Evaluation of Inertial Measurement Units and Non-Invasive Muscle Contraction Sensors for Monitoring Cycling Biomechanics

The paper deals with monitoring cycling biomechanics using different sensors for better fitting and injury prevention. A method for determining knee joint angle using inertial sensors is presented. Additionally, two sensor fusion methods are used: a complementary and a Kalman filter. Evaluated on a 5-minute measurement, the root mean square error is below 2°. Results indicate that the method could be used for position monitoring and improving one's technique and fitting. Cheaper and more efficient solutions could be developed for motion tracking with its adoption. Additionally, usage of non-invasive sensors, could provide valuable insight into cycling biomechanics, with the patterns resembling electromyographic ones.