In this project, the goal is to research control policies for drones swarms and develop a control solution to fly a swarm of drones in simulated environment around an aircraft (e.g. in a hangar) such that their cameras, collectively, cover an aircraft’s fuselage/exterior. The problem is how to control each drone in the swarm as to inspect each part of the aircraft in an efficient manner, allowing a high quality reconstruction of the aircraft to be produced (e.g. via NeRF or photogrammetry).
The project is focused on aviation and offers you the opportunity to work with cutting edge drone swarm simulations in the domain of aviation maintenance. The outcome of your work will contribute to a more environmentally friendly and sustainable aircraft maintenance operation.
Tasks to be foreseen:
- Get acquainted with the current developments regarding drone swarming technology and 3D reconstruction techniques by means of a literature study.
- Research, experiment and evaluate different flight planning and/or control algorithms for drone swarming (in simulation).
- Present the outcomes of your study at NLR.
- A master thesis report
- A presentation at NLR
- A motivated master student in Computer Science, Artificial Intelligence, Data Science or a related field.
- Solid Python skills, with preferably experience or affinity with OpenCV and/or Unity.
- Experience or affinity with computer vision and control problems (e.g. reinforcement learning).
- Assertive and self-motivated, able to be part of the project team and also proceed individually
- The Dutch nationality in order to comply with security standards
What we offer
- A challenging graduation project/internship in a high-tech result-orientated work environment
- Weekly supervision and availability of the technical staff for support
- An internship allowance
- Working in an actual R&D project as part of the team
- Internship results to be used in the current and future projects
- A ping-pong table, billiard table, and air hockey for the necessary exercise in between.
Royal NLR has been the ambitious research organisation with the will to keep innovating for over 100 years. With that drive, we make the world of transportation safer, more sustainable, more efficient and more effective. We are on the threshold of breakthrough innovations. Plans and ideas start to move when these are fed with the right energy. Over 650 driven professionals work on research and innovation. From aircraft engineers to psychologists and from mathematicians to application experts.
Our colleagues are happy to tell you what it’s like to work at NLR.
This assignment will be managed by the Modelling & Simulation group within the Aerospace Operations Training & Simulation (AOTS) department.
Unmanned Aerial Vehicles (UAVs), also known as drones, have in recent years greatly transformed the landscape of civil aviation. Advances in battery technology, sensing and onboard computing have brought about significant developments in drone technology, which has enabled versatile drone concepts to be developed. These advancements have elevated the practical utility of drones for many use-cases and are making their implementation into real-world operations increasingly viable.
One interesting concept is that of drone swarming, where a group of drones coordinate to achieve a given task. A promising use-case for this technology is aircraft inspection. Drone swarms, by their distributed nature, can perform inspection of an aircraft fuselage much faster than a single drone or human operator, making it a serious option for future maintenance operations. When equipped with small cameras, a swarm of drones may, for example, be able to quickly capture many angles of an aircraft’s fuselage and create a 3D reconstruction of the aircraft using photogrammetry or NeRF “from scratch”. This reconstruction may then be displayed in a dashboard for a remote inspector on the ground to see.
However, for drone swarms to be used effectively and safely for aircraft inspection, many current challenges have to be overcome. For one, the aircraft to be inspected may be geometrically complex, as such, may need to be captured from many suboptimal angles to be reconstructed faithfully. Moreover, to optimally inspect an aircraft in the least amount of time, drones must consider parts already inspected by other drones (as to not do double work) and perform obstacle avoidance in order to prevent mid-air collisions with other drones, other equipment or the aircraft.
Send your application, together with your motivation letter and CV to Thomas Bellucci (Thomas.Bellucci@nlr.nl) and we will contact you as soon as possible
Datum : 01/02/2024
Locatie : Amsterdam
Uren : 40
Opleidingsniveau : WO
Achtergrond : Computer Science, Artificial Intelligence, Data Science or a related field