Computing Velocity and Force in Strength Exercises Using a Wearable and Sport Performance Monitoring Machine (MYO-QUALITY)

Velocity, a fundamental parameter in sports performance, enables the evaluation of exercise effort, particularly power, defined as the product of velocity and known weight. This work presents a novel integration of commercial wearable technology and a specialised performance machine for the estimation of velocity. A high-precision sports performance machine engineered for strength training (M1 MYO-QUALITY) provides real-time data on force, velocity, and power. A commercial wearable device acquires data from Inertial Measurement Units comprising an accelerometer and a gyroscope through a specially designed application. Subsequently, our deep learning model accurately estimates the velocity with a mean absolute error (MAE) as low as 0.0245 and an R-squared value of up to 0.9811 in relation to the professional sports machine, thus facilitating the measurement of this parameter in scenarios where MYO-QUALITY is not accessible to athletes. In addition, four exercises were classified to support the incoming research, achieving a precision and recall of 1.00 in most cases.