CCTV Finance Report: TranXmart Builds on Multidimensional Safety to Support High-Quality Decision-Making for Autonomous Heavy Trucks

From April 22 to 24, the 16th China International Road Traffic Safety Products Expo was held in Nanjing. Approved by the Ministry of Public Security and the Ministry of Commerce and hosted by the China Road Traffic Safety Association, this year's expo focused on empowering traffic governance with technology and promoting the deep integration of new technologies with practical applications. During the event, the Blue Book on Traffic Safety Development of Intelligent Connected Vehicles (2026), edited by the Road Traffic Safety Research Center of the Ministry of Public Security, was officially released, further improving the safety supervision system for intelligent connected vehicles and underscoring the importance of building a multidimensional safety system.

During the expo, TranXmart was interviewed by CCTV Finance as a representative company in autonomous driving. Building on single-vehicle safety redundancy, TranXmart has created a multidimensional safety defense line through the deep integration of roadside models and cloud transportation large models, providing perception support for high-quality decision-making by autonomous heavy trucks and calmly handling rain, fog, bridges, tunnels, and other scenarios.

As a leading L4 autonomous freight capacity service provider for highway freight, TranXmart proposed an innovative full-chain safety system of “vehicle-side decision-making + roadside blind-spot compensation + cloud optimization.” Through centimeter-level precise positioning near the vehicle, long-range blind-spot compensation, millisecond-level high-quality decision-making, and cloud-based global optimization, it builds a multidimensional safety system for autonomous driving.

In vehicle electronic system architecture, TranXmart adheres to automotive-grade dual-system safety redundancy design, ensuring that when the primary system fails, the backup system can immediately take over the vehicle and keep it operating safely in unexpected situations.

With multidimensional heterogeneous safety redundancy as its foundation, TranXmart uses a dual-system architecture of generative models and optimization-algorithm safety baselines to continuously improve vehicle decision-making in complex scenarios, refresh the decision quality of the autonomous driving system, and gradually enable it to evolve from learning from people to approaching people and even surpassing human capabilities.

In the future, TranXmart will continue strengthening the safety foundation of L4 autonomous driving technology, steadily improve single-vehicle intelligence, and build new transportation productivity so that every kilometer of highway freight becomes safer, more efficient, and smarter.