In the research project QuMAL-KI - Quantum Accelerated Multi-Agent Learning for Long-Time Autonomous Robots - existing quantum algorithms will be evaluated to accelerate Deep Reinforcement Learning methods for multiple agents, and new quantum algorithms will be developed. These will then be integrated with classical deep reinforcement learning methods into a hybrid framework for quantum/classical multi-agent deep reinforcement learning. The algorithm development and integration here is explicitly exploratory and to be located in the area of basic research (TRL-1). In order to evaluate and demonstrate the practicability, but also the limitations of the developed methods, a test in a simple, but realistic scenario with at least two robots is planned in the project (TRL-2 - TRL-3).
|Duration:||01.12.2022 till 30.11.2026|
|Donee:||German Research Center for Artificial Intelligence GmbH & University of Bremen|
|Sponsor:||Federal Ministry for Economic Affairs and Climate Action|
|Application Field:||Quantum Computing|
Qubit-based data representations for classical machine learning and simulations (10.2022- 09.2025)QuBER-KI
Quantum Deep Reinforcement Learning for simple robotic behaviours (11.2022- 10.2025)QINROS
Quanten computing and quantum machine learning for intelligent and robotic systems (09.2020- 02.2022)Q3-UP!
Needs-oriented and low-threshold qualification modules for quantum computing and quantum machine learning (03.2022- 02.2025)