FishAI: Sustainable Commercial Fishing
The Norwegian Government aims to reduce CO2 emissions from domestic fishing 50% by 2030. Artificial Intelligence combined with your skillset might be the key to achieve this goal!
With a fishing zone of 2.1 million square meters, Norway is considered Europe's largest fishing and aquaculture nation. Every year, commercial vessels land fish to a total value of around NOK 20 billion in Norway.
At an overall level, the migration patterns of a variety of fish species are relatively predictable. A fisherman knows, for example, that the mackerel season starts in mid-September and plans accordingly.
On a daily basis however, fish can move over large distances, and with the main decision-making tool being the captain´s experience and intuition, boats typically search for days, even weeks, before making a catch.
In other words, there are great environmental benefits and opportunities in optimizing commercial fishing activities by reducing unnecessary transport distances. We believe that smarter use of publicly available data is key to transforming one of the oldest industries in the world.
We invite you to take a deep dive into the data and transform them into a powerful decision-making tool for the Captains of the 1100 vessels operating across the Norwegian fishing zone!
About the dataset:
We have made the following datasets available to you:
Catch Notes Data: The data contains catch notes collected by the Norwegian Fishing Directorate from 2000 to today for vessels larger than 15 meters. The notes consist of information about the catch that is manually logged during landing, e.g., when it was caught, where it was caught, what equipment was used, the species distribution of the catch etc. There are approximately 130~data fields and around one million notes each year
Salinity Data:
Monthly averages of salinity data from 2015 to present day is provided from the SMAP Salinity V4 dataset. Salinity (in combination with temperature) affects the growth rate of microalgae. This can potentially affect the migration patterns of fish. Eight-day running averages are also possible to obtain if needed (https://salinity.oceansciences.org/data-smap-v4.htm).Moon Phase Data:
The moon phase data consists of dates and exact times of full moon from 1900 to 2050. Lunar phases affect the migration and behaviour of fish due to water levels changing. Therefore, it is potentially possible to use this data source for modelling of the movement of fish. The dataset is published at https://www.kaggle.com/datasets/lsind18/full-moon-calendar-1900-2050.Temperature Data:
Sea surface temperature (SST) from 1981 to present has been collected by National Oceanic and Atmospheric Administration (US). It contains daily estimates of SST globally. The data was collected from satellite observations, and consists of daily data at 0.25 degree latitude x 0.25 degree longitude resolution. We have included the subset of data from 2000 to present day. The dataset is published at:https://www.psl.noaa.gov/data/gridded/data.noaa.oisst.v2.highres.html.
All datasets can be found and downloaded here: https://tinyurl.com/54w5bvxa
We invite you to submit your answer to the following tasks:
Task 1: Build a model that can predict which coordinates a vessel should prioritize in order to maximize the likelihood of catching a type of fish of your choosing (haddock or mackerel is most valuable for our industry partners). The prediction can be based on historical data.
Task 2: Create a report of your analysis that can be read by experienced fishermen; an user-friendly visualization that a captain can read to make a assessment of where the vessel should search for fish the next day
Task 3: Make a Sustainable Fishing Plan; a weekly plan that suggests the routes the fisherman should follow to optimize fish caught and fuel consumption
To compete for the prize awards all tasks are mandatory. Submission of only one sub-task is allowed, but will not be eligible for winning any of the prizes.
Proceedings and Awards
All participants are asked to submit a 2 page paper (double column, plus 1 additional page for references) 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. For more information about the NMI journal and to download templates, please visit the journal homepage https://journals.uio.no/NMI
We will also offer the following prices to the winning team:
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
Evaluation Methodology:
Task 1 will be evaluated using an unseen test set. We will use standard metrics such as precision, Fscore, accuracy, mean absolute error, etc.,
Task 2 and 3 will be evaluated from an expert team consisting of experienced fisherman and data scientists
Click here to access an overview article over the FishAI Competition and datasets, published in the Nordic Machine Intelligence Journal
Important dates:
May 9th, 2022: Competition launch and dataset release
June 14th, 2022: Digital Kick off and Q&A for registered teams
September 9th, 2022: Deadline for submitting results.
September 23rd, 2022: Jury evaluation of results.
October 21st, 2022: Deadline for submitting method description paper.
November 14th, 2022: Presentation and winner announcement!
f you have any questions, please email Birte Hansen ([email protected])
