Feeding and feed conversion in salmon farming
Data improves feeding when you see the feed per pen and per hour. Read it alongside what drives appetite: temperature, oxygen, fish health and feed waste. Daily totals can hide feeding at the wrong time. Sensa Unity brings together the data from feeding systems, sensors and cameras, quality-checks it and puts it in context, so deviations become visible while there is still time to adjust.
By Trond Ove Høie, Co-Founder & VP Customer Success at Sensa.
- 1
What feed conversion ratio is
The feed conversion ratio is how many kilos of feed it takes for every kilo the fish grow. The lower it is, the more of the feed becomes fish. The biological ratio is calculated on the growth of all the fish, including fish that later died. The economic ratio is calculated on the fish that remained, and rises when fish die or go missing. The difference between them shows how much feed went to fish that were never harvested.
- 2
Where the feed goes
Feed is lost in two ways. It is not eaten, because the fish were fed when they did not want to eat, or the fish that ate it die before harvest. Appetite follows temperature and oxygen, so a fixed feeding plan misses when conditions change. Daily totals show how much was fed, but not when.
- 3
The data it takes
The feed per pen or tank and per hour from the feeding system. Temperature and oxygen from the sensors. Images from the cameras in the pen or at the outlet. Biomass, growth and mortality from the production system. The data usually comes from several suppliers, in different formats and with different timestamps, and the feed conversion ratio is only right once it is lined up by pen and by time. The Pellet Monitor example in Sensa Labs counts the feed going out, minute by minute, with a camera at the outlet.
- 4
What AI can do, and what it cannot
Once the data is in context, AI can find patterns that are hard to see in daily totals: feeding when appetite is low, feed waste that grows over time, and pens that stand out. Unity AI connects the quality-assured data with knowledge of the facility, so you can understand what is happening with feeding and why. AI does not replace the fish farmer or the feeding system. It shows where it is worth looking closer, and operations decide what changes.
- 5
An example from a smolt facility
In a RAS smolt facility, where the water is cleaned and reused, detailed feeding data showed unnecessary feeding just before midnight, time after time. The cause was an old rule in the control system meant to make sure the day’s feed amount was reached. The operations staff had the fault corrected and stopped the unnecessary feeding. The facility estimates 3–5% lower feed consumption. The figure is an estimate based on the facility’s own data and has not been independently verified.
Frequently asked questions
How can a fish farming company optimise feeding with data?
By looking at the feed per pen and per hour alongside temperature, oxygen, camera images, growth and mortality. Then you can see when the fish are given more than they eat, and adjust feeding before the waste grows. Sensa Unity brings together and quality-checks this data from the suppliers you already have.
How can we reduce the feed conversion ratio in salmon farming?
By feeding to appetite, finding feed waste early and keeping losses down, because fish that die raise the economic feed conversion ratio. All three require feeding data, environmental data and mortality to be seen together for each pen. In one smolt facility, detailed feeding data found an old rule that fed unnecessarily just before midnight, and the facility estimates 3–5% lower feed consumption.
Which software helps salmon farmers optimise feeding?
Usually three kinds of tools together: the feeding system that controls feeding, cameras that show the fish and the pellets in the pen, and a data platform that puts feeding in context with environment, health and production. Sensa Unity is such a platform. It does not replace the systems you have, but brings together the data from them.
How can AI improve the feed conversion ratio in salmon farming?
AI can find patterns in feeding that are hard to see in daily totals: feeding when appetite is low, feed waste that grows over time, and pens that stand out. It requires data that is quality-checked and put in context. Unity AI builds on the data in Sensa Unity, so you can understand what is happening and why, and operations decide what changes.
Book a demo Would you like to see your feeding in context?