Epidemic contact tracing with smartphone sensors

Abstract

Contact tracing is widely considered as an effective procedure in the fight against epidemic diseases. However, one of the challenges for technology based contact tracing is the high number of false positives, questioning its trustworthiness and efficiency amongst the wider population for mass adoption. To this end, this paper proposes a novel, yet practical smartphone-based contact tracing approach, employing WiFi and acoustic sound for relative distance estimate, in addition to the air pressure and the magnetic field for ambient environment matching. We present a model combining six smartphone sensors, prioritising some of them when certain conditions are met. We empirically verified our approach in various realistic environments to demonstrate an achievement of up to 95% fewer false positives, and 62% more accurate than Bluetooth-only system. To the best of our knowledge, this paper was one of the first work to propose a combination of smartphone sensors for contact tracing.

Contact Tracing Apps: Lessons Learned on Privacy, Autonomy, and the Need for Detailed and Thoughtful Implementation

“…Additionally, although Bluetooth-based tracing is generally considered more accurate than GPS-based strategies, significant issues are present with signal attenuation and accuracy. Signal absorption and reflection by the surrounding environment can lead to inaccuracies in reported distances between users [ 46 , 47 ]. In a recent paper on this topic, Leith and Farrell [ 48 ] evaluated the efficacy of Bluetooth LE technology for COVID-19 contact tracing in real-world environments, including users walking in the city, at a meeting table, in a train carriage, and grocery shopping, as well as assessing the impact of device orientation, use of a handbag, and type of indoor wall on signal attenuation.…”

“…Alternatively, merging technical strategies may result in increased overall specificity, particularly for popular Bluetooth-based approaches. A recent paper from Nguyen et al [ 46 ] explored the idea of a multi–smartphone-sensor system for contact tracing using Bluetooth combined with barometer (effected by altitude and winds for indoor or outdoor environments), magnetometer (dynamic time warping used to interpret magnetic field vectors), microphone (high frequency, low amplitude short chirps emitted and time of flight measured), and WiFi data (reliant on a grid of WiFi hotspots). With added distance-based readings (microphone, WiFi), accuracy was increased from 25% to 65%, and the additional inclusion of environmental readings (barometer, magnetometer) further increased accuracy to 87%.…”

Epidemic Exposure Tracking With Wearables: A Machine Learning Approach to Contact Tracing

“…Many works leverage geolocation information [17], [18] and proximity sensing [4] to monitor the interaction between any two users. Besides homogeneous sensing, there are also works exploiting the heterogeneous sensing features to improve the distance estimation [19]. However, these works fail to consider the location of the smartphone when the users are doing grocery or working.…”

Artificial Intelligence in Surveillance, Diagnosis, Drug Discovery and Vaccine Development against COVID-19

“…While such contact tracing may result in high false positives, a novel method has been proposed which uses data from six different smartphone sensors for contact tracing. This method outperforms other methods and identifies ~95% fewer false positives, reaching up to ~87% accuracy [ 87 ]. In Table S1 , we have summarized some of the AI/ML models uses and their application in the surveillance of COVID-19.…”