NeMoQC

Accurate predictions are critical for the safety and reliability of critical infrastructure such as transport and power grids, or for forecasting extreme weather events, as well as for safety-critical components such as aircraft wings or rocket engines. Many properties of these complex systems are highly non-linear and chaotic. In the NeMoQC (Neuromorphic Quantum Computing) project, we are working to improve the prediction of these systems with innovative solutions based on quantum reservoir computing (QRC).

Predictions of the behaviour of complex systems feed directly into proactive control of system behaviour to intervene before a critical point is reached. The use of artificial intelligence has made great strides in this area, but at the cost of increasing computing power and energy consumption. Our human brains, on the other hand, can solve typical tasks for AI models much more energy-efficiently and effectively than today's hardware and software. This is where the field of neuromorphic computing comes in. It uses concepts inspired by the way the human brain works to create more powerful and energy-efficient hardware and software. Reservoir computing is one of these approaches, where the AI itself draws on a complex system to more easily map properties in the data and store information from previous time steps. Quantum computers are promising candidates for realising this reservoir and for quickly achieving accurate results, even with very high-dimensional initial data.

Contribution Institute for AI Safety and Security

As part of the NeMoQC project, the Institute for AI Safety and Security is researching new concepts for QRC to improve its performance in prediction and optimisation techniques. In particular, we are analysing the limitations, suitable systems for reservoirs and potential barriers to the use of QRC. The results of this project will not only provide a deeper understanding of the method, but will also give new impetus to the application of innovative computational methods.

During the project, we are investigating the applicability of the methods, particularly in the context of use cases from the Institute of Aeroelasticity, to explore the performance and requirement characteristics of the quantum reservoir. In particular, we are exploring the technical feasibility and performance of the quantum reservoir in these applications. We also address the question of whether a universal quantum reservoir is suitable for a variety of applications, or whether problem- and application-specific optimal solutions are required in each case.