Physicists with the University of Sussex have successfully developed a remote monitoring and control system blueprint for quantum devices and experiments. The new research pairs human intervention, data collection, and AI systems to detect and correct environmental fluctuations in vulnerable quantum computers (and their processing units, the qubits). The team hopes their work will eventually allow for the development of quantum computing solutions that operate across vast swathes of distance and in otherwise unworkable environments, such as space, or beside your desktop PC from the future.
The quantum computing environment is booming, with multiple research breakthroughs occurring on a seemingly daily basis. Looking at this trend, the researchers believe quantum computing is reaching a level of maturity that could allow for adoption by non-specialist users — or that could extend its deployments towards otherwise non-ideal but attractive scenarios, such as space. And it was derived from the scientists’ experience at setting up work-from-home solutions so that research within the University’s ultracold quantum laboratories never stopped.
“Whilst we have been able to exploit this system for controlling our experiments for homeworking, even more significantly, this technology provides a blueprint for monitoring the environment and carrying out research and operating technology in inaccessible, non-constant and unpredictable environments such as space, underground or below sea level,” said Dr. Thomas Barrett, Research Fellow in the group and lead author of the paper. “That’s where it gets really exciting and could provide far reaching impact.”
The research provides control and key experimental parameters, such as vacuum chamber pressure, laser beam power, or resistances of important conductors; all of which are required hardware in quantum computing environments. It also showcases how otherwise detrimental environmental factors can be mitigated by deploying automated control sequences paired with evolutionary algorithms and machine learning protocols. These solutions constantly m
