Imagine having a super-smart digital eye that can look at millions of underwater photos and instantly spot pollution or rare fish. Instead of scientists spending years manually tagging pictures, this system uses AI to do the heavy lifting. It's like a shared cloud library where researchers can pool their data and use powerful computers to find patterns in our oceans and rivers.
Manual analysis of aquatic imaging data is slow, expensive, and inconsistent. Researchers and companies lack the computing power and standardized AI tools to process millions of images efficiently.
A scalable AI platform (transitioned to AI4OS) and a set of production-ready image analytics services for water pollution and biodiversity monitoring.
If you are a water quality monitoring firm dealing with slow manual analysis of river health — this project developed AI-powered image analysis tools that process massive online data streams. This allows you to detect water pollution faster using a system that has already handled over 9 million images.
If you are an AUV manufacturer dealing with the challenge of real-time species identification — this project developed 8 AI applications and a compute layer with 132,000 GPU-hours. This enables your hardware to integrate high-performance image processing for biodiversity studies.
If you are a coastal management agency dealing with beach erosion and monitoring — this project developed a scalable IT platform integrated into the European Open Science Cloud. You can now use production-ready services to automate beach monitoring and climate change analysis.
The project provides a portfolio of image datasets and analysis tools that are free at the point of use.
Yes, the platform has demonstrated scalability by supporting over 9 million images and utilizing 4.2 million CPU hours across four cloud providers.
Based on available project data, the tools are integrated into the European Open Science Cloud and AI4EU, suggesting an open science approach, though specific commercial licenses are not detailed.
The system is built on the EGI federation infrastructure and AI4OS, providing a generic layer for deploying AI models that can be adopted by researchers.
The first public services were launched in September 2024 and are currently accessible to users.
The consortium is heavily weighted toward research and academia, with 11 research organizations and 8 universities. However, the inclusion of 2 SMEs and 2 industry partners, alongside the coordination by EGI (a major infrastructure entity), indicates a strong focus on technical deployment. With 24 partners across 11 countries, the project has a wide geographic reach for validating aquatic data across different European water bodies.
Contact STICHTING EGI in the Netherlands for infrastructure access.
Contact us to find the specific AI model for your aquatic monitoring needs.
