Li Keqiang Proposes Vehicle-Road-Cloud Integration to Break Through Intelligent Driving Safety Bottlenecks

[Technical Approach] At the 2026 High-Level Forum on Intelligent and Electric Vehicle Development, Academician Li Keqiang proposed 'vehicle-road-cloud integration' as a solution to overcome intelligent driving safety bottlenecks.

Core Trend: Vehicle-Road-Cloud Integration Establishes a New Safety Paradigm

Li Keqiang pointed out that while L2-level new vehicle penetration has reached 64.9%, single-vehicle intelligence faces challenges such as perceptual blind spots, data silos, and high-cost architectures, leading to frequent safety incidents. He emphasized the necessity of achieving holistic coordination among vehicles, roads, and cloud systems to establish beyond-line-of-sight perception and global decision-making capabilities, thereby building an intrinsic safety framework.

Key Metrics: Safety Performance Improved by Nearly 50%

The architecture can enhance the safety performance of connected forward collision avoidance assistance by 48.6%. Currently, cities including Beijing and Chongqing have established shared data infrastructures, with 15 automakers jointly running demonstrations across 17 scenarios, and 7 automakers already advancing toward mass production.

Strategic Foundation: Overcoming Siloed Development

Through layered decoupling and low-code platforms, vehicle-road-cloud integration will reconstruct R&D workflows, improve AI training efficiency, and foster cross-industry ecosystem collaboration, providing a sustainable commercialization pathway for intelligent connected vehicles.