In this abstract, results are presented from the study of the behaviour change of Atlantic salmon (Salmo Salar) in an industry scale fish farm when exposed several different influence factors. Data is gathered to study the fish behaviour change when exposed to structures with varying shape, size, light, color and sound, shown in Figure 1 . Utilizing these data, the fish reaction to the structure can be studied for different case studies to identify and quantify the distance that the fish keep from the different structures and how the fish reacts when exposed to the different impact factors .
This work was financed by the Research Council of Norway through the project: CHANGE and RACE Fish Machine Interaction .
The structure was in each case equipped a stereo camera, and two BlueRobotics Ping360 echo sounders set to collect data in 360⸰ swipes with a range of 5 meters . Four EK15 echo sounders and an underwater speaker were also installed on the structure for parts of the trials . Stereo camera videos and echo-sounder data have been obtained from cages in an industrial scale fish farm which is part of SINTEF ACE  during field trials in the fall of 2022 and compared to outcomes from results obtained in 2021.
Sound tests were carried out using the big yellow cylinder structure, which was lowered to 8 meters depth within the fish cage, as illustrated in Figure 1 . The fish was then exposed to frequencies of 200 Hz and 600 Hz, as well as propeller sounds, for a period 1 minute, followed by 10 minutes of silence between tests . Furthermore, e xperiments for light were done for the base structure without the yellow or white outer layer, as shown in Figure 1, with additional EK15 echo sounders installed. The structure was lowered to 8 meters depth within the fish cage and four different light intensities were tested to see the reaction of the fish. Finally, tests for shape, size and color were done using the six different structures shown to the right in Figure 1.
Sound tests from previous field trials showed that the fish generally had a flight response to 200 Hz and 400Hz when played for a shorter period (10 seconds). As it can be seen in Figure 2 , the distribution of the fish in much smaller when the sound is activated . These observations were also supported from statistical analysis outcomes and behavioural biologist analysis. However, b ased on the observations made from watching the stereo camera feed and 360 ping data during the trials in 2022 , it was not possible to see any clear flight response or other reaction from the fish in any of the sound tests , see Figure 3 . The same goes for the light tests, the fish did generally not show any clear rea ction based on what can be sa id from the visual observations from the stereo camera feed. Processing of the echo sounder data and the stereo camera videos can show if there was in fact any change in the general behaviour of the fish and the distance it kept to the structure that did not show clearly on the video feed.
Large data sets for different case studies have been collected from field trials in industrial scale fish farms to study the fish behaviour change under the influence of different impact factors. Both acoustic and video data were observed to check the fish behaviour change, however, the fish did not show any clear reaction to the sound, light or motion. Based on previous field trials, Atlantic salmon seemingly change beh aviour when exposed to audio signals with a frequency between 200 Hz and 400 Hz. I t is therefore interesting to investigate further why the fish did not show any clear reaction to the same frequency for the case studies presented here. If the data processing of the sonar data and formal statistical analysis gives the same results, the next step could be to look more closely into the size of the fish, as well as the time of day when the field trials were performed.
 RACE Fish-Machine Interaction. https://www.sintef.no/en/projects/2020/race-fish-machine-interaction/
[2 ] SINTEF ACE. https://www.sintef.no/en/all-laboratories/ace/