Fishing Prediction Maps

The interactive map displays the predicted fishing probability (FP) by zone and species for the next week, on a scale from zero to one. This variable represents how easy it will be to fish a species in a precise spot. Zero is assigned to areas where it will be more difficult to find and fish a species, while one to an area where it will be easier on a given day.

I worked closely with the data scientists to design and develop the interface of the webapp and its architecture. The goal was to deliver a simple and intuitive way to display their data without crowding the interface with too much information but still being useful to the end user.

The data produced by the model (running on the company server) get pushed every week on a AWS S3 bucket in the form of .pbf files. These files are then exposed at three different domains to improve performance. The front end then uses mapbox-gl-js to consume and render the .pbf files.

The webapp is developed in the form of a PWA (progressive web app) so that it can be easily installed and accessed.