Design, control and preliminary validation of an active hip exoprosthesis
(2026)
Files
thesis_LD_26_09_16.pdf
Embargo Access until 2027-09-30 - Adobe PDF
- 7.62 MB
Details
- Authors
- Supervisors
- Abstract
- (en) Hip disarticulation and hemipelvectomy are the most proximal levels of lower-limb amputation. These rare but profoundly mutilating interventions leave only a minority of patients able to walk with an exoprosthesis, and those who do face considerable challenges such as high metabolic expenditure and gait asymmetry. While the field of prosthetics has seen remarkable advances in active ankle and knee devices over the past decade, the hip joint has remained largely unaddressed. Active hip exoprostheses face specific design challenges such as physical access to the joint, large range of motion, and high torque requirements. The fundamental challenge of controlling a device deprived of volitional and rhythmic input from the missing limb also deserves attention. This thesis addresses these open challenges through the design and control of an active hip exoprosthesis, developed from mechanical concept to reality to real-time user tests. On the mechanical side, a novel exoprosthetic hip joint is introduced, built around a tilted double parallelogram mechanism that places the exoprosthetic center of rotation at the anatomical location of the hip joint. This remote center of motion architecture achieves both compactness and biomechanical fidelity, resulting in a device with a horizontal footprint of 140 mm, a frontal width of 136 mm, and a mass of only 3.30 kg excluding battery and power electronics. The range of motion spans from-30 deg to 90 deg, accommodating the full demands of walking, sitting and their respective transitions while remaining compatible with existing prosthetic shell designs. On the control side, a real-time kinematic estimator is proposed to infer the required exoprosthetic hip joint motion from signals recorded on the intact, contralateral leg, relying on only three IMUs placed on the pelvis and intact leg. This estimator fuses two complementary approaches, Complementary Limb Motion Estimation and Adaptive Oscillators, through a logistic weighting function. The fusion strategy allows the system to respond accurately during locomotion transitions while converging toward a more stable, oscillator-driven estimate during steady-state walking. On the user test side, an experimental session and case study with a subject with unilateral hip disarticulation demonstrate the impact of the exoprosthesis and its controller. Preliminary results suggest that the exoprosthesis may reduce the metabolic cost of walking compared to a commercial passive exoprosthesis. More time needs to be spent on analyzing the gait kinematics and kinetics, but preliminary results are encouraging. Together, these contributions form AUHRA (Active Unit for Hip Robotic Assistance), a complete active hip exoprosthesis system. This work lays a concrete foundation for restoring physiological locomotion to individuals living with the most proximal level of lower-limb amputation.
- Affiliations
Citations
Devillez, L. (2026). Design, control and preliminary validation of an active hip exoprosthesis. https://hdl.handle.net/2078.5/280839