Bojan Nikolic
Improving COVID-19 contact tracing using wearable humidity sensors
Abstract
Existing technological solutions for contact tracing rely on transmission and reception of decimeter radio waves (2.4GHz/12cm/"Bluetooth"). Unsurprisingly there are frequent real-world situations where propagation of this radiation is not a good predictor of transmission of a virus carried an aerosol: eg a thin non-conductive barrier such as an internal wall or window will hardly affect the radio waves while completely stopping the advection of an aerosol. I show that such shortcomings of radio-only contact tracing can be largely resolved by augmentation with additional sensor information, in particular with humidity and pressure measurements. Development of MEMS sensors, miniaturised storage and "IoT" technologies has made measurements of sufficient accuracy, storing and processing the data practical at surprisingly low cost. Therefore this technique provides feasible, efficient contact tracing in the near term for high-risk settings with controlled access (eg hospitals, care homes, office buildings). In this talk I will describe the prototype data acquisition system that we built and the data processing strategy to improve the risk estimation in COVID-19 contact tracing. Potential impact on COVID-19 contact tracing in practice will be discussed. The research and development of this project has been supported by Innovate UK.
Our speaker
Dr B Nikolic is a radio astronomer and expert in observational science and processing of very large data sets. After obtaining his PhD at the University of Cambridge he has played a key role in design and commissioning of the worlds largest radio telescope facilities: the 100m GBT, the Atacama Large Millimetre Array (ALMA) and currently, the Square Kilometre Array for which is one of the software architects. Dr Nikolic also runs a software and engineering consultancy business and has previously worked in the financial services industry. At the outbreak of COVID-19 Dr Nikolic realised that contact tracing will be a key way of reducing the incidence of the disease and, with support of Innovate UK, an agency of the UK government, begun the development of the technology described in this talk. Dr Nikolic is based at University of Cambridge and BN Algorithms Ltd.
