Notification: you are in staging mode.

Are you ready for take-off?

Last week, the Droonhaavn (West-Flemish for Droneport) business incubator was officially launched with the broad interest of the press and local industry decision-makers.

Based in the seaport of Zeebrugge, Droonhaavn aims to further establish the economic potential of drones in Flanders. ML2Grow also demonstrated the potential of Machine Learning in drone applications, which is the result of fruitful cooperation between ML2Grow and Droonhaavn and its founders at Noordzee Drones, but which is also a long-lasting experience of our staff that goes beyond the foundation of our company.

Optimal sample collection

So were our CEO and CTO, Joeri Ruyssinck and Joachim van der Herten, part of the SUMO (SUrrogate MOdeling) Lab of the IDLab research group at Ghent University – imec, from which our company is a direct spin-off. One of the excellent software tools the lab created is the Albatros algorithm. This Adaptive Line-based sampling TRajectOrieS (ALBATROS) algorithm, where our CTO contributed to and which is published in international research papers, is a sequential sampling algorithm to gradually increase the sampling density over the entire measurement space while minimising the traversed path length of the measurement probe. It consists of a combined coverage path planning algorithm and a path sampling algorithm and is functional in any convex 2D or 3D space.

This is especially useful in a drone context. It allows drones to collect samples (e.g. air quality measurements or photographic samples) in an optimal cost-saving way, reducing flight time. It outperforms traditional one-shot and sequential sampling methods, optimally spreading sampling points in the entire sample space.

Machine Learning on drone data delivers real business value

We do not only provide cost-efficient sample collection using drones, but our Machine Learning solutions also allow the transformation of the collected data (images, video or other measurements) into valuable knowledge and automatic decisions. Think of applications in agriculture (harvest prediction, disease monitoring, crop protection,…), surveillance, infrastructure monitoring, element detection and labelling, stock monitoring,… It’s not only about automatic measuring, but also the next step that gives real business value: prediction. Think of, for example, from stock monitoring to stock prediction.

Our staff gives an introduction of these techniques, workflows and ready-to-use tools and platforms as part of the Noordzee Drones RPAS courses to become a fully qualified Class 1 or 2 drone pilot.

At ML2Grow, we congratulate the founders of Droonhaavn on this great initiative and look forward to a great cooperation with the whole Droonhaavn community in order to make it a great success story with a real positive economic impact on our companies and society.

Have a safe flight!

Peter Dedecker

Newsletter

Sign up for our newsletter.