While the industry idolises human-shaped machines, Benmo Technology took the opposite bet in 2020: give a robot wheels and legs, then throw it into the messy places where neither wheels nor feet alone will do.
Direct drive, no gearbox
Founded by Zhang Di, a student of startup mentor Li Ze-Xiang, Benmo rejected the decades-old motor-plus-reducer norm and went all in on direct-drive precision motion. In 2021 it launched Xingtian, the world’s first fully direct-drive dual wheel-leg robot, with six freely moving M1502D joints. It balances at plus or minus 37 degrees of roll, peaks at 17 newton-metres per joint and cuts impact by 60 per cent through impedance control.
The logic is mechanical, not cosmetic. Wheels are fast and efficient on flat ground but fail at steps and gullies. Legs cross brutal terrain but burn their payload on their own weight. Benmo’s view: a wheel-leg base that carries load and adapts is the first real step toward general-purpose robots in actual sites.
Three generations, real deployments
In 2023 Benmo shipped TITA, the world’s first eight-degree-of-freedom wheel-leg robot, moving from labs to commercial inspection, with 18-inch wheels and the ability for two units to combine into a four-wheel-foot form. In 2025 came D1: obstacle clearance up to 70 centimetres, an 80-kilogram payload, switchable dual and four-wheel modes, and a box carrier for last-mile delivery in campuses and offices.
The pitch is “enter the environment first.” Instead of redesigning a site to suit a robot, Benmo builds a chassis that enters the site as it is: rubble, ramps, mud and standing water. That is the unglamorous backbone of scaled deployment, and it is where the company believes the real barrier was always sitting, in hardware, not algorithms.
Editor’s note: This is an adapted translation of the original OFweek Robotics report. It has been trimmed and restructured for readability for an international business audience. The full original (in Chinese) is at https://robot.ofweek.com/2026-09/ART-8321204-8120-30701232.html.
Translated and adapted from OFweek Robotics (https://robot.ofweek.com/2026-09/ART-8321204-8120-30701232.html).