Self Calibrating Wearable Active Running Asymmetry Measurement and Correction

We present a novel self-calibrating wearable device to improve running training by active vibro feedback. The system consists of a pair of spandex shorts embedding: a processor unit, 2 three-axial accelerometers, 2 vibro motors, a SD card reader\writer module and a real time clock. Two supplementary wireless accelerometers are located on the shoes. We present an algorithm to compute gait asymmetry from the four sensors data. The main novelty is the auto calibration algorithm which uses principal component analysis on each sensor based on kinematic assumptions. The system provides two important advantages: data logging and real time active correction. The active correction is performed sending signals to the user in real time via vibrations cells (rumble feedback). The vibration signals are sent to the specific leg and its intensity is proportional to the entity of required correction. This training system for running can be very useful to athletes and to sport medicine in order to improve speed, posture, fatigue and reduce joints osteoarthrotic problems.