Automated Exposure Notification: A Technically Beautiful but Empirically Useless Technology

During the COVID pandemic, we were told that one could get sick from close contact with an asymptomatic infected individual.  The problem was that the infected individual might not manifest any symptoms of illness for a good two weeks after initial contact.  How could a sick person, who might barely remember what she ate for breakfast, retroactively reconstruct a list of all contacts over a two-week period?

The Automated Exposure Notification problem formalizes this model, together with additional guarantees for privacy of the individuals involved and security of the system as a whole.  We will present our solution to this problem, that started with a core idea that was initially developed and implemented with colleagues at Boston University, grew into a multi-institution project, and was eventually adopted, in some form, in nearly all modern smartphones.  The solution, based on Bluetooth Low Energy advertisements, had provable privacy and security properties, while facing a number of fascinating engineering challenges in its development.  At the end of the day, despite beautiful theory and practice, the system ended up being largely useless.  Our post-mortem analysis of the project provided some important insight into the application of academic research to the broader world.

Bio:

Ari Trachtenberg is a Professor of Electrical and Computer Engineering at Boston University, where he is also affiliated with the Department of Computer Science and the Division of Systems Engineering.  He received his PhD and MS in Computer Science from the University of Illinois at Urbana-Champaign, and his SB from MIT.

At BU, Dr. Trachtenberg is the co-director of the Red Hat Collaboratory, a $20M five-year collaboration between industry and academia.  He is also a member of the Rafik B. Hariri Institute of Computing and the Center for Reliable Information Systems and Cyber Security, and a founding member of the Center for Information and Systems Engineering.  Dr. Trachtenberg has taken sabbaticals at Red Hat, TripAdvisor, and MIT Lincoln Lab, and served as a visiting professor at the Technion and a distinguished scientist visitor at Ben Gurion university, in addition to broader positions at HP Labs and the Johns Hopkins Center for Talented Youth. 

His research interests include but are not limited to security (side-channels, smartphones, cloud), distributed systems (data reconciliation, blockchain), and information theory (rateless codes, feedback).  He is an associate editor for the ACM Transactions on Privacy and Security, and recently completed serving as TPC co-chair for the 2024 IEEE International Conference on Blockchain and Cryptocurrency.