CSIC's researchers open the door to revolutionizing quantum technologies through artificial intelligence and cold atoms

Image

The revolution in quantum technologies hinges on the use of artificial intelligence and machine learning for their development and advancement. This is the core idea behind a new project led by the Spanish National Research Council (CSIC) through the Center for Research in Nanomaterials and Nanotechnology (CINN, CSIC-Uniovi – Principality of Asturias) and involving the Institute of Materials Science of Madrid (ICMM) and the Institute of Nanoscience and Materials of Aragon (INMA, CSIC - Unizar).

This project, titled 'Machine Learning in Quantum Simulations with Rydberg Atoms' and funded by the Ministry of Science, Innovation and Universities (MCIU), "combines quantum computing with artificial intelligence to build an advanced platform where hardware and algorithms work together to solve complex real-world problems more efficiently than classical computers," explains Miguel Pruneda, a CSIC researcher at CINN and the project coordinator. The hardware is based on Rydberg atoms, a type of highly excited atom (meaning it has a higher energy level than others) that emerges as a promising platform for the development of quantum computing.

"We know that this class of Rydberg atoms have unique properties that make them very good candidates for scalable quantum computing, as they combine strong interactions between them with long coherence times," explains Sigmund Kohler, a researcher at ICMM-CSIC and leader of one of the work packages. The challenge with this class of arrays lies in how to control these atoms. "And this is something we can address thanks to artificial intelligence and machine learning," adds Yue Ban, also a researcher at ICMM-CSIC.

The experts detail that this project constitutes "disruptive research" that uses artificial intelligence for the control and interpretation of quantum systems based on Rydberg atoms, which in turn can solve complex optimization problems using quantum algorithms: "This project has the potential to revolutionize both experimental design and the theoretical understanding of complex quantum phenomena," adds Jesús Carrete, a CSIC researcher at INMA (CSIC-Unizar).

Another interesting aspect of the proposal from these three CSIC centers together with the Computer Vision Center (CVC, from Catalonia) is the multidisciplinary approach that combines experimental innovation with computational advancement. In this way, the project is structured as a series of interconnected nodes through which experimental procedures for calibrating the systems for experiments will be optimized while simultaneously improving the fidelity and scalability of computational simulations and designing new quantum algorithms that, in turn, will be specifically optimized for the experimental platform.

Finally, they will also explore the use of these methods in other fields of application, such as quantum chemistry, the distribution of telecommunications antennas, or the stability of the electrical grid. "Overall, the synergistic study between quantum software and hardware will contribute to the creation of robust solutions" that will also be useful in other areas of knowledge.

From the Laboratory to the Computer, and Vice Versa

The project has been designed with a multi-nodal and multi-regional structure that will leverage the expertise of each working team. Thus, coordination is carried out from CINN, where a laboratory already exists that works with Rydberg atoms. This laboratory will serve as the experimental hub of the project, the space where previously designed quantum simulation protocols will be physically tested.

Meanwhile, the work at INMA will focus on using neural networks to estimate the properties of many-body quantum systems. These techniques will provide powerful theoretical tools for understanding complex quantum models formed by spins, and for optimizing the experimental platform. In this way, "within the project, the quantum simulator itself will be simulated."

In Madrid, the ICMM team will focus on the development of quantum algorithms, with special emphasis on quantum optimization methods and novel approaches. "This work will result in the creation of high-level quantum applications, linking theoretical development with a future hardware implementation," Kohler celebrates.

Finally, the CVC in Barcelona will contribute its expertise in both classical machine learning and quantum machine learning techniques to support the other nodes in executing methodological tasks. Specifically, the CVC will provide advice on the implementation of machine learning and, in turn, will structure the research pathways for technological use cases of quantum machine learning, for example in computer vision applications.