Electroluminescence inspection method and AI analysis of PV modules in daylight

Energie & Umwelt

The PVELspect@dAI project addresses the challenge of detecting defects in photovoltaic modules (PV modules) at an early stage, before they cause visible damage or impair performance. Traditional inspection methods are often labour-intensive as they must be carried out in the dark; this project, however, focuses on inspection in daylight. The project aims to develop an innovative and cost-effective measurement and testing method that extends the service life and efficiency of PV modules, enables higher energy production and contributes to the reduction of CO₂ emissions. As part of the project, a system is being developed that combines electroluminescence, photoluminescence, thermography and visible imaging spectroscopy (VIS) in a hyperspectral analysis. This system enables precise inspection even in daylight conditions, simplifies on-site inspections and increases efficiency. Artificial intelligence is used to analyse the collected data in order to identify patterns and anomalies and provide accurate diagnoses. The methodological approach involves the integration of various measurement technologies, supplemented by AI-based analysis, to ensure an accurate and efficient assessment of the modules. This reduces the need for manual checks, minimises human error and increases the reliability of the inspections.

Acronym

PVELspect@dAI

Project running time

01/02/2025 - 31/01/2028

Projectbudget in EUR

540,000

PVELspect@dAI is developing an innovative and cost-effective measurement and testing method for assessing the cell quality of photovoltaic modules, which is being tested in a real-world field environment. The system combines electroluminescence, photoluminescence, VIS and thermography in a comprehensive hyperspectral analysis. Electroluminescence inspection enables defects such as microcracks, hotspots and degradation to be detected at an early stage, before they cause visible damage or impair performance.

Unlike conventional methods, which require inspection in dark environments, PVELspect@dAI enables electroluminescence measurements to be carried out in daylight, which simplifies and speeds up on-site operations. In addition, photoluminescence is used to examine the optical properties of the PV cells, thereby revealing further defects or irregularities. VIS (Visible Imaging Spectroscopy) complements this analysis by providing detailed image data which, when combined with electroluminescence and photoluminescence analysis, enables more precise diagnoses.

Thermography is used to monitor temperature distributions across the modules and to identify hotspots or overheating, which may indicate electrical or mechanical problems. By combining these various measurement methods in a hyperspectral analysis, comprehensive data is collected and analysed by AI. Artificial intelligence recognises patterns and anomalies that indicate defects and provides precise diagnoses.

This integrated approach enhances the accuracy and efficiency of PV module inspection and analysis, reduces the need for manual checks and minimises human error. The timely detection and rectification of defects extends the service life of PV modules and optimises their efficiency. Improved performance and longevity of the modules lead to higher energy production and greater reductions in CO₂ emissions. PVELspect@dAI thus contributes to promoting decarbonisation and reducing CO₂ emissions by offering a comprehensive solution for optimising the performance and reliability of PV modules.