Energy efficient contact tracing and social interaction based patient prediction system for COVID-19 pandemic

Abstract

Due to the spread of Coronavirus disease 2019 (COVID-19), the world has encountered an ongoing pandemic to date. It is a highly contagious disease. In addition to the vaccination, social distancing and isolation of patients are proven to be one of the commonly used strategies to reduce the spread of disease. For efficient social distancing, contact tracing is a critical requirement in the incubation period of 14-days of the disease to contain any further spread. However, we identify that there is a lack of reliable and practical social interaction tracking methods and prediction methods for the probability of getting the disease. This paper focuses on user tracking and predicting the infection probability based on these social interactions. We first developed an energy-efficient BLE (Bluetooth Low Energy) based social interaction tracking system to achieve this. Then, based on the collected data, we propose an algorithm to predict the possibility of getting the COVID-19. Finally, to show the practicality of our solution, we implemented a prototype with a mobile app and a web monitoring tool for healthcare authorities. In addition to that, to analyze the proposed algorithm's behaviour, we performed a simulation of the system using a graph-based model.