Figure founder Brett Adcock has posted video of the company’s humanoid climbing 15 feet up a ladder and 15 feet back down with no human input, three weeks after its first autonomous ladder climb. Figure credits its Helix AI system, whose System 0 layer runs balance and contact control a thousand times a second using stereo camera vision alongside the robot’s own sense of body position. Figure has not released the technical details and the demonstration has not been independently verified.
Three weeks ago, Figure’s humanoid climbed a short ladder on its own.
This week it went fifteen feet up, and fifteen feet back down.
Founder Brett Adcock posted the clip on Saturday. No operator, no tether.
The climb up is impressive. The climb down is the part engineers will be staring at.
Going up, the robot sees the rungs it reaches for.
Coming down, it moves backwards into space it cannot see, feeling for each rung with its feet, while its own weight tries to peel it off the ladder.
Every step down is a controlled fall that three other limbs have to catch.
Figure says this runs on Helix, its in-house AI. The layer that handles balance and contact, System 0, updates a thousand times a second, and it now mixes stereo camera vision with the robot’s own sense of where its body is.
That is what lets it place a foot on a rung it cannot see.
Figure has not published the technical details yet. The robot is also slow.
The reason it matters is safety. Fixed ladders are everywhere in industry, on tanks, silos, telecom towers and warehouse racks. In the United States, falls, slips and trips killed 844 workers in 2024, out of 5,070 workplace deaths.
Height is where a machine that never gets tired and never gets complacent earns its place.
Figure is now building roughly one robot an hour and says it has delivered more than 350.
Adcock has called wheeled robots a dead end because they cannot handle stairs and ladders. Fifteen feet up and back down is that argument, on video.