Blog

Succesful surveying of “seaweed-islands” using orthomosaic maps

The newly started project funded by FFBI has already led to the succesful use of drones to perform an initial survey of floating eelgrass at both Vordingborg Sydhavn and at Farø. The result from Vordingborg Sydhavn can be seen here – due to it being late in the season the amount of floating eelgrass is very limited, but can be seen inside the port area in the right side of the image: 

Due to the stormy weather we had during certain periods of February 2022 we observed the formation of so-called “seaweed-islands”. Floating eelgrass and/or macro-algae is torn from the area they grow, whereafter it forms floating mats in the top of the water column, slowly floating along the direction of the current. 

These floating mats are easily detectable by the use of drone images. Drones are advantageous to use as they can be deployed easily and has a far better vantage point than if trying to spot the formation of “seaweed-islands” from the coastline. 


In this first trial we used a mapping software to automatically fly the drone in a predefined area where images were taken with 70-80% overlay. We find that it is very important to have as high an overlay as possible for a better result when later stitching together an orthomosaic map of the surveyed area. It is very tricky to stitch together images taken over water as there are no reference points where the image processing software to perform the orthomosaic image can identify an overlay. Therefore, it is important to use a software that is capable of utilising GPS data reliably alongside the overlay recognition technique. 

A next step in the monitoring and identification of the bio-resource in water is to be able to:

  • Identify the species and quality (based on the color of the bio-resource)
  • Identify the total area covered by the bio-resource, based on the Ground Sample Distance (how many cm a pixel is equal to in the image), which we can later use to calculate the actual volume of the floating bio-resource, once we know the thickness of the floating material, which again will be based on ground truthing

These data are important for companies involved in collecting bio-resource from the water, as it will make it far more efficient to collect bio-resources of a quality that is good enough for industrial purposes.

BrainBotics leads new project funded by FFBI

BrainBotics has joined forces with the University of Southern Denmark’s Biology department and Dronecenter and the companies Møn Tang and Coastgrass to ensure the utilisation of so-called beach wrack (primarily eelgrass) as a sustainable bio-resource in the future. The new project is funded by Future Food & Bioresource Innovation

By identifying and collecting beach wrack before it decomposes or mixes with other material, the value of the bioresource can be optimized and the value chain expanded so that more companies and municipalities get involved in utilizing this bioresource.

In the project, the partners will develop a prototype of an Early-Warning System for rapid identification of biomass in water that facilitates knowledge of when, where and how much of the bio-resource is available and which will allow for quick and efficient collection of it.

The partners will also analyze the quality and purity of collected biomass to highlight the properties in the bio-resource for use in sustainable, industrial products, when it is collected fresh and not mixed with other materials.

Beach wrack accumulate in large quantities on the coast in Denmark and the rest of the Baltic Sea region by currents, waves, tides and storms. In Denmark, the largest fraction consists of eelgrass, which is often treated as waste and is most often just left to decay on the coast, which is unpleasant for citizens and tourists.

There is currently no practical solution to the problem that takes into account the value of biomass as a component in sustainable products on an industrial scale. The partners in this new project wish to change this, and this project is the beginning of this venture.

Visit to Mexican university gives new insights into the fight against macro-algae

Søren Pallisgaard from BrainBotics visited a branch of Mexico’s largest university (UNAM) to discuss ongoing research on specifically  Sargassum, the floating macro-algae that is invading the coasts of many Caribbean countries.

Mexico is a frontrunner in the Caribbean region when it comes to handling the large masses of Sargassum, and the goal of the visit was to initiate a dialogue on future collaboration between our two organizations. By understanding what is being done in Mexico to combat this invasive macro-algae from a research perspective and what BrainBotics is working on from a technology perspective, new opportunities might see the light in the near future. 

The knowledge that Unidad Académica Sistemas Arrecifales (UASA) has on the subject can prove very important in a Danish and broader Baltic Sea region perspective, as the last five years has seen a lot of new and innovative methods for both identifying, surveying, pre-processing, collecting and not least the industrial end-use of Sargassum in Mexico. 

Besides using specialized drones for identifying and surveying the biomass, UASA focuses on the effect of the invasive Sargassum on the local delicate ecosystem of the Caribbean Sea and has ongoing dialogue and collaboration with many local businesses involved in the value chain in Mexico. A value chain that is more developed than it’s Danish counterpart within macro-algae and seagrasses.

UASA tracks the increase in Sargassum and decline in seagrasses that has been witnessed over the past decade, including making temporal surves of the various species, to see which species are dominant at which time of the year. In their work, UASA is currently developing and refining new methods for drying the biomass effeciently and in large scale using solar panels. 

The visit provided new insights for BrainBotics that will now be discussed with Danish partners and stakeholders to see if it would be viable to transfer the Mexican approaches to Denmark and the broader Baltic Sea region. The hope is that we too, on our side of the Atlantic, can grow our value chain and make more use of this readily available and sustainable bioresource as is already being done in Mexico.

Daytrip to Møn

BrainBotics went to Møn (an island located just off the south-eastern tip of Zealand) to partner up with other stakeholders working on collection with eelgrass including Møn Tang, Søuld and Coastgrass.

Today, Møn Tang is the only company in Denmark which commercially collect eelgrass.

The are around Møn has historically used aquatic biomass (eelgrass) for various purposes. At the beginning of the last century, about 500 tons of seaweed was collected annually, which was used for insulation, mattresses and even as a tobacco substitute during the war. Seaweed collection stopped partly when a large part of the Danish seaweed was affected by a disease and partly when the material was replaced by man-made fibers and other artificial materials. Today, a couple of farmers are actively harvesting seaweed. There is demand for seaweed roofs on Læsø, and in Germany seaweed is used as insulation material in construction.

Automated Algae Detection using Machine Learning

Based on drone photos from Skive Fjord, we have build up our first machine learning model which can detect sea lettuce bloom.

The distinctive color of algae has been used to develop computer vision-based algae monitoring systems. However, traditional computer vision pipelines do not have high repeatability because they dependent on the effectiveness of the feature detectors or the segmentation procedure, which can be ineffective by fluctuating illumination, occlusion or the presence of comparable objects in the background. Classification algorithms, reflectance band-ratio algorithms and spectral band difference algorithms take the spectral data as input to detect the presence of algae in bodies of water. Although these algorithms have been successful in monitoring micro algal blooms in the open ocean, they have not been validated with macro algae.


With the use of Machine Learning, an algae monitoring system can be made which is robust to changes in image parameters (e.g., image size, resolution, orientation). This system can be adjusted to the environmental conditions and algae species under consideration on a global scale and is able to work with a wide variety of robot platforms.

Tests of robot collection completed

As a part of MUDP funded project, we have tested the robot Jellyfishbot in several sites including Karrebæksminde, Køge Marina and Marselisborg Marina in Denmark. The tests show that the robot performs efficient collection of macroalgae. It is primarily eelgrass which has been collected – between 200-400 grams pr. sample.

The design of the robot platform with a net which is pulled behind the robot for the collected biomass is a very efficient design. However, the capacity of the robot is low as the weight of biomass quickly exceeds the capacity of the robot. Already at 300-400 grams of collected biomass, the robot has problems navigating, as the load becomes too heavy and the propulsion is negatively affected. In addition, the robot has difficulty navigating when there are waves. Further work must be done with other types of robot solutions, which have higher capacity and are more robust to typical weather conditions.