Three-dimensional motion analysis of upright bipedal walking android model

• Published in: Clinical biomechanics (Bristol) Vol. 129; p. 106620
• Main Authors: Sanaka, Kouji, Sekiguchi, Yusuke, Kurosawa, Daisuke, Sugimura, Seiichi, Hashimoto, Ko, Takahashi, Kohei, Ebihara, Satoru, Murakami, Eiichi, Aizawa, Toshimi
• Format: Journal Article
• Language: English
• Published: England Elsevier Ltd 01.10.2025
• Subjects: 3D motion analysis; Android model; Bipedal walking; Center of pressure; Physical Medicine and Rehabilitation; Trunk movement; Android model; Center of pressure; 3D motion analysis; Bipedal walking; Trunk movement
• ISSN: 0268-0033; 1879-1271; 1879-1271
• DOI: 10.1016/j.clinbiomech.2025.106620

We previously developed a bipedal android model driven by trunk motion via psoas major contractions. A mechanically stabilized principle model was created to preserve gait mechanics, enabling autonomous bipedal walking and reliable center of pressure measurement by addressing knee and foot–ankle joint instability. This study investigated whether the center of pressure trajectory generated by the trunk-driven model approximates that of humans using a three-dimensional gait analysis system. Thirty-five markers were attached to healthy subjects versus 24 markers to the principle model. Ground reaction force data were captured at 1200 Hz and analyzed using motion analysis and numerical software. The center of pressure trajectory of the right foot during the stance phase was compared between the principle model and healthy subjects. Center of pressure trajectories were generally similar during the double-limb stance and single-limb support phases. The principle model showed differences such as a backward deviation of the center of pressure in the direction opposite to walking progression during the single-limb support phase, especially between 47.5 % and 61.5 % of the phase. The trajectory did not extend toward the forefoot, likely due to the shorter stride length, slower walking speed, and prolonged single-limb support duration (0.91 ± 0.05 s vs. 0.41 ± 0.05 s in healthy subjects). Conventional gait analysis assumes passive trunk motion following lower-limb activity. In contrast, the principle model demonstrates trunk-driven motion with passive leg swing, partially replicating human center of pressure trajectories. This suggests a trunk-driven approach may offer insights for gait analysis. •The bipedal android model, driven by trunk motion, mimics human center of pressure.•The trunk is crucial in bipedal walking, not just passively following limb motion.•The model aids biomechanical studies post-treatment and predicts leg abnormalities.