COVID-19 Contact Tracing Analysis with Bluetooth Technology Using Raspberry Pis
Dean Zhang · bioRxiv (Cold Spring Harbor Laboratory) · 2022
Abstract Contact tracing, a method for detecting and preventing the spread of a disease, can become more efficient by becoming automated, rather than being done manually. Bluetooth based contact tracing is a potential method for creating an automated method for contact tracing. However, Bluetooth signals cannot always predict something with complete accuracy due to many subtle obstructions. Received signal strength indicator (RSSI), a value produced when Bluetooth devices send and receive signals, generated from Raspberry Pis can predict the distance between the two devices. By running many experiments with variations of obstructions, I was able to successfully create models to correlate RSSI values and distance with a potential success rate of 91.973%. For my research, a success is determined to be anything but a false negative. Despite the limitations when conducting my research, the inaccuracies of my results prove Bluetooth based contact tracing to not be a reliable method for determining people who have been in contact with an index case in the real world.