Auto-Calibrating Foot Vertical Trajectory With Known Velocity and Displacement Estimated by Smart Insoles
Year: 2026
Type: article
Abstract: Accurate estimation of the vertical trajectory of the foot is important for prosthetic control, fall risk assessment, pedestrian dead reckoning (PDR), and localization. Although inertial measurement units (IMUs) offer a practical solution for foot trajectory estimation, their measurements are prone to drift errors. This study proposes two novel auto calibrating approaches called known velocity update (KUPT) and known displacement update (KDPT) to address these errors. By leveraging the known location of the IMU within the insole and the continuously measured foot angle, the velocity during heel contact and toe off can be calculated, which enables KUPT to correct drifting errors. Compared to the traditional zero velocity update (ZUPT), KUPT extends the calibration period from 24% to 60% of the gait cycle. Additionally, during level ground walking, which is the most common walking condition, the net vertical displacement over one gait cycle is known to be zero. Using terrain context information detected by the insole, the KDPT method is proposed to further improve calibration performance. To validate the proposed methods, walking data from 10 participants were collected using a motion capture system as the reference. The KUPT approach achieved a mean absolute error (MAE) of 9.294 cm across all speeds, representing a 17.5% improvement over the conventional ZUPT method. The KDPT method achieved an MAE of 1.335 cm, corresponding to an 88.1% improvement. These results demonstrate the potential of the proposed methods to support applications that require long term and reliable tracking of the foot’s vertical trajectory.
Source: IEEE Access
Authors: Stella A. Ansah, Diliang Chen
Institution: University of New Hampshire
Language: English
Subfield: Aerospace Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: gold
Estimated APC paid: $2,075
Funder: National Science Foundation
OpenAlex ID: W7128310927
