At EU PVSEC conference 2025 in Bilbao, researchers from DTU and CIEMAT presented innovative approaches to make photovoltaic (PV) systems more intelligent, efficient, and sustainable. Their work spans AI-based diagnostics, digital twins, tracking optimisation, and lifecycle sustainability assessment, all contributing to Europe’s clean energy transition. 

Influence of irradiance and drone altitude in infrared thermography inspections of photovoltaic plants  

By Rodrigo del Prado Santamaría 

This study analyses how irradiance and drone altitude affect infrared thermography (IRT) for PV plant inspections. The dataset, collected at DTU’s Riso PV plant, includes drone-based IRT images, IV curves, and electroluminescence (EL) references for multiple modules and defect types. The work advances drone-based defect detection methods and supports standardised inspection procedures for solar installations. https://doi.org/10.11583/DTU.29665346.v1 

CartesiaNet: a Lightweight Neural Network for PV Module Corner Detection 

By Thøger Kari, DTU 

CartesiaNet enables fast and precise identification of PV module corners using a custom lightweight neural network. It achieves outstanding accuracy (Weighted F1 > 0.92, mAP > 0.98) and can be tuned within hours for specific PV plants. Tests on unseen RGB images show excellent robustness to varying conditions, making it a promising tool for automated inspection and digital twins of solar assets. 

Can Regular String IV Measurements Complement MPP Monitoring Data? 

By Martin Bartholomäus, DTU 

This study explores how inverter-based IV scans can enhance fault detection and system diagnostics in PV plants. IV traces provide valuable insights at low cost, complementing Mpp monitoring. The research identifies key challenges, such as reliability and downtime and seeks to quantify how many IV measurements are needed to improve O&M strategies. 

Modelling, Implementation and Validation of Solar Tracking Algorithms in Horizontal Single-Axis Trackers 

By Nuria López Peña, DTU 

The work develops and validates an Astronomical Tracking algorithm for horizontal single-axis trackers at DTU’s Riso PV plant. Each tracker, equipped with bifacial modules and custom controllers, was independently tested to assess energy yield and dynamic response. The study provides a robust model for optimising tracking performance and energy consumption. 

Evaluating the Accuracy of Single-Camera Irradiance Forecasting 

By Jacob Krum Thorning, DTU 

This research assesses the accuracy of all-sky imaging for irradiance forecasting. Using a calibrated single camera at the DTU Riso testbed, results show strong correlation between projected and measured global horizontal irradiance (GHI), with an nRMSD of around 1.5%. The findings support the use of low-cost visual sensors for real-time solar forecasting. 

Sustainability Strategies for Optimising the Use Phase of Photovoltaics: an Overview from a Life Cycle Perspective 

By Daniel Garrain, CIEMAT 

This study provides a comprehensive overview of the sustainability assessment of PV systems during the O&M stage, based on a literature review. The main proven strategies and findings from recent studies based on life cycle assessment (LCA) and life cycle costing (LCC) were identified. The results guide strategies for extending system lifetimes and enhancing circularity within the solar sector. 

Together, these works showcase how digital innovation, AI, and sustainability analysis can transform the way PV systems are designed, monitored, and maintained, strengthening Europe’s leadership in renewable energy technologies.