The project has made it possible for BrainBotics to create and train a Machine Learning algorithm that can reliably and automatically detect eelgrass floating in water, and is in itself a huge step towards the goal of collecting eelgrass in water for use as a green and sustainable raw material that can for example substitute plastics and certain building materials.
Here we showcase how the algorithm has automatically detected floating eelgrass in a sample image:
Through more than 50 drone flights, that the University of Southern Denmark has assisted on, we have furthermore been able to observe eelgrass every single day we have performed flights from June until early October, documenting a fact that until now has not been known; that even though eelgrass is not necessarily arriving at the coastline, there are still huge quantities floating in the water, available for collection.
We have been able to establish a direct relationship between current and wind conditions and how floating eelgrass behaves in the water, a relationship that has never before been studied in detail.
Based on our many observations we have also been able to reliably estimate the amount of eelgrass in a certain area, based on input from a single drone image. This provides direct and valuable input for companies to base future collection efforts on.
We here showcase the output from our many automated calculations:
The potential for our technology shown here, with the example of eelgrass, is however only the tip of the iceberg though – we are planning to use the technology to also detect, monitor and quantify macro-algae, micro-algae, plastic debris and oil spills, so that our oceans and waterways will face a brighter future!