skip to main content

CIRI physicists harness quantum computing to safeguard critical infrastructure

4/18/2019

University of Tennessee Physics Professor George Siopsis and University of Calgary Physics Professor Barry Sanders, both researchers in the Critical Infrastructure Resilience Institute, are working at the forefront of their field to tackle novel solutions for critical infrastructure security with machine learning and quantum computing. 

Through their newly-funded CIRI project, “Hybrid Quantum-Classical Reinforcement Learning in Controlled Quantum Networks,” the team aims to develop a reinforcement learning framework with wide applicability for the Dept. of Homeland Security, which funds much of CIRI’s research. Siopsis describes reinforcement learning as “a flavor of machine learning in which the machine learns while taking actions and thereby gaining experience.”  

Quantum computing is heralded as a technology that could transform cybersecurity, though many advancements in physics and computing will be necessary to achieve this vision. One such advancement involves quantum machine learning, which leverages hybrid classical-quantum machine algorithms to train neural networks to recognize patterns. For example, current technology may struggle to differentiate between a bird and a drone; however, heat or acoustic signatures could be utilized to discriminate between objects in the sky. Visual detection (via a camera) would be paired with machine learning (algorithms that recognize patterns) in order to correctly identify drones. 

Siopsis and Sanders’ team sees wide applicability for such pattern recognition, including the aforementioned drone detection, malware detection, power grid resiliency, data mining and analysis for first responders, and crisis management for the Federal Emergency Management Agency (FEMA). For example, in terms of crisis management, algorithms could be developed for the optimization of allocation of limited resources in crisis situations, such as in the aftermath of a hurricane.   

Quantum computers at present have limited capabilities, according to Siopsis. However, “They are expected to develop quickly and we should position ourselves in a way to take advantage of their power,” he says. 

At present, multiple companies are competing to produce the best quantum computer architecture, including D-Wave Systems, IBM, and Microsoft. The team intends on developing algorithms with various functions (for example, drone detection) that can be run on these quantum computers of the future.  Siopsis explains, “DHS has already begun working on and applying machine learning. What we are hoping to do is to add the quantum ingredient.” 

The project of Siopsis and Sanders, who is also the Director of the Institute for Quantum Science and Technology at University of Calgary, sprang from CIRI's International Workshop on Artificial Intelligence and Quantum Information Applications in Homeland Security, which brought together leading researchers in the field from the US and Canada.