Solar Panels Identification & Defects Recognition using Artificial Intelligence (SpiderAI)

The project at a glance

Start date:

01 Feb 2024
Duration in months:

36
Funding:

SolarCleano
Principal Investigator(s):

Raphaël FRANK

About

The partnership project between SolarCleano and SnT aims to develop an innovative, fully autonomous robotic solution for cleaning and inspecting solar panels, specifically targeting utility-scale photovoltaic (PV) installations in remote and dusty regions. The core of the project is the SolarCleano B1, an all-electric gantry cleaning robot designed to integrate advanced defect detection capabilities using artificial intelligence. This integration will allow simultaneous cleaning and inspection, significantly reducing operational costs and increasing the efficiency and lifespan of solar investments. The project involves creating a digital twin for PV modules, developing a machine learning architecture for defect detection, and establishing a secure, scalable data processing and reporting infrastructure. By leveraging MLOps best practices, the solution ensures continuous improvement and deployment of detection models, providing clients with detailed, actionable reports to enhance maintenance decision-making. The project addresses key technical challenges, including data handling, real-time processing, and modular scalability, ultimately aiming to deliver a robust, cost-effective inspection and cleaning service for large-scale PV installations.