Liong Robotics (Shenzhen) has closed a tens-of-millions yuan strategic round from Lifud, with Yunyue Capital as adviser. The company also signed a research partnership with the AI school of the Southern University of Science and Technology to work on embodied intelligence, spatial perception, world models and high-quality embodied data.
Liong’s thesis is that the real gap in embodied AI is not the model or the body but the middleware between them and the real world. Its LA Capture System is a wearable rig that records a worker’s first-person video, hand motion, spatial position and touch at once, turning normal work into training data without tying up an expensive robot. A companion platform auto-cuts action sequences, aligns multimodal timestamps and scores data quality.
An industrial partner as a live testbed
The sharper move is the investor. Lifud opens two industrial parks, three R&D centres and a sales network across more than 70 countries serving over 10,000 clients, handing Liong not just capital but a stream of real failure cases: slipped parts, misaligned assemblies, the long-tail conditions where robots drift. Those failures are the scarcest training material there is.
Wearable capture is not a free lunch. Quality varies with the worker, and heavy data cleaning remains. But it does not stop the production line or rebuild the floor. In a field starved of real interaction data, a live factory beats a perfect simulation, because sim-to-real decay is exactly what breaks robots on the floor.
Editor’s note: This is an adapted translation of the original OFweek Robot report. It has been trimmed and restructured for readability for an international business audience.
Translated and adapted from OFweek Robot (https://robot.ofweek.com/2026-09/ART-8321202-8420-30701510.html).