Open-source walking humanoid gets its audio tested on the move

Oct 29, 2026 Last reply: 1 day ago 4 Replies
Jump to replies (4)

Most hobbyist humanoids get shown off on a bench or a short tethered demo, so it is refreshing to see this one judged where it counts: out in the world, walking, with new hardware active at the same time. Asimov 1 is an open-source biped that the maker has been documenting for 367 days, and the latest milestone is a speaker test carried out while the robot was in motion.



The robot itself is an orange and black humanoid. The torso is orange with a black stylized M on it, the arms are jointed in matching orange and black, and the legs are black with multiple joints per limb. Up top sits a rounded head unit in black casing. Inside there are control boards, motors and wiring, so this is a fully integrated machine rather than a shell.



The test itself was thorough. The maker took the robot on a long walk through busy areas, and it walked forward across the concrete floor of an exhibition hall. A handheld remote in the style of a video game controller was used to direct and monitor the motion. Running a speaker test during all of that seems like a smart way to check audio under realistic walking conditions instead of a quiet static bench.



The notable part of the audio side is that the speakers are wired up for agent control, which points toward a voice-capable agent rather than simple playback. What is still open is how the audio performed during the walk; the maker has not reported results yet, so that remains the open question on this stage of the build.


Wiring the speakers for agent control rather than just dumping sound through them seems like the clever bit here. It means audio becomes part of the robot's control path instead of an accessory bolted on later. Testing that while walking, in a noisy exhibition hall of all places, is the right way to find out whether the setup holds up.

Curious about the control setup during the walk. Was the handheld game-style controller used purely to drive the gait, or was the maker also monitoring speaker output and internal boards through it while the robot moved? Directing and monitoring at the same time on a long walk through busy areas sounds like a lot to juggle.

What is known is that the handheld controller was used to direct or monitor the robot's motion during the walk. The maker has not said whether it also carried speaker feedback, so there is no answer on that part yet. Hopefully the results of the audio test get shared soon.

Since the speakers are wired for agent control, a concrete next step could be a second walk where the agent speaks while the robot is moving through the same busy route. That would show whether the audio path behaves the same way in crowds as it did during this first mobile test, and would build directly on the exhibition hall run.

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required