DONGGUAN, China, Oct. 23, 2026 /PRNewswire/ -- DataBeyond showcased PICKINGMAN, its embodied AI robot for mixed municipal solid waste sorting, during the robot's first global public livestream at 8 a.m. Beijing time on October 22. The broadcast took place at the Liaobu Municipal Solid Waste Integrated Treatment Center in Dongguan.

Built for Waste-Sorting Conditions

The broadcast demonstrated the robot working on an operating sorting line, highlighting the application of embodied AI to demanding industrial tasks involving mixed materials, dirty conditions and repetitive manual labor. PICKINGMAN began operating at the facility on September 22.

PICKINGMAN features dual arms with a total of 14 degrees of freedom, IP65-rated dust and water protection, and a maximum picking rate of 3,600 picks per hour. Designed to support continuous 24/7 operation, it can sort recyclables including PET bottles and trays, PP containers, PE items, aluminum cans and glass. It coordinates perception, decision-making and motion control to identify and pick target materials on a moving conveyor.

Project Results

Jointly developed by DataBeyond and Guangdong Dongshi Environment Co., Ltd., the Liaobu project combines robotic picking with AI optical sorting and other processing equipment. According to project data, the complete sorting line can process 50 metric tons of mixed municipal solid waste in four hours, achieving approximately 35% waste reduction and 25% resource recovery. Previously, sorting recyclables from the same volume required four skilled workers working for eight hours.

The integrated line separates recyclables for downstream recovery before residual waste is compressed and sent for further treatment, connecting resource recovery with existing waste management operations.

Workforce Benefits and Next Steps

For operators facing recruitment and retention challenges, robotic picking offers an alternative to repetitive manual sorting. The project aims to reduce workers' direct contact with waste and enable personnel to focus on equipment inspection, production coordination and other technical roles.

"During the livestream, we identified areas where the robot still needs improvement," said DataBeyond's Mo Zhuoya. "Bridging the gap between R&D and industrial deployment requires engineering experience. We will accelerate product iteration and conduct longer global livestreams."

About DataBeyond

DataBeyond develops AI-powered optical sorting equipment and robotic solutions for recycling and solid waste processing. The company integrates AI and multisensor technologies to improve material separation and support resource recovery.

Media Contact

Felix WU
marketing@databeyond.com