Human Digital Twin for Fitness Management

Barbara Rita Barricelli,Alessandro Petrini,Stefano Valtolina,Elena Casiraghi,Jessica Gliozzo

doi:10.1109/access.2020.2971576

Abstract

Our research work describes a team of human Digital Twins (DTs), each tracking fitness-related measurements describing an athlete’s behavior in consecutive days (e.g. food income, activity, sleep). After collecting enough measurements, the DT firstly predicts the physical twin performance during training and, in case of non-optimal result, it suggests modifications in the athlete’s behavior. The athlete’s team is integrated into SmartFit, a software framework for supporting trainers and coaches in monitoring and manage athletes’ fitness activity and results. Through IoT sensors embedded in wearable devices and applications for manual logging (e.g. mood, food income), SmartFit continuously captures measurements, initially treated as the dynamic data describing the current physical twins’ status. Dynamic data allows adapting each DT’s status and triggering the DT’s predictions and suggestions. The analyzed measurements are stored as the historical data, further processed by the DT to update (increase) its knowledge and ability to provide reliable predictions. Results show that, thanks to the team of DTs, SmartFit computes trustable predictions of the physical twins’ conditions and produces understandable suggestions which can be used by trainers to trigger optimization actions in the athletes’ behavior. Though applied in the sport context, SmartFit can be easily adapted to other monitoring tasks.

Highlights

Nowadays, the extension of Internet connectivity into physical devices and everyday objects has radically transformed interactions and communications happening in all the aspects of human life
THE TEAM OF DTS we describe the system developed to create the artificial intelligence (AI) of a Digital Twins (DTs), which monitors the health status of a generic person, described by a set of parameters, and provides predictions about the fitness outcome and suggestions to improve it
EXPERIMENTAL RESULTS we describe the experimental evaluation of a team of DTs used by trainers for supervising the activity and fitness level of a team of 11 athletes

Summary

INTRODUCTION

The extension of Internet connectivity into physical devices and everyday objects has radically transformed interactions and communications happening in all the aspects of human life. To help coaches and trainers in the monitoring task, we have extended SmartFit with machine learning techniques that enable predictions and we have developed a method, based on counterfactual explanations, to compute suggestions for improving athletes’ performance. THE TEAM OF DTS we describe the system developed to create the artificial intelligence (AI) of a DT, which monitors the health status of a generic person (i.e. a physical twin, which is an athlete in this case), described by a set of (fitness related) parameters, and provides predictions about the fitness outcome and suggestions to improve it. When novel measurements are provided, they are coded by eventually imputing missing data and by selecting only the most informative features (identified during training) At this stage, they are input to the trained classifier that provides a fitness score prediction and retrieves the coded historical data from the data storage to compute suggestions for improving the predicted score. Whenever some measurements with their ground truth label are provided by the expert trainers, they are added to the historical data and the training phase is repeated to update the DTs knowledge with the novel labeled data

TRAINING THE DT

EXPERIMENTAL RESULTS

ANALYSIS OF COMPUTED RESULTS

CONCLUSION