China's first AI platform designed specifically for the automotive industry has been launched in the Beijing Economic-Technological Development Area (BDA), also known as Beijing E-Town, offering automakers shared access to computing power, data services and AI models without having to build the full infrastructure themselves.

The platform was unveiled by the National AI Application Pilot Base for Manufacturing (Automotive), led by China Mobile and established in Beijing E-Town last year. It is a key achievement from the base's first year of development, with its services designed to be shared across the automotive industry.

The platform currently manages 112.7 exaFLOPS of computing capacity and supports four major areas: research and development, manufacturing, intelligent driving, and testing and evaluation. It also features six AI agents for vehicle styling design, manufacturing process planning, visual quality inspection, intelligent operations and maintenance, intelligent driving data generation, and intelligent driving model training.

Shared AI infrastructure addresses common industry barriers

Automakers face several persistent obstacles to wider AI adoption, including limited computing resources, fragmented data and the high technical barriers to developing AI models. Building dedicated infrastructure requires substantial investment and long lead times for major manufacturers, while smaller companies may struggle to afford the cost.

The new platform brings together computing, data and model services as a shared resource, allowing automakers to access capabilities according to their operational needs.

At the computing level, companies can draw on the platform's shared resources rather than investing in their own large-scale computing centers. At the data level, automakers retain full ownership of their core data, which can be exchanged securely with ecosystem partners within applicable compliance frameworks. At the model level, the platform provides tools covering the full workflow from AI development and validation to deployment, lowering the barriers to applying AI in automotive operations.

By integrating these capabilities, the platform aims to make advanced AI resources more accessible to manufacturers of different sizes and accelerate the transition from individual AI projects to broader industrial use.

Six AI agents deliver measurable results in factories

The six AI agents are designed to translate the platform's computing, data and model capabilities into practical improvements in vehicle development and manufacturing. Initial applications have been tested in production and research settings, with reported gains in efficiency and equipment operations.

In Shanghai, a vehicle designer used the styling-design agent to generate designs from sketches. Its multimodal generation capabilities tripled design efficiency for each set of design drawings.

At another automaker in the same city, the process-planning agent was deployed in an assembly workshop. A process-planning document for a single vehicle model that previously took five days to produce could now be completed in four hours.

At an automaker in Chongqing, the intelligent operations and maintenance agent can identify more than 20 types of faults and operational risks, reducing equipment downtime caused by failures by 30%.

These applications illustrate how shared AI infrastructure can support work across the vehicle lifecycle, from early-stage design and manufacturing planning to factory operations and intelligent-driving development.

Building a wider automotive AI ecosystem

A representative of China Mobile said the next phase would focus on expanding access to the platform's core capabilities, matching services more closely with automakers' operational needs and supporting further pilot projects and application validation.

The base also plans to develop a portfolio of replicable demonstration projects and work with companies across the automotive supply chain to strengthen the industry's AI innovation ecosystem.