Researchers at the Ames National Laboratory are developing a new approach to designing high-performance permanent magnets that does not rely on rare-earth elements.
The effort combines artificial intelligence, advanced simulations, and physics-based models to identify promising materials faster. The work supports broader US goals to strengthen critical material supply chains and reduce dependence on foreign sources.
Permanent magnets are essential components in many modern technologies. They are used in electric vehicles, wind turbines, electronics, industrial equipment, and defense systems. Many of today’s strongest magnets depend on rare-earth elements because these materials offer high magnetic strength and resistance to demagnetization.
However, rare earth materials can be expensive and are often sourced from regions outside the US. This creates concerns about supply chain stability and long-term availability. Finding alternative materials has therefore become an important scientific and economic priority.
For more than two decades, scientists have searched for rare-earth-free permanent magnets. Traditional research methods often involve producing materials in the laboratory, testing their properties, and gradually building knowledge through repeated experiments. While effective, this process can be slow, costly, and resource-intensive.
READ ALSO: NASA Trials Innovative Supersonic Parachute Delivery Technology
Ames Lab scientist Prashant Singh recently outlined a different strategy in a published article. His roadmap combines physics-informed artificial intelligence, high-throughput computer simulations, and advanced computational tools. The goal is to predict promising materials before researchers spend time and money creating them in the laboratory.
The approach begins by studying how atoms are arranged inside a material and how electrons behave. These factors directly influence key magnetic properties, including magnetic strength, energy storage capacity, resistance to demagnetization, and performance at higher temperatures. Understanding these relationships helps researchers identify materials with the most desirable characteristics.
By embedding this scientific knowledge into AI systems, researchers can move beyond simple trial-and-error methods. The models can examine a much larger range of potential materials in a shorter period. This allows scientists to focus laboratory testing on the most promising candidates.
Ames Reinvents Future Magnets
According to Singh, Ames Lab holds a unique advantage in this area. The laboratory has spent decades building expertise in magnetic materials and collecting valuable research data. This combination of knowledge and data provides a strong foundation for training advanced AI systems.
Researchers are also developing new ways for scientists to interact directly with these computational models. Instead of manually evaluating large datasets, scientists can ask design-focused questions and refine material requirements through AI-assisted tools. This can speed up and improve the exploration process.
READ ALSO: US Navy’s latest 6th gen jet could have 25% higher range
One example is DuctGPT, an agentic AI system previously developed by Singh. The tool was designed to support the exploration and design of interactive materials. Similar technologies are now being adapted to help identify future magnet materials.
A major challenge for AI-based materials research is ensuring that models learn from reliable scientific information. General-purpose data alone is often not enough for accurate materials design. Researchers need experimental measurements and physics-based calculations that reflect real-world material behavior.
Singh emphasized that physics remains central to the discovery process. AI systems trained only on existing data tend to make predictions within known boundaries. When physical principles are included, researchers can explore a much wider range of materials and identify options that have never been tested before.
Another important feature of the new approach is its ability to account for economic and supply chain factors. Material prices, availability, and market conditions can change rapidly. Including these variables in the design process helps researchers focus on materials that are both practical and high-performing.
This broader perspective connects scientific discovery with manufacturing realities. A material with excellent performance may not be useful if it is too expensive or difficult to obtain. AI tools can help balance performance, cost, and availability from the earliest stages of development.
The research is also connected to the US Department of Energy’s Genesis Mission. The initiative brings together national laboratories, universities, and industry partners to apply artificial intelligence to major challenges in energy, science, and national security. Securing access to critical minerals and materials is one of its key objectives.
WATCH ALSO: Chinese humanoid robot learns to stitch detailed embroidery patterns autonomously
Ames National Laboratory has played a significant role in critical materials research for decades. Its long history in magnet science, theoretical modeling, and simulation provides researchers with specialized knowledge and proprietary datasets. These resources strengthen efforts to accelerate materials innovation using AI.
The combination of physics, simulation, and artificial intelligence represents a shift in how advanced materials are discovered. Instead of relying mainly on laboratory experimentation, scientists can now use computational tools to guide decisions more effectively. This reduces development time and improves research efficiency.
As demand for advanced magnets continues to grow, the need for alternative materials is essential. Industries ranging from clean energy to defense depend on reliable supplies of high-performance magnetic materials. Faster discovery methods can help address these growing requirements.
Researchers believe integrating AI with decades of scientific expertise will expand the search for new magnetic materials and strengthen domestic supply chains. By accelerating discovery while accounting for manufacturing realities, Ames National Laboratory is helping lay the foundation for the next generation of permanent magnet technologies.
