JD Industrial Releases Industrial World Model to Boost On-Device AI Decision-Making

On October 13, JD Industrial unveiled its industrial world model at the 26th China International Industry Fair in Shanghai, targeting high-value scenarios such as manufacturing, equipment maintenance, logistics and quality inspection. The model is designed to extend industrial AI from centralized analysis to real-time on-device decision-making, with initial deployments in industrial safety warning and human-machine collaboration.

According to the company, the world model leverages JD Industrial's accumulated expertise in industrial data, domain knowledge and operational workflows, combined with high-performance computing, to support industrial applications capable of scene understanding, dynamic reasoning and risk prediction. This enables complex on-site tasks such as risk assessment, change simulation, task comprehension and decision support to be handled locally, reducing response latency, enhancing data security and improving deployment flexibility.

JD Industrial also released a joint study with the China Academy of Industrial Internet on AI-enabled enterprise value quantification, proposing a business-scenario-oriented value assessment framework. The research links scenario-level business metrics with corporate operational data across quality improvement, efficiency gains, cost reduction and revenue growth, providing a methodology for enterprises to evaluate AI effectiveness and optimize resource allocation.