IoT System for Monitoring and Analysing Physiological Variables in Athletes

This is the authors' abstract. We don't add key points for this paper.

IoT has had a wide diffusion in monitoring variables of interest in applications such as health, agriculture, environment, and industry, among others. In the context of sport, although wearable devices can monitor physiological variables, they are limited by the fact that they are linked to proprietary applications, have limited storage and perform analyses based on descriptive statistics without including the application of data analytics models. In this paper, we present the construction of an IoT system for monitoring and analysing physiological variables in athletes based on the use of unsupervised learning models. This system is articulated in the IoT four-layer architecture (capture, storage, analysis and visualization). It has the advantage of benefiting from the data provided by commercial devices, storing them in a non-relational database and applying clustering algorithms to the historical data. The proposed system is intended to serve as a reference to be replicated in sports training contexts in order to take advantage of the data provided by commercial wearable devices for decision-making based on the use of machine learning models.

May enable open-source athlete monitoring platforms; leaves open prospective validation before clinical adoption.