One of the biggest challenges facing industrial embodied intelligence is where data comes from, said Yu Chunhua, head of the shared factory at XbotPark Robotics Base, highlighting persistent difficulties in collecting, sharing, and reusing industrial data.

Speaking at an industry forum on embodied intelligence during the 2025 T-EDGE, Yu noted that while China has a rich industrial ecosystem and diverse manufacturing scenarios, turning these environments into usable data remains a major hurdle.

"A critical issue for industrial embodied intelligence is data availability," Yu said. "China has abundant industrial chains and application scenarios, but how to collect this data, and more importantly, how to share and reuse it, is extremely difficult."

He added that many industrial environments treat data as confidential, limiting cross-factory and cross-industry data flows and slowing the training and deployment of embodied intelligence systems.