What behaviour do we need from autonomous drones

6 september 2021

When it comes to innovation in aviation, drone development is set to make the biggest strides in the near future. Especially autonomous drones, whether they are used for last-mile delivery or security purposes. Access to reliable data and machine learning will be pivotal for their use. Our newest colleague Jerry is specialising in this.

In an earlier blogpost we spoke about the future of drones in aviation and U-space; the regulatory framework of communication standards to make sure that drones can safely use the airspace and communicate with each other and with ATC. One of the key elements in drone development will be the behaviour of autonomous drones. How will they function in a network of man operated aircraft and ATC, how will they interact with controllers and what is the best way to relay commands to them?

Distributed Machine Learning

Our newest colleague Jerry Knuyt, has taken a special interest in this. As a student at TU Delft, Jerry is writing his MSc thesis on the effective coordination of surveillance through online learning using autonomous drones. “Machine learning (ML) is essential for autonomous drones. In my opinion however, the focus should be on distributed ML, rather than on a centralised perspective, i.e. the entire network of (autonomous) drones, ATC, aircraft and the communication between them.” Take surveillance drones, for instance. “Airports could use them to monitor birds along active runways. If you can ‘learn’ an individual drone to recognise a runway change and move between different areas to autonomously identify hazardous bird (flocks), without the need of a controller to give it these commands, it alleviates the workload of controllers and makes the drones more flexible in their use. This can only work if you have reliable, accessible data.”

Aviation in his blood

Drone development is just one of the interests of Jerry. “I always wanted to become a pilot, but unfortunately that did not work out. That is when I learned about aviation engineering. A professor at TU Delft got me in drones. At Adecs I found a place to use my skills to help airports innovate. The fact that they also advice airports on other topics like noise reduction, really ignited a spark in me. Now I can work on both these topics,” Jerry says with a smile.

Autonomous drone technology is developing rapidly and airport operations can benefit from these technologies when it comes to safety. Reliable data is key. Adecs helps airports to make as much use of these technologies as possible. And with Jerry as our newest team member, we will all benefit.

Schrijf je in voor onze blog updates