Ground Reaction Force Analysis of Robot-Assisted Walking Using Inertial Sensors

Abstract

The growing elderly population faces increasing mobility challenges, particularly knee joint problems during walking, which limit independence and increase the risk of falls. Traditional force plate measurements for ground reaction force (GRF) analysis are constrained to laboratory environments, limiting the assessment of assistive technologies in natural settings. This study developed a mathematical framework to estimate the GRF during robot-assisted walking using inertial sensors, with the BART LAB AI-Elderly Supporting Robot designed specifically for elderly mobility assistance. Ten healthy participants performed walking trials under controlled and robot-assisted conditions to validate the proposed measurement framework. Inertial measurement unit sensors captured motion data, and mathematical models estimated the GRF by incorporating gravitational, dynamic, and robotic assistance components. Robotic assistance achieved a 28.9 ± 8.6% peak GRF reduction, with 90% of the participants showing clinically meaningful improvements (>20%). Individual reductions ranged from 12.4-39.8%, equivalent to a substantial joint load reduction. The framework successfully quantifies the effectiveness of robotic assistance without force plate constraints, providing a foundation for the objective evaluation of elderly mobility support systems before clinical implementation.