Shanghai Advances Autonomous Driving Test Platform with Simulation Tech as Key Enabler

On July 19, 2026, during the World Artificial Intelligence Conference (WAIC 2026), the "High-Level Autonomous Driving Innovation and Development Forum" was held at the Shanghai Expo Center. Co-hosted by the Shanghai Municipal Commission of Economy and Informatization, the Shanghai Municipal Transportation Commission, and the Shanghai Municipal Public Security Bureau, and organized by Intelligent Vehicle Innovation & Development Platform (Shanghai) Co., Ltd., the forum centered on the theme “Data-Intelligence Driven · Accelerated Smart Mobility.” It unveiled eight key outcomes, including the *Shanghai Implementation Measures for Testing and Demonstration of Intelligent Connected Vehicles*, an autonomous driving public service platform, a compliance testing platform for road traffic regulations for intelligent connected vehicles, and the *Guidelines for Functional Simulation Evaluation of Autonomous Driving Systems*. These achievements span policy and regulation, public platforms, testing and evaluation, data sharing, and industry standards. The forum highlighted that as L2/L3 autonomous driving access standards are rolled out and L4 commercialization advances, the competitive focus of the intelligent driving industry is shifting from algorithms to capabilities in data, testing, and infrastructure. Public testing platforms are recognized as critical infrastructure linking data, R&D, testing, regulatory approval, and oversight, with simulation capability serving as the core technological foundation of this ecosystem. During a roundtable discussion, Bao Shiqiang, CEO of 51Sim, stated that the industry has entered a phase where “data drives models, and models drive capabilities,” yet the scarcity of high-value data constrains model evolution. He emphasized that the core value of simulation lies not in replacing real-world data but in amplifying high-value data. 51Sim’s newly launched SimOne 4.0 platform leverages neural reconstruction technology to transform real-world roads into high-fidelity, interactive, and editable digital scenarios. Combined with generative AI, it expands coverage of long-tail and high-risk scenarios, establishing a closed-loop workflow: “real data → reconstruction → generation → validation.” In anticipation of the upcoming 2027 implementation of the *Safety Requirements for Combined Driving Assistance Systems of Intelligent Connected Vehicles* and the *Safety Requirements for Autonomous Driving Systems of Intelligent Connected Vehicles*, the SimOne platform has been upgraded to meet regulatory compliance needs. It now includes 136 new typical L2/L3 test scenarios covering complex conditions such as tunnels, highway toll plazas, and extreme weather. Additionally, the platform has established a simulation credibility validation system, enabling quantitative calibration of sensor models and vehicle dynamics. The platform also features an upgraded 3DGS hybrid rendering engine, significantly improving multi-sensor rendering efficiency while reducing GPU resource consumption. Bao Shiqiang noted that 51Sim will continue to support the development of Shanghai’s autonomous driving public testing platform by providing technical expertise in scenario construction, simulation testing, and capability evaluation, thereby advancing standardization and reusability across data, scenarios, testing processes, and evaluation frameworks. Looking ahead, simulation platforms will evolve beyond mere R&D tools to become foundational infrastructure connecting public data platforms, testing facilities, and regulatory systems—facilitating trusted circulation among real-world, synthetic, and simulated data to underpin large-scale production of intelligent driving systems. Currently, Shanghai is accelerating the establishment of a comprehensive capability system encompassing policy management, public data, testing and validation, standardization, and regulatory platforms to foster an innovative ecosystem for high-level autonomous driving.