Stefano Riva receives two prestigious international awards for his research work

Stefano Riva, PhD candidate at the Department of Energy of the Politecnico di Milano, has been awarded the Young Professional Award at the international conference NURETH-21(21st International Topical Meeting on Nuclear Reactor Thermal Hydraulics), held from 31 August to 5 September 2025 in Busan, Korea. The award was granted for the paper “Verification and Validation of Shallow Recurrent Decoders for State Estimation in the DYNASTY facility”, co-authored by S. Riva, Andrea Missaglia, Carolina Introini, J. Nathan Kutz and Antonio Cammi.

The paper presents an innovative application of Shallow Recurrent Decoder (SHRED) networks, a novel paradigm for state estimation in complex systems that combines experimental observations with high-fidelity model data. The approach was validated for the first time on a real facility, DYNASTY, located at the Department of Energy and dedicated to the study of natural circulation in internally heated fluids, with a particular focus on applications for liquid-fuel Generation IV nuclear reactors.

Earlier in May 2025, Riva had also received the ENEN PhD Prize, awarded annually by the European Nuclear Education Network, for his doctoral thesis entitled “Advanced Data-Driven Techniques for State Estimation in Nuclear Reactors”, carried out under the supervision of Prof. Antonio Cammi, with the co-supervision of Dr. Carolina Introini and Prof. J. Nathan Kutz.

 

These recognitions highlight the international impact of the research carried out at the Department of Energy in the field of data-driven techniques and their innovative applications to the nuclear sector.

Shallow Recurrent Decoder (SHRED) networks

are an innovative artificial intelligence method that enables the estimation of the state of a complex system (such as an energy facility) from a limited number of measurements. In practice, with just a few local observations it is possible to reconstruct the entire system state over time, including quantities that cannot be directly measured.

DYNASTY

In the award-winning work, this technology was tested for the first time on a real facility: DYNASTY, at the Department of Energy of the Politecnico di Milano DYNASTY is an experimental facility reproducing natural circulation of internally heated fluids, a phenomenon underlying the operation of Generation IV nuclear reactors, particularly liquid-fuel reactors. Thanks to this study, SHRED networks proved their effectiveness not only with computer-simulated data but also under experimental conditions, opening new perspectives for the safe and accurate monitoring of future nuclear reactors.