Imagine a flight simulator, but for nuclear power plants during a worst-case disaster. Currently, the software that predicts these disasters is too slow to run in real-time. This project uses AI to speed up those calculations so operators can practice responding to emergencies in a virtual environment without any real-world risk.
Nuclear accident codes are too computationally slow for real-time use, making them impractical for operator training and rapid safety analysis. This creates a gap in workforce readiness and emergency response planning.
A proof-of-concept desktop simulator interfacing ASTEC with TEAM_SUITE®, featuring a GUI and AI-driven surrogate models for real-time execution.
If you are a plant operator dealing with the risk of fuel degradation and workforce turnover — this project developed a real-time simulator that allows staff to practice managing severe accidents like station black-outs. This reduces the learning curve for new employees and improves safety guidelines.
If you are a software vendor dealing with slow physics-based calculation codes — this project developed a way to interface the ASTEC code with the TEAM_SUITE® platform using machine learning. This enables the creation of fast-running surrogate models that maintain calculation accuracy.
If you are a training center dealing with the lack of accessible severe accident tools for students — this project developed a desktop simulator with a graphical user interface. It allows non-specialists to visualize in-vessel and ex-vessel accident phases in real-time.
Based on available project data, no pricing or cost information is provided as this is a research project funded by EURATOM.
Yes, the project explores solutions to extend the models developed for the Western-type PWR to other reactor types and severe accident codes.
Based on available project data, the training database for machine-learning sequences will be made openly available, but specific licensing for the ASTEC/TEAM_SUITE® interface is not detailed.
The project specifically demonstrates integration by interfacing the ASTEC calculation code with the commercial TEAM_SUITE® simulation platform.
The project period runs from 2022-11-01 to 2026-10-31, indicating it is currently in the development and proof-of-concept phase.
The consortium is well-balanced for technology transfer, consisting of 15 partners across 11 countries. With a 33% industry ratio (5 industrial partners, including 3 SMEs), there is a strong link between academic research (3 universities, 5 research centers) and commercial application, specifically through the use of commercial platforms like TEAM_SUITE®.
Contact the Autorite de Surete Nucleaire et de Radioprotection (ASN) in France.
Contact us to explore licensing opportunities for the AI surrogate models developed in ASSAS.
