risks to users or other road users. On this basis, it defines the safety requirements for the system executing dynamic driving tasks in various scenarios and clarifies the triggering and execution requirements for minimum risk strategies. Thirdly, it standardizes human-machine interaction and user notification requirements to prevent misuse and abuse risks. The standard mandates that the activation and deactivation processes of the autonomous driving system should be safe, and it specifies the notification standards for readiness, activation, deactivation, and other status information and signals. For Level 3 autonomous driving systems, a driver takeover capability monitoring function should also be included. Additionally, vehicle manufacturers should communicate key information to users about the driving automation level, capabilities, limitations, control strategies, and role transitions through channels beyond the product user manual, such as official websites and in-vehicle display terminals, guiding users to understand and use the system correctly. Fourthly, it improves multi-dimensional inspection and testing methods to ensure that the standards can be implemented. The standard constructs a multi-faceted inspection system integrating "enterprise capability inspection + safety file inspection + confirmatory testing" to evaluate the integrity and robustness of the enterprise's safety assurance capabilities and product safety files. Based on this, third-party testing organizations shall conduct confirmatory tests using site tests, road tests, and simulation tests based on the recommended national standards GB/T 41798 "Testing Methods and Requirements for Autonomous Driving Functions of Intelligent Connected Vehicles at Test Sites," GB/T 44719 "Testing Methods and Requirements for Autonomous Driving Functions of Intelligent Connected Vehicles on Roads," and GB/T 47025 "Testing Methods and Requirements for Simulation of Autonomous Driving Functions of Intelligent Connected Vehicles," ensuring a comprehensive inspection and testing of the actual capabilities of the autonomous driving system.