Recently, Xu Wenqiang, Director General of the Department of Science and Technology of the Ministry of Transport, responded to hot topics related to the "Artificial Intelligence + Transportation" initiative.
How to advance the "Artificial Intelligence + Transportation" initiative?
1. Strengthen technological research to break through key core technologies. The national major science and technology projects during the "15th Five-Year Plan" period will further focus on building an intelligent comprehensive three-dimensional transportation network. They will increase efforts to deploy and develop key technologies for the integration of artificial intelligence and transportation, organize collaborative research by innovation entities, tackle major technologies, develop key equipment, and innovate development models.
2. Consolidate the technological foundation to improve digital and intelligent capabilities. Accelerate the development of a comprehensive transportation large model and build a unified scheduling platform for multiple types of models, heterogeneous computing power, and general data to provide basic capabilities for the development of intelligent agents across the industry. We will also develop intelligent agents for road, waterway, and comprehensive transportation sectors, covering functions such as road network operation monitoring, safety early warning, and emergency command and dispatch. This will create a "smart brain" and "data hub" with broader coverage, stronger functions, smoother collaboration, and safe and controllable operations.
3. Deepen scenario-based applications to foster new drivers of industry development. Fully implement the "Ten-Hundred-Thousand" innovation initiative for artificial intelligence. Targeting the 860 typical scenarios identified earlier, focus on key areas such as intelligent driving, smart highways, and smart shipping. Organize a number of representative pilot demonstrations and landmark innovation projects to promote the full-scenario application of artificial intelligence technologies.
4. Optimize the industrial ecosystem to create a sound innovation environment. Adhere to the principle of "market leadership and government support", encourage the participation of various innovation entities, and better leverage the main role of enterprises in innovation. Accelerate the improvement of the standard system for "Artificial Intelligence + Transportation" to support the development of new quality productive forces.
How to promote the development of the comprehensive transportation large model?
Xu Wenqiang stated that the comprehensive transportation large model is the technological foundation of "Artificial Intelligence + Transportation", a basic strategic resource, and an "intelligent brain" capable of thinking, analyzing, and making decisions.
"The transportation large model we are building does not refer to a specific product. Instead, it adopts a '1+N+X' architecture, integrating innovative forces from all parties to develop an algorithm library, dataset, and toolchain for the innovative application of artificial intelligence."
Xu Wenqiang noted that the development of the large model will be promoted from three aspects:
l Lay a solid foundation model base. Formulate and issue the "Construction Plan for the Comprehensive Transportation Large Model" to further clarify the construction model and key contents. Promote the development of the foundation model through engineering methods, and accelerate the construction of a common technological base with functions such as computing power scheduling, knowledge hub, training and simulation, and evaluation and certification to lay the groundwork for the development of vertical models and intelligent agents.
l Build an open innovation platform. Launch the first Comprehensive Transportation Large Model Intelligent Agent Innovation Application Competition. Through "promoting research and application through competitions", stimulate innovation vitality across the industry, and discover and cultivate a number of high-quality intelligent agent solutions with practical value.
l Establish a collaborative industrial ecosystem. In August this year, 55 leading industry enterprises, top artificial intelligence companies, universities, and research institutes initiated the establishment of the Transportation Large Model Innovation and Industry Alliance.
Where will the accelerated scenario-based innovation applications be applied?
Xu Wenqiang emphasized that accelerating scenario-based innovation applications is the core starting point and ultimate goal of the "Artificial Intelligence + Transportation" initiative. Guided by scenario applications and driven by technological innovation, efforts will be made at two levels to systematically promote the large-scale innovative application of artificial intelligence technologies in the transportation sector.
In terms of major innovation projects, six key initiatives have been formed: intelligent driving for trucks on major corridors, autonomous formation navigation of inland freight ships, intelligent management and disposal of major highway disasters, free-flow toll collection on highways, high-speed maglev systems, and low-altitude air-ground coordination.
In terms of pilot and demonstration applications, the "Ten-Hundred-Thousand" innovation initiative for artificial intelligence + transportation has been fully launched to systematically promote scenario opening and demonstration applications. Focusing on ten key areas—including intelligent driving, smart highways, intelligent railways, smart shipping, smart civil aviation, smart postal services, intelligent construction and maintenance, smart travel services, smart logistics, and intelligent safety supervision—over 100 pilot demonstrations will be deployed, driving the participation of over 1,000 innovation entities. This will create a vibrant innovation atmosphere where "thousands of ships race forward". Considering technological maturity and scenario representativeness, the pilots will be divided into two categories: "research breakthrough type" and "application demonstration type". Promoted in a localized and classified manner, they will deepen the integration of technological innovation and industrial innovation, form a number of replicable and scalable solutions, continuously spawn and lead the development of new quality productive forces, and provide strong support for accelerating the construction of a transportation power.
