A real-time gait measurement system for wearable ankle-foot assistive device

Wearable robotics is a promising approach for assistance to people with gait deficits. The efficient sensory feedback is of important for closing the system loop of adaptive robotic control. This work presents a novel real-time gait measurement system for ankle-foot assistive device based on data from wearable sensors. This system incorporates two inertial measurement units (IMUs) attached to the shank and foot respectively. Recognition of gait stance and swing phases is performed by a high-level probabilistic method using angular rate from the IMU attached to the shank. The ankle planar-/dorsiflexion angle is measured by fusing the measured angular rate and acceleration from the two sensors based on a complimentary filter algorithm and sensor-to-segment identification method. The proposed system was experimentally validated in real-time while 10 healthy subjects walking on the treadmill with three different speeds. The outputs were compared to the reference measured by a Vicon motion capture system. Results show that the Bayesian recognition method achieves the recognition of gait phases with an accuracy above 94% and a delay response time below 20ms and accurate IMU-based angle measurements (RMSE below 3.5◦). The results demonstrate the potential of the wearable system for perception in the robotic control of anklefoot assistance.