Industry: Manufacturing
Our client is India’s first and largest vertically integrated energy company that offers products and services across the spectrum of photovoltaic manufacturing. The fully integrated, high-efficiency Solar Panel manufacturer is currently building the world’s first-ever 10 GW Solar PV manufacturing ecosystem in the country.
Business Challenges
Equipment failure can occur on a small part of a large, complex plant and go unnoticed for a long time.
The main goal was to collect the images from the cloud and add them to a centralized space where results could be predicted.
The traditional method of detecting damage to solar panels was an overhead for the customer due to the huge labor costs involved.
The client wanted to identify the damage classes for their solar panel system by analyzing images of solar cells and predicting the type of defect.
Solutions
Celebal Technologies proposed a data-driven solution to help anticipate failures and optimize maintenance interventions, resulting in increased plant performance.
Data was transmitted from their servers to Blob Storage.
An AI model was developed to detect two damage classes in EL images.
A user interface was designed for uploading the data and displaying the inference results.
Technologies Used
Azure ML
Azure Blob
Key Vault
Cosmos DB
Azure DevOps
AI Models
Key KPIs
Business Impacts
A prominent renewable energy firm successfully harnessed the potential of predictive maintenance and remote monitoring through a collaborative partnership with Celebal Tech.
