(Yicai) Nov. 5 -- JD.Com has made embodied intelligence a major strategic priority, the chairman of its technical committee said, highlighting the Chinese e-commerce and supply chain technology company’s growing ambitions in robotics and artificial intelligence.

The Beijing-based firm’s roadmap for embodied AI focuses on three areas: collecting embodied data and developing data infrastructure, researching and deploying hardware in real-world environments, and developing cutting-edge embodied AI foundation models, Cao Peng said yesterday in a keynote speech at the Robot and Embodied Intelligence Sub-forum of the World Internet Conference Wuzhen Summit.

JD.Com is building embodied AI data hubs and aims, within two years, to collect more than 10 million hours of high-quality video data captured from real-world human activities and over one million hours of robot operational data, Cao said.

This work will be carried out by JD.Com's frontline staff and ecosystem partners, with all data collection activities conducted in strict compliance with applicable laws and regulations to ensure they are lawful, compliant, secure and trustworthy, he pointed out.

JD.Com’s supply-chain capabilities and extensive range of businesses give the company a distinctive data advantage, he pointed out. These businesses generate massive volumes of real-world data every day, he noted, adding that the diverse, multidimensional, and multimodal datasets provide a rich source of training material for embodied AI.

Cao offered examples of JD.Com's businesses to illustrate this advantage. The company has more than 3,800 warehouses, with a logistics network covering the entire country. It works with more than 200,000 partner retail pharmacies and operates over 10,000 outlets, including JD Malls, 7Fresh supermarkets, and JD Auto service stores. And JD Domestic Services has over 50,000 in-house cleaners, serving nearly 10 million households a year.

Meanwhile, JD.Com is working to integrate AI and robots throughout its entire supply chain, Cao said. JD Logistics plans to buy three million robots, one million autonomous vehicles, and 100,000 drones over the next five years, while the group’s offline retail, healthcare, and industrial park businesses also offer extensive opportunities for using robot, he added.

Robots could be deployed at scale in commercial settings within the next three to five years, but mass adoption by households will take far longer, potentially five to 10 years or even more, Cao told reporters.

Editors: Tang Shihua, Martin Kadiev