The competition proposed three tasks:
- To build a model that can predict which coordinates a vessel should prioritize in order to maximize the likelihood of catching a specific type of fish
- To write a report based on the conducted analysis that can be read by experienced fishermen and to create a user-friendly visualization that a captain can read to make a assessment of where the vessel should search for fish the next day
- To make a Sustainable Fishing Plan; a weekly plan that suggests the routes the fisherman should follow to optimize fish caught and fuel consumption
The teams had access to various types of data, such as catch note data collected by the Norwegian Fishing Directorate, monthly averages of salinity data provided from the SMAP Salinity V4 dataset, moon phase data consisting of dates and exact times of full moons from 1900 to 2050 and sea surface temperature (SST) which had been collected by National Oceanic and Atmospheric Administration (US). The teams could also include and supplement available data with data from other publicly available data sources as they saw fit.
The team behind the FishAI competition recommended an exploratory approach when solving the set tasks and encouraged innovation when developing solutions. All participating teams were asked to submit a 2 page paper describing their method and results. The submitted papers will be reviewed single blind and will be published. Outstanding submissions will be invited to submit a full length paper to a special issue about the competition in the Nordic Machine Intelligence Journal.
The competition saw great interest and over 30 teams registered to compete for the winning prize. The FishAI Team have been impressed by the dedication of each team, by their creative approaches to each task, and for the innovative results submitted!
We want to congratulate Poseidon for winning the FishAI: Sustainable Commercial Fishing Competition 2022! As the winner, Poseidon will receive:
- 1 year start-up membership at NCE Seafood
- Mentorship and support with commercialization
- Support and practical work with soft-funding (up to NOK 100.000)
- Pre-seed funding network and support
We would also like to extend our congratulations to FishMaze, who was the second-runner up in the competition! In third place we have Intito and in fourth we have Craig Syms. Congratulations to all on your excellent efforts!
Below, you can meet the people behind Team Poseidon and Team FishMaze.
1st Place: Team Poseidon
The team: Jonas Dammen, Ludvig Løddesøl, Kristian Andersen Hole, Åsmund Brekke, Tomas Roaldsnes, Julia Ortheden
- How did your team come across the competition?
Jonas was the one who saw the competition first, on a post on Linkedin. Having worked on similar things with Ludvig before, he reached out to see if he was keen on a new challenge. We discussed how the challenge could be solved and what kind of team would be needed. Then we reached out to a group of friends whom we knew from work and studies that perhaps could be interested in taking part in this. And that was the beginning of team Poseidon.
- What triggered your team to participate in the competition?
The first time we all met to discuss the challenge, the team started to dive straight into possible solutions, challenges, and other applications of the dataset. One could easily see that this was something that intrigued the team and that we thrived on challenges like this. We also wanted to find out how well we work together as a team. We all thought Fish AI sounded like a good opportunity for this.
- What was your biggest take away from the competition?
The learning was a big takeaway, both in terms of domain and technical knowledge. Working with AIS data, geodata, and data from PO.DAAC,and looking into technical solutions surrounding these datasets was interesting as well as challenging. Funnily, everyone on the team now has a favorite vessel on the AIS map. It was also a lot of fun to learn about the fishing industry and to gain valuable insights from the datasets.
- This competition focused on increasing sustainability in commercial fishing. What is your thought on using AI in the fishing industry? What are the benefits?
The potential for innovation in commercial fishing is huge. There are a lot of individual actors and vessels who all might have their practices in terms of catching fish, but with new technology and publicly available data we can help skippers make better decisions to reduce fuel consumption and practice more sustainable fishing. There are also other interesting use cases related to the AIS data of fishing vessels, for instance detecting illegal fishing, which could be interesting to investigate further.
- What is the most exciting part of working with AI for you?
The difficulty is quite exciting. For example, taking on a task you are uncertain about how to solve. It is hard when you're working on it, but finding a solution feels very rewarding. Then you have the innovation aspect of it, where you get to take part in developing solutions that haven't been created before. Lastly, you have the intrigues connected to developing something that is “smart” and that can do a task better than for example a human being. All of these combined makes working with AI very exciting and it keeps you going when things are tough.
Since winning the competition, Team Poseidon have established and launched their own startup featuring their solution, namely Catchwise! Since their launch, Catchwise has had 800 users on their platform with over 70 shipping companies and are working closely with a small number to further develop the service. Read more about Catchwise and their journey here!
2nd Place: Team FishMaze
- How did your team come across the competition?
Earlier this year, we won Climate hackathon where we utilized HubOcean (https://www.hubocean.earth/press/news/climate-hackathon-2022) to create project AVERA: Automated Vessel Emission and Route Analysis. After the hackathon, Aldrin joined HubOcean public slack channel based on pure curiosity and to connect with the community. He accidentally found FishAI competition in one of the threads. Then he presented the idea and timeline of the competition to the team and the rest is history.
- What triggered your team to participate in the competition?
The topic is interesting. We are exploring domains in which we can apply our data science skills and we thought this is a great opportunity. We believe there is great value in pursuing AI in the fishing industry. To be able to create solutions and deliver a positive impact to the fishermen, it is definitely worth our efforts.
- What was your biggest take away from the competition?
Experiment. We will never know what works until we try. There are a couple of methods we tried which went wrong. But with continuous iterations (a lot of back and forth really), we eventually learned what works and what fails leading us to what is FishMAZE now.
- This competition focused on increasing sustainability in commercial fishing. What is your thought on using AI in the fishing industry? What are the benefits?
It is a novel idea. We believe there are a lot of untapped opportunities for AI in the fishing industry. Precision fishing is one of the benefits. Knowing when and where it is best to fish (which is what exactly the challenge is all about) would save our fishermen from unnecessary costs. Aside from this operational benefit, the use of AI tools can improve data collection in fisheries and this is beneficial in managing our resources.
- What is the most exciting part of working with AI for you?
Making complex data simple, that’s the exciting part. The process of transforming an overwhelming amount of data into insightful stories and valuable solutions, it’s fulfilling. Discovering patterns and trends never seen before, automating labor intensive tasks, supporting decisions with data-driven approaches – these are not easy, but knowing that we make data insights more accessible and comprehensible to people, it makes AI more exciting and meaningful.
NORA.ai would like to thank all partners involved in the FishAI Competition for their dedication throughout the process! A special shoutout to Michael Riegler (Simula), Steven Hicks (Simula) and Tor-Andre Schmidt Nordmo (UiT) for their time and effort!
For more information about the FishAI competition, click here.
This articles was originally published on the 17th of November, 2022.