Workshop at ACM e-Energy 2026

The growing integration of renewable energy, distributed energy resources, and cyber-physical infrastructure has created an urgent need for advanced modeling, optimization, and control tools for sustainable energy systems. At the same time, the rapid expansion of high-resolution operational data (e.g., from smart meters, PMUs, and building sensors) and recent progress in machine learning and AI are transforming how complex engineering systems are analyzed and operated. However, standard ML methods often struggle to incorporate physical constraints, ensure safe operation, and generalize reliably in safety-critical energy systems. To bridge this gap, this workshop focuses on physics-informed learning for optimization and control of sustainable energy systems.

Physics-informed learning incorporates physical laws, engineering constraints, and domain knowledge into the training and design of machine learning models. By coupling data-driven flexibility with physically grounded structure, these approaches enable improved sample efficiency, interpretability, and reliability, properties that are essential for safety-critical energy systems such as power grids, buildings, and industrial processes. Recent advances demonstrate the potential of physics-informed methods to accelerate optimization, enhance predictive modeling, support real-time control, and enable end-to-end learning frameworks that respect engineering constraints. This workshop aims to bring together researchers in this space to highlight recent work on principled approaches that integrate physical constraints, domain knowledge, and structural information into data-driven models for sustainable energy applications.

This is the first edition of this workshop at ACM e-Energy.

Workshop Organizers:

• Yuanyuan Shi, University of California San Diego, yyshi@ucsd.edu
• Wenqi Cui, New York University, wenqicui@nyu.edu
• Nicolas Christianson, Johns Hopkins University, christianson@jhu.edu