SerpApi.Org
Chinese scientists have made a bold move at the World Artificial Intelligence Conference (WAIC) in Shanghai, unveiling a roadmap that integrates artificial intelligence (AI) with the full life cycle of advanced nuclear energy systems. Dubbed AI for ADANES, the aim of the roadmap is to deploy AI in the field with zero tolerance for safety failures.
As countries look to balance growth and development with carbon emissions, nuclear energy is poised to become the backbone of such a new world. Unlike renewable sources, nuclear reactors can generate energy on demand and can be connected directly to the grid. This is why countries like China are adopting them at scales like never before, since they solve the problem of energy demand without increasing carbon emissions.
While nuclear energy has its multiple advantages, it is also fraught with a fair amount of risk. Disasters like Chernobyl and Fukushima have reminded us time and again, that the risk of an accident remains with this technology, and we need to prepare for the worst scenarios. A technology like AI is well suited for this role as it can process large number of signals coming in from an operational reactor and shut it down in the earliest stages of a mishap.
China’s push for AI
For AI to be deployed in such a role, we first need to overcome the opacity of how the systems work. The Large Language Model (LLM)-based AI systems that are commonplace these days work much like a black box and are not compatible with the strict safety standards of the nuclear industry.
To improve our understanding of how AI can play a role, researchers at the Chinese Academy of Sciences (CAS) have developed ADANES – Accelerator-Driven Advanced Nuclear Energy System, which integrates nuclear fuel breeding, spent fuel transmutation and power generation into a single platform.
The system is designed to operate in subcritical mode and fundamentally changes the safety logic that governs conventional nuclear reactors. Since this also introduces new challenges for the long-term stable operation of the facility and system coupling issues, China is constructing CiADS, a national-scale science and technology infrastructure that will serve as an engineering verification platform for ADANES.
AI for ADANES
At the WAIC in Shanghai, researchers from the Institute of Modern Physics at CAS unveiled AI for ADANES roadmap that combines data, physics-based models and expert knowledge into a “triple-driving” paradigm.
The AI architecture consists of five layers with a unified data infrastructure, physics-native world models, physical-system control, intelligent-agent coordination and continuous evolution designed to embed AI in the full life cycle of the system, from design and commissioning to operation and maintenance.
Experts like Wang Shoujun, the president of the Chinese Nuclear Society, argue that digital and intelligent transformation of the nuclear industry is inevitable. To further the cause, the WAIC forum also launched the AI for ADANES Alliance that brings together research institutes, nuclear enterprises, financial institutions, and AI companies to bridge the gap between scientific validation and industrial deployment.
“AI will play a core role throughout the full life cycle of nuclear energy by improving quality, efficiency and safety,” said Wang in a report by China Daily.
