At the JD.com Global Tech Explorer Conference (JDD Conference), Cao Peng, Chairman of the JD Technology Committee and President of JD Cloud, stated that physical AI is moving towards large-scale implementation, with the focus of competition shifting to computing power, data, and scenarios, transitioning from single-point competition in models and entities to systemic competition.

He emphasized that JD's approach to AI is not just "in papers," but in production lines, warehouses, delivery stations, and throughout the entire supply chain process, developed through one order at a time. Focusing on the construction of a super supply chain, JD has officially launched the "JD Physical AI Acceleration Program," continuously investing in six key areas: data, machines, manufacturing, sales, logistics, and service for embodied robots.

Six Goals: From Ten Million Hours of Data to 300,000 Robots

In terms of data, JD aims to build the "largest embodied data training center in the world," planning to collect more than ten million hours of real-world human scenario data within two years, and integrate network data, simulation data, and real robot data to build an integrated embodied intelligence technology model that supports "understanding—generation—action."

Regarding base construction, the goal is to establish the "largest robot manufacturing base in the world," planning to set up 80 bases nationwide over the next five years, creating a full-life cycle service platform covering display, delivery, maintenance, servicing, R&D, design, assembly, data collection, production, and iteration. In terms of manufacturing, the goal is to become the "largest supplier of core robot components in the world," investing in core components and building its own factories over the next three years, helping 100 original equipment manufacturers reduce costs across all product categories and driving 100 component suppliers to double their performance.

In terms of sales, by 2028, JD plans to invest over 10 billion yuan to help 100 brands achieve independent sales exceeding 1 billion yuan, driving robots into millions of terminal scenarios. In the logistics field, JD has already implemented joint development clusters covering the entire supply chain, including warehousing, sorting, transportation, and delivery. Over the next five years, JD plans to purchase more than 300,000 robots, 100,000 unmanned vehicles, and 100,000 drones. In terms of service, over the next five years, JD will establish a after-sales and service network covering more than 100 countries, creating over 100,000 positions for robot after-sales service engineers.

Cao Peng said that as these goals gradually take shape, JD will drive AI from thousands of industries to households around the world, building the largest physical world collaboration system globally, ultimately becoming a leading global physical world operation center.