In Norway, the fishing industry is based on the principle of sustainable fishing and strict fishing regulations. Knowit created a solution helping the Norwegian Coast Guard efficiently identify illegal fishing.
Fishing is a large industry in Norway and has faced challenges due to illegal fishing. Such acts can lead to overfishing, which threatens species and destroys marine habitats. Knowit felt that this was a meaningful challenge to work on.
Knowit saw this as an opportunity to help out in the fight against illegal fishing. The solution was also Knowit’s entry in the competition Nordic SAS Hackathon.
This challenge was well-suited for use of open automatic identification system (AIS) data. Available AIS data were gathered and supplemented with further data on vessel type from marine websites. This was done to determine if a vessel was a fishing boat, a ship, or one of the Coast Guard’s own.
With the help of robotic process automation, RPA, all the necessary information about the vessels was gathered. This information was fed into the SAS Viya Visual Analytics platform, for analysis of each vessel’s movement patterns. Machine learning was used to train and refine the model, so it could determine if a movement pattern could be classified as a normal or abnormal fishing movement.
With the SAS Viya platform, it was easy to identify fishing patterns. We could easily spot the difference between vessels that were fishing and those that weren’t. The data we got from SAS Viya was used as training data for machine learning in Python. The information was then sent back to the platform, to automatically determine if a vessel was fishing or not. The data power and services in SAS Viya come in useful when you have large amounts of data.
Team Knowit
Using the model, historical data in an area of interest can be interconnected, to predict and highlight vessels that show behavior patterns which might indicate illegal fishing.
A further development of the solution was to connect an AIS data stream in real time. This creates a model that can automatically warn the coast guard if it discovers and identifies illegal activities.
The model was Knowit’s entry in the competition Nordic SAS Hackathon and was awarded first place.
Knowit has presented the solution to the Norwegian Directorate of Fisheries, the Swedish Agency for Marine and Water Management, and the Danish Fisheries Agency, among others, to provide inspiration as to what is possible.
