Norway-based aquaculture technology supplier Aquabyte has introduced a new camera-based system capable of measuring biomass and scoring the welfare of smolt as small as 50 grams.
The system, which employs machine learning algorithms and over one million daily images from smolt tanks, is designed to assist fish farmers in making data-driven decisions throughout the entire lifecycle of their stock.
“It’s fantastic that customers can now monitor fish from the start of the smolt phase. For farmers, it’s important to know the size and quality of the fish they receive so they can make the right decisions at the pen from day one,” says Kristin Aase, Insight Manager at Aquabyte.
Automatic weight measurement and welfare scoring with cameras also reduce the need for manual handling at the smolt facility.
“Less stress and fewer injuries from manual handling lead to improved welfare and health, making the smolt better prepared for transfer to the sea,” adds Kristin Aase.
The Aquabyte system uses a specially adapted camera that captures over one million images from the smolt tank each day. The images are analyzed using machine learning and newly developed algorithms for smolt as small as 50 grams.
“Monitoring smolt is different from monitoring fish in pens. The fish are smaller, the fish density in the tank is higher, and the behaviour of the fish is different. The algorithms used for weight measurement and welfare scoring must be specifically developed for these conditions, says Mochi Liu, Senior Data Scientist at Aquabyte.
Using the system in smolt facilities also places unique demands on the camera that is submerged in the tank. – The self-cleaning lens technology and adaptations we’ve made to the camera are critical to ensuring the system operates optimally in smolt facilities over time without the need for extra maintenance, says Mochi Liu.
Aquabyte, founded in 2017, provides systems for automatic lice counting, biomass estimation, and fish welfare monitoring, serving nine of the top ten aquaculture companies globally.
