SciTransfer
Bluegrove Welfare Shield · Project

AI-Powered Salmon Welfare Monitoring System to Reduce Fish Mortality and Boost Growth

foodPilotedTRL 7

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.

By the numbers
25%
reduction in mortality rates
15%
improvement in fish growth
60
production cages validated
13
sites validated
25,000
feeding days analyzed
The business problem

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.

The solution

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.

Audience

Who needs this

Salmon farming enterprisesAquaculture technology integratorsSustainable seafood producersFish health consultants
Business applications

Who can put this to work

Aquaculture
enterprise
Target: Industrial Salmon Farm

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%.

Agri-Tech
mid-size
Target: Aquaculture Equipment Provider

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.

Fisheries Management
any
Target: Sustainable Seafood Producer

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.

Frequently asked

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.

Consortium

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.

How to reach the team

Contact CageEye AS in Norway

Next steps

Talk to the team behind this work.

Contact us to explore licensing or partnership opportunities with CageEye AS.

More in Food & Agriculture
See all Food & Agriculture projects