Multimodal data on bipedal locomotion during prolonged treadmill recordings at varying speeds

Abstract

This dataset provides curated multimodal recordings of human bipedal locomotion collected from 18 healthy adults during prolonged treadmill activity at four controlled speeds (4 km ∙ h−1, 6 km ∙ h−1, 9 km ∙ h−1, and 12 km ∙ h−1). Each participant completed three continuous treadmill trials comprising multiple speed intervals, totalling approximately forty minutes of recorded activity. Kinematic data were captured using an optical motion capture system with a full-body marker setup. Additional movement-related signals were recorded using inertial measurement units attached to the sacral crest and to the lateral side of each shoe. These units provided acceleration and angular velocity data. Kinetic data were recorded simultaneously from dual-belt treadmill force plates, and plantar pressure data were recorded using pressure-sensing insoles. The resulting dataset provides synchronised, multimodal data from all systems, including detailed marker and force trajectories, as well as pressure, acceleration and angular velocity signals. Vicon processing pipelines and MATLAB scripts are provided to ensure full reproducibility. The dataset adheres to FAIR principles and supports a range of biomechanical and movement science applications.