Here we include IMU-driven (OpenSense) and optoelectronic-driven (OpenSim) inverse kinematics from 14 healthy young adults (7 males and 7 females) who performed five 7-minute trials of walking on a treadmill: (i) at preferred speed and with preferred arm swing, (ii) at 70% preferred speed and with preferred arm swing, (iii) at 130% preferred speed and with preferred arm swing, (iv) at preferred speed and with active arm swing, and (v) at preferred speed and with the arms bound to the sides. Models were based on the Rajagopal 2015 full-body model and were modified to lock the toe joints.

Our manuscript (https://www.mdpi.com/1424-8220/21/22/7690) reports on the concurrent validity and sensitivity of the OpenSense model for joint angle timeseries, ranges of motion, and several (stride-to-stride) motor variability features from these angles: (a) magnitude of variability, (b) local dynamic stability, (c) persistence of range-of-motion fluctuations, and (d) regularity.