Direct User Control of Knee Exoskeletons (In progress thesis)

This project aims to advance the field of assistive lower limb exoskeletons by developing a knee exoskeleton capable of direct user control. The proposed system introduces a novel control strategy that allows the user to manually operate the exoskeleton’s actuators via a human input device (HID), enabling real-time customization of torque profiles tailored to specific movements. This approach diverges from traditional methods, which optimize exoskeleton control based on metabolic cost or pre-programmed motion profiles, and instead prioritizes user comfort, adaptability, and responsiveness.

The exoskeleton will be built on the foundation of the Modular Backdrivable Lower-limb Unloading Exoskeleton (M-BLUE), an open-source platform designed for lower-body assistance. While M-BLUE introduced backdrivable actuators and lightweight mechanical design, it faced limitations such as sensor inconsistencies, discomfort, and the inability to accommodate direct user control. These limitations will be addressed through mechanical upgrades, improved actuator control, and enhanced sensor integration to ensure stable attachment and user comfort.

The control architecture will employ an Extended Kalman Filter (EKF) for real-time gait estimation and sensory fusion, enabling the exoskeleton to adapt seamlessly to varying terrains and dynamic activities. By combining the EKF with direct user input, the system will be capable of both automatic and manual adjustments, optimizing torque profiles in real-time based on the user’s immediate needs and preferences. This dual-layered control strategy, integrating human-in-the-loop feedback with advanced sensor fusion, aims to create an exoskeleton that not only improves mobility for individuals with lower limb impairments but also enhances the user’s ability to perform tasks requiring strength and endurance.

The project’s ultimate goal is to develop an assistive lower limb exoskeleton that provides a versatile, comfortable, and responsive solution, moving beyond lab-bound prototypes to real-world applications. The successful implementation of this system could pave the way for commercially viable, adaptive exoskeletons that enhance mobility and quality of life for a wide range of users.

Recommended citation: Z. Bucknor-Smartt, J. Mustafa, W. Bannick, L. Graves, M. Korwani, G. C. Thomas, “Direct User Control of Lower Limb Assistive Exoskeletons”, IEEE Robotics and Automation Letters, June 2025 (In progress) [Journal paper + undergraduate thesis]