Magnetics play important roles in almost all power electronics systems. The multidimensional and interconnected complexities of material physics, geometric structure, operating conditions, and circuit functionality make magnetic components a major bottleneck in high-performance design. Data-driven methods and artificial intelligence (AI) techniques have demonstrated strong capabilities in modeling complex physical systems. Large-scale, high-quality, well-labeled data sources are the foundations of modern AI. A collaborative environment enables the community to break traditional knowledge barriers and unlock forward-looking breakthroughs. This presentation will explore the historical evolution and emerging intersections between power magnetics and artificial intelligence, and introduce the MagNet Challenge as an international collaborative research platform that bridges materials and design, data and algorithms, academia and industry, and the past and future of power electronics.
