Article

Neural Networks for Reconstructing Fissile Material Production

Fabian Unruh at the Annual Meeting of the German Physical Society

Fabian Unruh gave a talk at the a annual spring meeting of the German Physical Society (DPG) in Bonn. In his talk titled “Applying neural networks in nuclear archaeology with nuclear reprocessing waste,” presented to the working group Disarmament, Fabian Unruh analyzed recent research results to physicists from German universities. In nuclear archaeology, researchers use nuclear reprocessing waste a by-product of the separation of uranium and plutonium from spent nuclear fuel to obtain information about the amount of plutonium produced in a nuclear reactor. Fabian Unruh presented this approach and showed how to conduct the analysis with neural networks. He confirmed the validity of the neural network's results and provided insights into further optimizations. Finally, he presented some initial results for more complex scenarios to the audience.

Apart from the talk, Fabian Unruh participated in the Member's Assembly of the working group Disarmament.

The spring meeting is an annual conference that the German Physical Society, the largest physical society worldwide with over 50,000 members, hosts. The conference serves as a place for scientific exchange and the presentation of recent scientific findings in physics.

Project partners