PiPACT: Examining the Feasibility of Automatic Bluetooth Proximity Detection for Digital Contact Tracing

As cases of the coronavirus disease (COVID-19) soar worldwide, contact tracing has proven to be an essential and effective method of slowing the virus propagation. This paper explores the feasibility of an automatic contact tracing technology that uses Bluetooth signals to determine proximity between two devices. While without obstruction Bluetooth signal strength is effective at proximity sensing, obstructions and variations in device orientation provide sufficient ambiguity such that Bluetooth signal strength alone cannot reliably determine proximity. It is also found that the duration of contact can be accurately measured by the number of transmissions received. More factors other than signal strength, such as those accounting for suboptimal conditions, must be considered to reliably determine proximity.