If you are an industrial salmon farm dealing with high fish loss and unpredictable growth — this project developed an AI monitoring system that can reduce mortality rates by 25% and improve fish growth up to 15%.
AI-Powered Salmon Welfare Monitoring System to Reduce Fish Mortality and Boost Growth
Imagine having a 24/7 security system for fish that doesn't just watch them, but understands how they feel. It uses underwater sound waves and cameras to tell if fish are hungry or stressed by analyzing their movements. This lets farmers fix problems before the fish get sick, much like a fitness tracker for an entire fish farm.
What needed solving
Salmon farmers struggle to monitor fish welfare in real-time, leading to unexpected mortality and suboptimal growth rates due to undetected stress or appetite issues.
What was built
An AI-driven monitoring system consisting of digital hydroacoustic sensors, feeding camera integration, and a remote dashboard for welfare scoring and appetite tracking.
Who needs this
Who can put this to work
If you are an equipment provider dealing with a lack of smart data for customers — this project developed a digital hydroacoustic sensor and dashboard that provides live analysis of fish behavior and appetite.
If you are a seafood producer dealing with strict welfare regulations — this project developed a statistical welfare scoring system that allows you to benchmark performance against the most productive cages.
Quick answers
What is the cost or pricing model for the WelfareShield system?
Based on available project data, specific pricing or cost details are not provided.
Has the system been tested at an industrial scale?
Yes, the system was installed and validated in 13 sites and 60 production cages across Norway.
What is the IP or licensing status of the AI models?
Based on available project data, the project developed deep learning models using YOLOv8 and heuristic movement models, but specific licensing terms are not mentioned.
How does the system integrate with existing farm hardware?
The system integrates digital hydroacoustic sensors and standard feeding cameras to collect data for the dashboard.
How long did the development and validation phase take?
The project period ran from 2022-07-01 to 2024-12-31.
Who built it
The project was led by a single SME, CageEye AS from Norway. This 100% industry-led consortium indicates a strong focus on commercial application rather than academic research, with the company managing all development and industrial validation internally.
Contact CageEye AS in Norway
Talk to the team behind this work.
Contact us to explore licensing or partnership opportunities with CageEye AS.