AI (Artificial Intelligence) in EL Testing
When electricity is run through Solar PV Cells, they convert this energy to light. This phenomenon is called Electroluminescence or EL.
This property of solar PV cells can be used at any stage of their life cycle to look for cracks, defects, and damages in the cells. Something like an x-ray report.
EL testing is conducted twice during the solar panel's manufacturing process, before and after lamination. This test is carried out by passing an electrical current through the panel and capturing the image of the panel, similar to an X-Ray, using a specialized camera.
The images thus captured are then inspected by a trained professional. It requires alertness and a high degree of attention to detail.
However, when done manually only, fatigue can set in quickly because of having to look at a similar image repetitively.
When conducted manually, even after an EL defect detection inspection, there is a possibility that a few modules with small micro-cracks may pass through the inspection without being noticed. We have been witnessing advances in machine learning and artificial intelligence that can benefit solar module manufacturing. A key area that these technologies are being applied to is EL Testing using a combination of Machine Learning and Artificial Intelligence.
The Automatic AI-enabled EL tester helps by highlighting areas with micro cracks.
RenewSys has been leveraging these technological advances, and is successfully using AI-based EL defect detection software. It detects all types of defects (linear cracks, spider cracks, AAL, among others) and works in sync with existing systems, allowing us to process EL images faster and with nearly 100% accuracy in real-time.
The system takes an image of the module while passing a current through it and the defects are highlighted on the screen that the operator is viewing. Each type of defect is highlighted in a different colour.
At RenewSys every single module that we manufacture goes through EL inspection twice. This amounts to thousands of modules at each of our plants, every day.
Leveraging AI-based EL testing helps us establish the highest levels of accuracy and decreases the time for this inspection process as well. Both of these factors directly benefit the end customer.
AI-based EL testing has helped improve both the speed and accuracy of EL testing assuring customers of accuracy levels that would not have been possible otherwise and benefitting installers, project owners, and homes owners with Solar PV Systems installed.