Place two walking demonstrations side by side and the movement is the first thing you notice. A more useful comparison begins by asking what each demonstration was designed to establish.
Then, isolate the walking problem
Honda’s archive presents E0 as its 1986 experimental walking machine and follows successive efforts toward more stable, flexible movement. The chronology lets readers examine walking as a sequence of engineering problems. Honda’s development archive ↗
Now, examine the platform and configuration
Unitree’s G1 page presents a humanoid platform and distinguishes configurations. When watching G1 footage, record the version and equipment shown alongside the task. Unitree’s product page ↗
The basic step
Ask what the experimental setup isolates and how each movement is controlled.
Open the archival video →A modern platform
Ask which motions are demonstrated, the operating mode, and how the setup varies between runs.
Open the introduction →Keep the comparison fair
Walking speed alone leaves out terrain, external support, body dimensions, stability, and recovery. These sources do not form a shared benchmark. The historical comparison shows how the visible task and the surrounding engineering questions have changed.
For our next walking comparison, we would look for continuous trials, disclosed assistance, repeated conditions, and recovery after an unexpected disturbance. Those details help turn a visual impression into a technical assessment.
Sources & reading
Sources reviewed November 5, 2026. Company claims and author-reported research are attributed in the text.