Running

A size-adjustable instrumented track to measure vertical ground reaction forces: a validation study

INSTRACK — R. Roinson, A. Karamanoukian, J.-P. Boucher, N. Vignais. IEEE Sensors Journal, 2025.

Two-step validation of the size-adjustable instrumented track: a static protocol on each sensor (mean error of 1.27 N, i.e. 0.3%, and repeatability < 2 N), then a running protocol compared with a reference force plate (excellent agreement, ICC > 0.98, for all vertical force parameters).

Download the PDF · DOI: 10.1109/JSEN.2025.3552967

Mesure du centre de pression à partir d'une piste instrumentée en course à pied : étude de la précision

INSTRACK — R. Roinson, A. Karamanoukian, J.-P. Boucher, N. Vignais. Oral communication, Société de Biomécanique congress, 2026.

(Centre of pressure measurement from an instrumented track in running: an accuracy study.) Comparison of the centre of pressure (CoP) measured by INSTRACK with that of two reference force plates (AMTI), on 13 runners at two paces. Mean difference of 7.1 mm along the running direction and 3.6 mm across the width, with the same heel → forefoot path for both devices. Accuracy decreases at the beginning and end of the step, when the force is low. The CoP calculation is described on the INSTRACK page.

Prediction of running anteroposterior ground reaction forces from the vertical component through a deep learning approach

INSTRACK — R. Roinson, A. Karamanoukian, J.-P. Boucher, N. Vignais, 2026.

A convolutional neural network estimates the anteroposterior force (braking and propulsion) from the vertical force alone, as measured by INSTRACK. Trained and tested on 9,500 running steps from 19 runners at 5 speeds, it reaches a mean relative error of 7.4% (correlation r = 0.98). The effect of this error on joint moment calculation is small at the ankle and larger at the knee and hip.

PDF available soon.