Automated in-line defect classification and localization in solar cells for laser-based repair

Defective crystalline silicon solar cells may be repaired using laser-based techniques if the problem is properly identified and characterized. This paper presents a novel system for the automation of solar cells repair that carries out the following tasks: 1) It detects and localizes cracks and shunts in solar cells from electroluminescence images; 2) It takes a decision on the laser process to repair faulty cells; 3) It automates the operation of a laser machine for processing solar cells. Regarding the analysis of electroluminescence images of solar cells, the proposed solution is able to discriminate the type of defect, which means a step-forward compared to state-of-the-art approaches. Moreover, it is to our knowledge the first solution that takes the results of such analysis to automate a process of laser-based repair. The proposed system paves the way for waste reduction in the production of solar cells by using repaired cells in custom-made solar modules.