Next Step Fusion

⚡ Tokamak simulations can now be done online with our Fusion Twin Platform For months, we have been developing an innovative DeepTech solution, the Fusion Twin Platform. Opening it up to all fusion researchers is a big step for our company and an exciting opportunity to engage the fusion community in exploring its potential. Our cloud-based platform aims to accelerate scientific insights by streamlining the preparation, execution, analysis, and sharing of experiments conducted on tokamaks, magnetic confinement fusion devices that are key candidates for evolving into fusion reactors, a source of clean and abundant energy. As the Fusion Twin Platform is free for all fusion researchers, now everyone can run precise tokamak simulations using NSFsim, our simulator, and digital replicas of famous and less famous tokamaks. Register at https://fusiontwin.io to: 🟢 Run highly customizable magnetic equilibrium simulations with the digital replicas of DIII-D, ISTTOK, NSF NTT, and other tokamaks. We keep adding existing and new yet-to-be-built devices to the platform and will soon launch the inverse problem simulation and discharge scenario builder. 🟢 Upload your experimental data or download simulation results as HDF5 files. We do not share your data with anybody else and let you download it freely. 🟢 Easily map your uploaded data to the platform’s variables for visualization. The embedded Graphs tool includes 100+ pre-defined graphs for all major plasma parameters and diagnostic signals and allows flexible simultaneous work with multiple graphs. 🟢 Analyze and transform data within JupyterHub, using Python Notebooks and integrated extensions. 🟢 Try out Plasma Boundary Prediction using an ML model trained on DIII-D tokamak data. 🟢 Enjoy a web-based solution that does not require any additional software or hardware. 🟢 Improve your work with upcoming collaborative features, more digital replicas of tokamaks, significantly upgraded NSFsim, an API for simulations, an embedded LaTeX editor, a discharge database, and more. We hope that the Fusion Twin Platform will streamline fusion experiments, data management, and collaboration. We also believe it has great potential as an educational tool to help prepare the next generation of a diverse workforce for the rapidly growing fusion industry. The platform can close the gap between theoretical education and practical work on real tokamaks by providing students with a digital environment to study and practice fusion simulations, scenario building, fusion data analysis, and similar skills. Check it out at https://fusiontwin.io and ask any questions in the comments. #FusionTwinPlatform #DigitalTwins #FusionEnergy #EnergyResearch

Dmitri Orlov thank you for your always-collaborative spirit and hard work. We are happy to work together and proud of the results. Happy 2025!

Hey #Fusion! Many of us have seen a #LEGO model of #ITER, but AFAIK, it was never available for purchase. Let's support MAST Upgrade #tokamak on the LEGO Ideas website: https://lnkd.in/dXwVhi8N! the LEGO Group, #FusionEnergy is the next frontier for humanity. There must be a LEGO set for us!

🎄🎄🎄 𝐃𝐨 𝐧𝐨𝐭 𝐦𝐢𝐬𝐬 𝐭𝐡𝐞 𝐅𝐮𝐬𝐢𝐨𝐧 𝐂𝐡𝐫𝐢𝐬𝐭𝐦𝐚𝐬 𝐓𝐫𝐞𝐞! What an incredible year it was for us and for fusion! In 2024, we pushed the boundaries toward robust and reliable fusion power plant control, demonstrating the power of our technologies and launching new products: ✅ 𝐌𝐋-𝐛𝐚𝐬𝐞𝐝 𝐏𝐥𝐚𝐬𝐦𝐚 𝐂𝐨𝐧𝐭𝐫𝐨𝐥 𝐒𝐮𝐜𝐜𝐞𝐬𝐬: Our shape and position control ML model, trained using reinforcement learning, successfully controlled plasma in L-mode, H-mode, and H-L transition for over 3 seconds on the DIII-D tokamak. A paper is coming soon, and since the module is now part of the DIII-D Plasma Control System, we will soon start supporting DIII-D users with training models to help them achieve and control their desired plasma shape and performance characteristics. ✅ 𝐂𝐨𝐦𝐩𝐥𝐞𝐭𝐞𝐝 𝐍𝐞𝐠𝐚𝐭𝐢𝐯𝐞 𝐓𝐫𝐢𝐚𝐧𝐠𝐮𝐥𝐚𝐫𝐢𝐭𝐲 𝐓𝐨𝐤𝐚𝐦𝐚𝐤 𝐃𝐞𝐬𝐢𝐠𝐧: Work on our NSF NTT—a negative triangularity private research tokamak—progressed rapidly, culminating in the completion of the full preliminary design by the end of the year. We’ve shared many updates through our blog and LinkedIn posts and plan to continue developing tokamak design and optimization tools in 2025. ✅ 𝐎𝐩𝐞𝐧𝐞𝐝 𝐔𝐩 𝐅𝐮𝐬𝐢𝐨𝐧 𝐓𝐰𝐢𝐧 𝐏𝐥𝐚𝐭𝐟𝐨𝐫𝐦 𝐟𝐨𝐫 𝐄𝐯𝐞𝐫𝐲𝐨𝐧𝐞: We made precise simulations with NSFsim, our advanced simulator, more accessible, enabling faster and easier work with fusion data for researchers worldwide. Digital replicas of DIII-D, ISTTOK, and NSF NTT are already available at 𝐟𝐮𝐬𝐢𝐨𝐧𝐭𝐰𝐢𝐧.𝐢𝐨, and 2025 will bring many more! ✅ 𝐋𝐚𝐢𝐝 𝐭𝐡𝐞 𝐅𝐨𝐮𝐧𝐝𝐚𝐭𝐢𝐨𝐧 𝐟𝐨𝐫 𝐭𝐡𝐞 𝐌𝐨𝐝𝐞𝐫𝐧 𝐏𝐥𝐚𝐬𝐦𝐚 𝐂𝐨𝐧𝐭𝐫𝐨𝐥 𝐒𝐲𝐬𝐭𝐞𝐦: Recognizing the fusion industry’s need for modern, fusion power plant-relevant, high-quality plasma control software, we initiated the development of a state-oriented Plasma Control System. This system will incorporate all our technologies and deliver a reliable solution for the fusion energy industry. ✅ 𝐃𝐞𝐦𝐨𝐧𝐬𝐭𝐫𝐚𝐭𝐞𝐝 𝐄𝐱𝐜𝐢𝐭𝐢𝐧𝐠 𝐑𝐞𝐬𝐮𝐥𝐭𝐬 𝐰𝐢𝐭𝐡 𝐒𝐮𝐫𝐫𝐨𝐠𝐚𝐭𝐞 𝐌𝐋 𝐌𝐨𝐝𝐞𝐥𝐬: Using surrogate models trained on DIII-D experimental data, we demonstrated the promises and limitations of this approach based on historical datasets. This work has opened a new chapter in our R&D, focusing on building highly reliable surrogate models for both existing and future tokamaks. Stay tuned for a publication! 👥 We were thrilled to work with DIII-D National Fusion Facility, UC San Diego, General Atomics, Columbia Engineering, Instituto Superior Técnico, General Fusion, visit tokamaks in person, present at the 50th EPS Conference, the 66th APS DPP, and other conferences, and meet so many of you along the way! ❤️ Special thanks to Dmitri Orlov, Randall Clark, Carlos Paz-Soldan, Chris Hansen, Ph.D., Oak Nelson, Himank Anand, David C. Pace, PhD, MBA, Richard Buttery, Celso Ribeiro, Horácio Fernandes, Pavel Aleynikov. Enjoy the holiday season, and see you soon in 2025!

🔬 Applying a Surrogate Model to an RL-Controlled Pulse After conducting an experiment on the DIII-D tokamak with our Reinforcement Learning (RL) model for plasma control, and training a surrogate model to reconstruct plasma shape, we became curious how these models work together. 🔗 See the links to our blog posts in the comments below. So, we applied the surrogate model to reconstruct the shape of the plasma controlled using our RL controller. The surrogate model demonstrated promising performance in reconstructing the plasma boundary, achieving a mean boundary point displacement of 0.07 meters during the discharge. Interestingly, this is slightly lower in quality compared to our model’s reconstructions for DIII-D experiments with conventional control algorithms, where the model achieves a mean error of 0.05 meters. ‼️ This observation reminds us the importance of providing representative and diverse training datasets for ML models, as their generalization power is closely tied to the quality and coverage of the data used during training. Overall, we see the great potential in integrating supervised ML and RL techniques in fusion research to advance plasma control and optimization. See the chart with the comparison of the ground-truth boundary (blue) and the reconstructed boundary (red), illustrating the capabilities of our model. #Fusion #FusionEnergy #PlasmaPhysics #Tokamak #MachineLearning

Last week, Maxim Nurgaliev from Next Step Fusion delivered an excellent presentation during the Plasma Control Group meeting on the recent experiments conducted by our UC San Diego and Next Step Fusion teams on the DIII-D National Fusion Facility. Even during challenging piggyback experiments, where the discharges were not ideal and unexpected H-to-L mode transitions occurred, the Reinforcement Learning module integrated into the DIII-D Plasma Control System performed remarkably well, exceeding our expectations in holding plasma shape. These achievements demonstrate the robustness of the controller, and we hope that more researchers in the DIII-D program will adopt this tool for their experiments. Next Step Fusion will be presenting these results at the DIII-D Science, Engineering, and Technology meeting in January. This collaboration exemplifies the kind of public-private partnerships envisioned and supported by the U.S. Department of Energy (DOE) Office of Fusion Energy Sciences, fostering innovation and progress. With tools like these, we are advancing the Technology Readiness Level (TRL) of plasma control systems for future Fusion Pilot Plants (FPP). #FusionEnergy #PlasmaControl #ReinforcementLearning #NextStepFusion #DIIID #PublicPrivatePartnership #FusionPilotPlant #TechnologyReadinessLevel #DOE #PlasmaPhysics Richard Buttery Scott Hsu Matthew Lanctot Michael Halfmoon Aleksei Zolotarev Jean Paul Allain

🔊 Another blog post about our journey in designing a negative triangularity tokamak. We like sharing about our work, so if you're interested to know more or meet us, just let us know! #Fusion #FusionEnergy #Tokamak #PlasmaPhysics #NegativeTriangularity

I’m thrilled to share that our manuscript, “Validation of NSFsim as a Grad-Shafranov Equilibrium Solver at DIII-D”, has been accepted for publication in Fusion Engineering and Design! NSFsim is an incredible code, and the results in this paper prove it. By successfully validating its capabilities to model plasma shapes, flux distributions, and diagnostic signals under various conditions, we’ve demonstrated its potential as a transformative tool in fusion research. We envision that NSFsim will be adopted across many different tokamaks in the future, setting the standard for plasma control on Fusion Pilot Plants (FPP). This is just the beginning of its journey toward becoming a cornerstone of fusion technology. A huge thanks to our colleagues at Next Step Fusion, who did a tremendous job developing this code. Your expertise and dedication have been key to this achievement. This collaboration exemplifies the power of public-private partnerships in driving fusion forward and making this clean energy dream a reality. The preprint is available on arXiv at https://lnkd.in/gXPb5atC. Dmitri Orlov, Georgy Subbotin, Maxim Nurgaliev, Eduard Khairutdinov #FusionEnergy #PlasmaPhysics #NSFsim #PlasmaControl #PCS #DIIID

🎯 First publication of our result with UC San Diego: We've validated NSFsim against DIII-D tokamak data! The results demonstrate our simulator's capability to accurately model fusion plasma behavior. Thank you Randall Clark #FusionEnergy #PlasmaPhysics #DIIID #SimulationScience Maxim Nurgaliev Eduard Khairutdinov Georgy Subbotin Dmitri Orlov Anders Welander

A great model of public private partnership! Public sector working with private sector to share the fruits of our advances, jointly made, for all, and befitting each is our programs. Well done to Next Step Fusion and their pioneering work at DIII-D National Fusion Facility ! Scott Hsu