The 6th Conference on Controlled Nuclear Fusion and Artificial Intelligence Technology (AI4Fusion 2027) will be held in Wuhan, Hubei, China, on April 20–23, 2027. It is supervised by the Chinese Nuclear Society and the Chinese Physical Society, organized by the Nuclear Fusion and Plasma Physics Branch of the Chinese Nuclear Society and the Plasma Physics Branch of the Chinese Physical Society, hosted by Huazhong University of Science and Technology (HUST), and co-organized by China Fusion Energy Co., Ltd. and the Southwestern Institute of Physics.
The conference brings together researchers in fusion energy and artificial intelligence to share progress in applying AI to controlled fusion: device experiments and plasma control, intelligent simulation, diagnostics and data analysis, fusion engineering, and foundation models and AI-driven research. It also aims to strengthen links between academia and industry. Young researchers and students are especially welcome.
Previous Editions
- 1st2021 · Chengdu
- 2nd2023 · Chengdu
- 3rd2024 · Hangzhou
- 4th2025 · Langfang
- 5th2026 · Shanghai
- 6th2027 · Wuhan
Topics
1. AI for fusion device experiments and operation control
- AI-enabled experiment design, operational decision-making and process optimization
- Plasma state awareness, prediction and early warning, and intelligent control of complex processes
- Coordinated optimization and autonomous control of multi-parameter, multi-objective, multi-actuator systems
2. Intelligent simulation for controlled fusion
- Surrogate models, reduced-order models and data-driven simulation of fusion physics
- Modeling methods that combine physics models, experimental data and AI
- AI-assisted multi-scale, multi-physics simulation and high-performance scientific computing
- Digital twins and intelligent simulation platforms for experiments, design and operation
3. AI for plasma diagnostics and data analysis
- AI-enabled diagnostics and parameter inversion for plasmas and fusion processes
- Fusion of multi-source, multi-modal experimental data for integrated analysis and state recognition
- Processing, feature extraction and pattern discovery in fusion big data
- Joint use and knowledge mining of experimental, simulation and cross-device data
4. AI for fusion engineering
- AI-assisted design and optimization of devices, key components and engineering systems
- Condition monitoring, anomaly detection, fault diagnosis and predictive maintenance
- AI in fusion materials, energy conversion, the fuel cycle and other key technologies
- Intelligent manufacturing, robotics, remote maintenance and life-cycle management
5. Foundation models and AI-driven research for fusion
- Domain, foundation and multimodal models for fusion science and engineering
- AI-driven autonomous experiments and simulation, scientific reasoning and knowledge discovery
- Fusion knowledge platforms, digital research infrastructure and human–AI collaborative research
Format
The conference will feature invited plenary talks, oral presentations in topical sessions and poster sessions.
Highlights
These activities are planned; details will follow in later circulars.
Roundtable: AI × Fusion
Fusion device teams, AI researchers and industry leaders on one stage: what kind of AI does fusion research need, and what can AI gain from fusion?
Industry–Academia Meeting: needs, pitches and one-on-one meetings
Fusion and AI companies present technical needs, university and institute teams pitch solutions, and investors and companies give feedback. An exhibition area runs alongside. Company sign-up will open soon.
AI4Fusion Data Challenge
Tasks built on real device data, open to teams from universities, institutes and companies. Tasks and data will be released with the second circular; top teams present in the final at the conference.
Best Talk and Best Poster Awards
Selected by the Academic Committee during the conference and presented at the closing ceremony.