arXiv paper: Bridging Artificial Intelligence and Power Systems Education Using a Hands-On Executable Framework

A new arXiv paper presents an open, executable framework for teaching AI in power-systems engineering. The authors argue that existing work often focuses on specialized applications and offers limited reusable material for newcomers, while engineering-grounded AI should incorporate established power-system rules rather than operate as a task-agnostic black box. The framework provides progressively difficult Jupyter notebook modules, runnable locally or in Google Colab, covering neural-network function approximation and load-curve fitting, a convolutional-neural-network power-flow surrogate for a five-bus system, DNN-assisted optimization, reinforcement learning for battery-storage control, and physics-informed neural networks for the swing equation. The paper says a community survey found that 92% of respondents faced at least one barrier before running an AI model and 94% wanted a power-specific hands-on course. The materials are being delivered through an IEEE online course and IEEE Power & Energy Society webinars; one webinar had more than 590 live attendees, while the repository recorded more than 344 visits in two weeks.