Development and Validation of a Power Meter for Functional Fitness Athletes

Sammanfattning

Power meters have been available in endurance sport since the 1980s and provide useful insights to individualized training programs in order to improve the effectiveness of training sessions. This also helps in reducing injury risks. In gym context, especially for functional fitness (FF), nowadays it is not possible to measure the power output (PO) of athletes, thus making it difficult to understand how different training sessions can affect athletes’ bodies, both in terms of performance and fatigue.The main aim of this study is to develop a new inertial measurement unit (IMU)-based power meter to be used in FF contexts and to validate it against a gold standard, a combination of a motion capture system (MoCap) and 4 force plates. Additionally, an investigation about the best configuration of sensors to be used is carried out.Seven well trained FF athletes took part in this study. They were asked to perform some FF movements (burpees, clean and jerks, lunges, snatches and thrusters) while being recorded by 3 IMUs placed on them (chest, wrist and ankle) and by the gold standard system. A Python algorithm applied to IMU data estimated PO and this was compared with the PO computed from the gold standard. The new power meter estimated the PO with two different approaches: the force approach (FA), based on the formula P=F*v, and the energy approach (EA), based on mechanical energy.Both FA and EA showed very high to almost perfect correlation with the gold standard, with the best correlation (0.927) provided by FA with the configuration with 2 sensors at chest and wrist, when comparing the full PO curve. Analyzing average positive and negative power, total positive and negative work and peak positive and negative power, FA showed better estimations despite a general underestimation of around 10% for all the metrics analyzed and with a RMSE around 25% (after the correction of the raw estimations, while raw PO overestimates the MoCap by around 33%).The correlations between IMU and MoCap found in this study are in line with the values found by other studies. A comparison of the other metrics with literature is difficult due to a lack of similar types of studies. Despite that, studies comparing IMU and MoCap to estimate PO during countermovement jumps found a percentage overestimation around 25%, close to the raw estimation of around 33% of this study.This study showed that the best method to estimate PO with IMUs is FA. Moreover, even if the 3-IMU power meter showed promising results, the configuration with just two sensors (at chest and wrist) provided slightly better estimations.