This dataset reports the lower-limb kinematics and kinetics of twenty able-bodied participants walking at multiple stiffness levels (1000 kN/m, 80 kN/m, 40 kN/m, and 25 kN/m) and speeds (0.8 m/s, 1 m/s, and 1.2 m/s). Data were recorded by a Vicon motion capture system and, a unique robotic treadmill, the Variable Stiffness Treadmill 2 (VST 2). The data can be used to identify specific gait patterns, balance strategies, and muscle activation profiles that individuals adopt over compliant surfaces. This information is crucial for designing better controllers for prosthetic limbs, improved rehabilitation protocols, and adaptive assistive devices that can enhance mobility on compliant surfaces, and for optimizing controllers for robotic walkers.
The experiment’s data is saved in twenty (20) MATLAB structs, each one corresponding to the data collected from each of the 20 participants. Each struct is named “SUBXX”, where XX represents the single or double-digit participant ID as listed in Table 1 of the accompanying manuscript, e.g. “SUB01” represents the data for participant ID #1, “SUB10” for ID #10 etc. Each struct file contains the anthropometric and demographic information of each participant within the cell “SUBXX.subjectInfo”, as well as the raw experiment data of all trials parsed in gait cycles. The sampling rate for the structure by data type is as follows:
• Markers, Joint Angles, Forces, Moments, Powers, CoM - 100Hz
• Force mats (GRF, CoP) – 65 Hz
• EMG, IMU– 2000 Hz
• Heart rate data – 1 Hz (data was resampled to 100Hz for synchronization purposes)