Human-Centric AI Webinar #4: AI for Sustainability: From Solar Cells to River Systems

This webinar features two research talks on using interpretable machine learning to understand real-world physical systems. The first talk looks at solar cells: predicting their internal electrical behavior is normally done with complex physics-based simulations that are hard to invert to recover the underlying material properties. This talk shows how training ML models on simulated data can instead directly extract these physical properties from real measurements, while also revealing which experimental conditions are most informative for which parameters. The second talk turns to river systems, comparing modern deep learning models against a traditional hydrological model for predicting streamflow (useful for flood and drought monitoring), and shows how explainable AI techniques can make these predictions more transparent and trustworthy for water management decisions. Together, the talks show how interpretable ML can turn accurate predictions into real physical understanding, making models more useful for real-world decisions.

Participation is free!

Date: 15 July 2026, 14:00-15:00 CEST
Location: Online (Teams)
To register, you must fill in the following form by 14 July 2026 (EOB CEST).

Target audience

AI researchers, practitioners, and students interested in explainable and interpretable machine learning, as well as professionals in photovoltaics, hydrology, or other physical/environmental sciences applying ML to complex systems.

Speakers