Find defective panels from home

Find defective panels  from home

How do you know that your panel is broken? - ”If your panel is broken, it will heat up”, - says Otto Fabius, member of the Sobolt AI-team. But how do you know that it is heated up? - “That’s why images play an important role in the detection of defects”. Nevertheless, the devil is in the details: knowing the importance of infrared data is only the beginning, the hardest part is to perfect the automated tool, something Sobolt has achieved.

Sobolt is launching Defect Detect! This is a unique tool that takes into account over 570 fault parameters, identifies defective solar panels in solar parks. The product combines state of the art machine learning with knowledge in how PV panels work.

The process is automated. The images/video are collected and the techniques are applicable everywhere. Sobolt’s technology is based on a series of neural networks which are trained using large amounts of data

The importance of Defect Detect for operation and maintenance service is being tested the coming months. It provides a client with a report on the status of the solar installation and points out at the most significant faults. This technology can save a significant share of the O&M costs.

         

Comments are closed.