LYNX M20 takes a lighthouse trial shift

On Qiliyu, near Ningbo, China, a LYNX M20 robot carried 15 kilograms up 145 steps. Guangzhou Daily’s November 12 report describes DEEP Robotics’ lighthouse trial. The load equals three five-kilogram bags of rice. A path that is manageable empty-handed becomes a different job with that much on your back. Compared with a stage routine, carrying something useful changes where the benefit goes: to the person who would otherwise shoulder it.

The station has a history longer than any robot product cycle. Historical reporting by Ningbo News records an original lighthouse built in 1865, later relocated for preservation, and a replacement completed in 1997. For a vessel offshore, the light provides a reference point. For the people maintaining it, the station is also equipment that needs attention. Scenic photographs tend to emphasize the first view. Keeping supplies and maintenance moving belongs to the second.

The robot also returned equipment images and data to a management platform. Its underlying hardware combines wheels with articulated legs. DEEP Robotics lists a 15-kilogram working payload, 2.5 hours of loaded runtime and replaceable batteries for the M20. Those are manufacturer specifications. Putting such a machine into a full working shift means planning charging, battery changes and the order of its assignments. A delivery route competes with later jobs for the same stored energy.

Think of the wheel-legged design as a small cart with knees. Wheels handle rolling travel, while changing the leg position lets it negotiate level changes that would stop an ordinary cart. Carrying supplies and carrying inspection equipment can then use the same mobile base. A camera brings a scene back to a screen; a person uses it alongside an understanding of the equipment. Getting to the right place and helping someone decide what to do are useful stages in their own right, distinct from physically repairing a fault.

There is a relevant US precedent in Boston Dynamics’ work with bp. Its case study describes tests at a simulated industrial site in Texas, US, where grated stairs initially caused trouble, followed by adjustments, a refinery trial and plans for at least three months offshore. The aim was to make the robot familiar as a working tool. That experience points to the real test of usefulness: fitting into repeated work rather than completing a single route. Qiliyu remains a capability trial. For lighthouse keepers, progress would mean fewer loaded climbs and fewer trips simply to read equipment, while maintenance decisions and responses to problems stay with people.

Original source published: Guangzhou Daily, November 12, 2026, 23:41 Beijing time / 10:41 AM EST.