BLOG – Three Max Planck Institutes join forces to develop a next generation contact tracing system
A blog post by Heiner Kremer, Ph.D. student in the Empirical Inference Department at MPI-IS
In a recent work published at Nature Scientific Reports, researchers from the Max Planck Institutes for Intelligent Systems, Software Systems, and Security and Privacy as well as other institutions joined forces to devise PanCast, a novel privacy-preserving and inclusive system for epidemic risk assessment and notification.
The mitigation of the Covid-19 pandemic required unprecedented restrictions of individual freedoms and interactions among the global population. While the current pandemic slowly comes to an end and restrictions are gradually lifted in many places, it revealed society's vulnerability to such occurrences and emphasized the necessity to prepare for similar events in the future. In the context of the Covid-19 pandemic, it became clear that successful epidemic mitigation requires effective testing, contact tracing, and isolation of infected individuals. Among these measures, contact tracing can be an important tool to help direct limited test resources to those most likely to be infected. Unfortunately, the success of recently developed digital contact tracing systems has been limited, partly owing to poor interoperability with manual contact tracing, low adoption rates, and a societally sensitive trade-off between utility and privacy that prevents the use of valuable location information.
Building on their combined expertise in software systems, machine learning and security and privacy, the researchers developed PanCast with the goal of addressing some of these limitations. The paper “Listening to bluetooth beacons for epidemic risk mitigation” was published in April 2022 at Nature Scientific Reports.
In contrast to many previous digital tracing systems, which focus on person-to-person encounters, their system relies on a person-to-infrastructure design. To set up PanCast, Bluetooth beacons which continuously broadcast ephemeral IDs would be placed in strategic locations (e.g., shops, restaurants) where infections are likely to happen. Individuals would carry devices that listen to these beacons passively (i.e., without transmitting anything), store the beacons' ephemeral IDs, and later compare the stored IDs against risk information that is broadcast via a subset of the beacons, called network beacons. The user devices could be inexpensive, zero-maintenance, small electronic dongles in the form of cards or key fobs, or more sophisticated devices such as smartphones. When an individual tested positive for the infectious disease, they could choose to disclose (a selected subset of) the list of ephemeral IDs stored in their user device. The individual would have to explicitly authorize the transmission of data from their device via a trusted terminal, which could be part of a kiosk installed in a test center, clinic, or doctor's office, or a personal device (e.g., a smartphone or a computer) owned by the user. The information about which locations were contaminated at which times (which depends on users' visits as well as on location features) would then get included in the broadcast risk messages. A subset of PanCast beacons would act as network beacons, which broadcast risk information associated with times when individuals who tested positive were near specific beacons. A user learned their risk of contagion due to their presence at sites visited by diagnosed individuals, but would not learn anything about locations visited by other individuals.
The Figure below shows an overview of PanCast's architecture.
The researchers argue that, by design, PanCast facilitates participation of technologically or economically challenged individuals who cannot or do not wish to use smartphones. Moreover, it ensured data minimization in accordance with existing regulations for (manual) contact tracing—a healthy individual would use the system in a purely passive “radio” mode which allowed PanCast to achieve a similar level of privacy as existing digital contact tracing systems.
The study cites multiple lines of evidence showing that for infectious diseases like Covid-19 different locations (e.g. indoor vs. outdoor) admit vastly different transmission rates, a fact that cannot be accounted for in smartphone-to-smartphone based approaches. Based on epidemiological simulations the researchers show that the use of PanCast’s environmental information could significantly increase the accuracy of contact tracing and isolation measures and thus lead to stronger epidemic mitigation while imposing a smaller burden on the population. Moreover, the work shows that by making use of location information PanCast could achieve bidirectional interoperability with manual contact tracing: If a PanCast user tested positive, they could use the information saved in their device to better recall visited locations during a contact tracing interview. Conversely, if a diagnosed individual did not use PanCast, a human contact tracer could manually create an entry in the risk database for any locations the individual recalls visiting. The epidemiological simulations suggest that due to this, PanCast required a significantly smaller level of adoption than existing smartphone-to-smartphone based systems to achieve the same level of epidemic mitigation. Furthermore, the study shows that equipping only a small fraction of 5-10% of the sites with the highest risk with beacons would already be sufficient, which is an important aspect for the practical feasibility of the proposed system. The authors conclude that their system could also be combined with existing smartphone-to-smartphone based approaches and be employed in an incremental and local fashion, e.g., by companies or educational institutions on their respective grounds.
The recent events during the Pandemic have revealed modern society’s vulnerability to infectious diseases like Covid-19. While similar pandemics will most likely happen again in the future, one can hope that the insights gained in the current pandemic will enable humanity to better prepare for such occurrences. In this context, contact tracing systems like the one developed by the Max Planck researchers might get to play an important role for the mitigation of future pandemics.