In short
The global emphasis on sustainable energy underscores the importance of maximizing the efficiency and lifespan of photovoltaic installations. Current inspection methods, while effective, have limitations in precision and predictive maintenance capabilities.
The MULTISPECTRAL team aims to bring to market an advanced drone-based luminescence imaging technology tailored for outdoor photovoltaic inspection. This cutting-edge method promises superior performance over existing inspection technologies in the market. Together, they aim to develop comprehensive tools spanning the entire value chain – from robust hardware components to user-centric cloud software solutions. Their collective goal is not only to enhance solar panel inspections quality but also to redefine standards for accuracy and efficiency for operations and maintenance in the photovoltaic sector.
Important research to society
Investing in solar, wind and thermal power, improving energy productivity, and ensuring energy for all is vital if we are to achieve Sustainable Development Goal 7 by 2030. Expanding infrastructure and upgrading technology to provide clean and more efficient energy in all countries will encourage growth and help the environment. The MULTISPECTRAL project will ensure less faults are introduced into photovoltaic modules and are identified earlier, minimizing power loss from the green energy sources and therefore accelerating the transition towards green energy. Minimizing loss will also lead to lower prices in the end.
Currently solar powered modules pay their embodied energy back 20-40 times over their lifetimes, thus contributing to mitigation of climate change. This project has the potential to increase the lifetime of solar power products, with higher longevity of the solar panels leading to a higher return on energy. All in all, therefore this project also, to a certain extent, contributes to climate action beyond the state-of-the-art.
How it works
Electroluminescence imaging is one of the most accurate and versatile solar panel diagnostic tools to date. As electroluminescence imaging is based on capturing light emitted by the solar cells operated as an LED device, it is highly sensitive to ambient light noise, and works best during dark/night time conditions. However, nighttime solar farm inspection poses safety concerns, therefore daylight electroluminescence inspection methods have been developed, successfully validated and applied in solar farm inspection in the last six years by researchers at DTU, allowing inspection of up to 2000 modules per day. However, scaling up for large solar farm inspection has reached a limit to what the stationary electroluminescence inspection method is capable of, meaning that mobile inspection and automation is necessary to bring it to the next level. Electroluminescence doesn’t include other imaging techniques from other sections of the electromagnetic spectrum that are of the interest of inspection professionals and installation owners and they often face similar limitations regarding upscaling, safety and automatic defect detection.
Ultimately, drone inspections are the safest option for building added or building integrated solar panels, where climbing roofs and cranes would be required otherwise. The drone will be equipped with a high-speed shortwave infrared camera, a high resolution visual camera, stabilization gimbal, communication interfaces and a high performance embedded computer for image acquisition critical processing. Moreover, it will be integrated with a high precision GPS positioning system and sensors for geo-localization of the modules. The operator will have full remote control of the position and speed of the rover, the camera pan/tilt, and over image acquisition. In addition, a path planning software will be developed to automate the inspection process.
A large extent of the development will include the implementation of new inspection technologies, installing a laser system to perform electroluminescence and photoluminescence imaging in a contactless way, generating spatially resolved images of solar panels that highlight poorly connected cell areas, such as finger failures, areas with high series resistance or disconnected cell cracks. The inspection speed is 10-20 times faster compared to solutions where the electrical system behind the modules have to be manipulated.
Image stabilization, correction and enhancement methods are required in order to automate the electroluminescence image acquisition and inspection process by use of mobile inspection platforms. The team will develop these to fill the current technology gap and provide accurate electroluminescence images and panel level diagnosis from drone captured data during daylight.
The images will be used to infer the electrical characteristics and system performance utilizing four different machine learning architectures with varying validation accuracy to detect cracks and shaded areas. The system examines the electroluminescence image of a solar cell and determines its acceptance or rejection status based on the presence and size of the crack. The system was validated with thermal testing using real-world cases, such as shaded areas and microcracks, which were accurately predicted by the system.
Team
The primary result of the project is an end-to-end service in the most accurate way to do solar cell troubleshooting in the world, namely electroluminescence imaging. The world-leading partner Above Surveying in the field of visual and infrared inspection participates in the project, which will expand their market area with the new luminescence technology and thus make use of multispectral technology for their customers. The hardware for this is being developed by AIR6 Systems, which will adapt one of their special drones so that it can carry both a luminescence camera and an IR camera, and thus the pilot on the same mission will be able to record both image types and thereby achieve the optimal analysis possible by utilizing both spectra – hence the name MULTISPECTRAL.
DTU and UxV Technologies also bring a completely new and ground-breaking way of making luminescence images into the project, where the solar panels do not have to be removed to make them emit electroluminescence, which is the current state-of-the-art. In the mentioned solutions, they are pumped with laser light from the drone and thereby emit electroluminescence that can be seen with a special camera from the drone, which is far ahead of the state-of-the-art but also a more risky track, and therefore developed two solutions based on adaptation of traditional technology, but also the more risky disruptive solutions in parallel.
The University of York participates together with DTU in the realization of image processing and analysis, which takes this beyond the state-of-the-art and implements it in Above Surveying's asset management software SolarGain, where a digital twin of the solar farm is implemented, and the new luminescence images are realized and imported and analyzed and can thus offer the customer the most accurate troubleshooting tools in the world at the present time.
With the University and York and DTU's AI-based analysis tools, the breakdown of the solar cells can even be predicted and thus result in energy loss over time, so that the customer gets a solid basis on which to decide on service on the plant. A very few players offer to make luminescence service today, but this project is the first time that it is sought to be realized in a hardware and software service that can be scaled and thus effectively make this market grow to become mainstream. Above Surveying already has a solid customer base within infrared inspection which can be used effectively as a route to market.
Facts
- Project title: Outdoor Luminescence and Infrared Imaging of Photovoltaic Systems
- IFD investment: 8.993.156 DKK
- Partners: Above Surveying, AIR6 Systems, University of York
