Planning and development of a bio-feedback music advisory system for training activities

Physical and mental wellness represents one of the most pursued and desired aims nowadays. Health, in its most general sense, is of a major concern in the present world scenario. People usually perform running activities to keep the body fit and healthy; within this context, the heart rate results in a vital measure of a person’s health and it can be used to monitor the body’s present condition and hence provide ways to improve it. The heart rate, in particular, is a revealing indicator for the training intensity and stabilizing it can help the user increase the efficiency and guarantee the safety of exercise. According to the desired intensity training level, certain limits for the heart rate can be determined using the maximum heart rate. The objective of this thesis is to demonstrate the feasibility of a support system able to rely on music application to enhance trainers' performances while performing a running activity. Music influences human physiology in many different ways, from both mental and physical points of view. It is then reasonable to engineer a system that emphasize music effects to adapt to a specific use case and drive the interested users to reach their fixed targets in an effective and pleasant way. A motivational support is then provided to the final user to generate a positive perception of the physical effort. Thus, some advanced solutions can be adopted to achieve satisfactory results when deploying music-based support systems. This project aims at providing a fully operative service to users interested in reaching desired training performances with the help of properly selected music tracks. The novelty of the proposed solution consists in an adaption of individual physiologic characteristics to provide an always updated and fully personalized experience that continuously keeps track of cardiac responses to external audio stimulus. After a preliminary evaluation of results achieved by administrating different audio tracks to a trainers' sample, an improved solution based on artificial intelligence and regression analysis was developed for adapting the support service with individuals' performance and cardiac responses. This approach provided acceptable outcomes that justified the choice of a linear model and regressors subset to describe and predict the future heart rate evolution in future time instants.