Case Study - Fishwell background
Case Study - Fishwell background
Case Study - Fishwell logo

A leading technology company in fish farming in Norway

Fishwell develops technology services for fish welfare in aquaculture by monitoring fish health and reducing mortality rates to increase profitability and sustainability.

Case Study - Fishwell background
Case Study - challenges icon

Challenge

Traditionally, monitoring fish welfare relies on manual observation, which can be subjective and time-consuming.

That's why Fishwell sought a solution that could provide real-time insights into the health and well-being of their fish populations. They needed a more objective and efficient approach to count the dead fish and measure fish stress levels to ensure optimal fish health and growth in their sea cages.

They turned toward Neurond, a leading AI company, for help.

Case Study - Fishwell background
Case Study - Solutions icon

Solution

Receiving Fishwell’s request, we quickly assembled a dedicated team of nine experts, including AI Engineers, Data Engineers, Backend Developers, and Frontend Developers, led by a seasoned Project Manager for overall consulting and management.

Fish welfare proved to be a complicated area due to the unstable and massive amount of data provided for analysis. Understanding this, our team spent the first few weeks researching and giving probable solutions.

The team meticulously researched the behavioral and physical indicators of stress in fish, including unusual swimming patterns, changes in skin color, and erratic movements. We leveraged computer vision (CV) technology to develop a system that can recognize these stress signals in real time, allowing Fishwell to pinpoint areas of concern. It also integrated alerts, notifying fish farmers of potential issues like high stress levels or deceased fish detection to intervene to improve fish well-being swiftly.

Technologies Used

  • YOLOv7
  • OpenCV
  • AzureML
  • ByteSORT
  • Python
  • Pytorch
  • Tensorflow
Case Study - Fishwell background
Case Study - Impact icon

Impact

The Fish Welfare system yielded remarkable results. It achieved an impressive over 95% accuracy in detecting dead fish and fish stress levels and providing diagnostic insights. This accuracy was rigorously validated by independent fish experts, solidifying the system’s effectiveness.

Fishwell was thoroughly impressed by Neurond’s leadership in developing a groundbreaking stress recognition model specifically designed for the aquaculture industry. This innovative technology empowered Fishwell to enhance fish welfare, boost growth rate to maximize yields, and optimize resource utilization.

Both were committed to continuous improvement and future expansions, paving the way for a future of sustainable and ethical fish farming.