join Our Workshop at sofe 2025
Location: SOFE, MIT Student Center, Room - 20 Chimneys
Date: Wednesday, 20th June, 2025
Time: 12:00-14:00
Spaces are limited - don't miss this interactive, hands-on session.
Machine learning doesn’t just solve moonshot problems — it can also help with the practical challenges fusion scientists and engineers face every day.
In this workshop, our digiLab Fusion experts will show you how to apply intelligent, uncertainty-aware ML methods to reduce your simulation workload while improving the quality and relevance of your results. Whether you're running design studies, sensitivity analyses, or integrated models, this session will show you how to get up to 10× faster convergence without sacrificing rigour.
1. How machine learning can help you choose simulation inputs more intelligently.
2. How to build ML models you can trust using uncertainty quantification.
3. How our low-code platform, the Uncertainty Engine, makes these tools accessible to everyone.
1. Fewer simulations needed to answer complex technical questions.
2. More converged simulations more often
3. Greater relevance and insight in your results.
4. No prior ML experience required - just bring your domain expertise.
This session is ideal for fusion professionals and engineers looking to maximise their simulation efforts:
- Plasma Physicists
- Integrated Modellers
- Neutronics Engineers
- Thermal/Structural Engineers
- Computational Scientists
- Nuclear Simulation Engineers
- Model-Based Systems Engineers
- Anyone exploring ML in high-stakes physics and engineering contexts
Cyd Cowley
digiLab Solution Engineer (Fusion)
Dan Greenhouse
digiLab Solution Engineer (Fusion)