By Yvonne Gerster
18. July 2024

Technical project report: “Swarm Init”

Project Number: 5.504.20001

Participating cooperations:
Lakeside Labs GmbH
University of Klagenfurt
AIRlabs Austria GmbH

Summary: LSL-drone flight without GPS Support

In this project, a drone from Lakeside Labs (LSL) was integrated into the infrastructure of the drone hall at the University of Klagenfurt (AAU). Previously, LSL drones could only be used for outdoor flights with GPS support. This project has made it possible for LSL drones to support other types of localisations. AAU’s drone hall offers an OptiTrack motion capture system that allows drones to be localised using cameras. During the project, the drone was adapted in terms of both hardware and software and integrated into the OptiTrack system. Final test flights demonstrated the functionality of the system.

Project goals and context

The aim of the project was to prepare an LSL drone for flights without GPS support. This should lay the foundation for future projects in which drones and drone swarms have to localise themselves without GPS support. This project was intended to gather the necessary experience and prepare the hardware and software to be able to set up a GPS-less drone system more quickly in the future.

Project implementation and fulfillment

The project began with a kick-off meeting on 13 November 2020, attended by Andreas Kercek and Micha Sende from LSL and Stephan Weiss from AAU. The project objectives were defined and the technical challenges discussed. LSL was given a brief overview of the capabilities of the OptiTrack motion capture system installed in the drone hall and the preparations required to integrate a drone into the infrastructure of the drone hall. It turned out that both hardware and software adaptations were required on the drone. The necessary software packages were provided by AAU [1]. On the hardware side, reflective markers are required for installation on the drones [2].

Next, a twinFOLD SCIENCE drone from LSL’s inventory was prepared for the test flights in three steps. The first step was to install the reflective marker spheres on the drone (see Figure 1). It is important that at least three spheres are attached in linearly independent positions.

Drone from Lakeside Labs (LSL) was integrated into the infrastructure of the drone hall at the University of Klagenfurt (AAU). In order to make this possible with the OptiTrack Motion Capture System, adjustments to the drone's hardware and software were necessary. The built-in reflective marker spheres can be clearly seen here.
Figure 1: Drone with installed reflective marker spheres.

The second step of the customisation consists of a user-defined firmware for the flight control. This is based on the PX4 autopilot [3] whose parameters were adjusted and the feedback from the OptiTrack motion capture system was used for localisation.

Als drittes wurde die notwendigen Software-Pakete am Bordcomputer installiert und konfiguriert. Zentral ist dabei die Virtual Reality Peripheral Network (VRPN) Software [1]. Diese empfängt die vom OptiTrack Motion Capture System berechnete Position der Drohne über eine WLAN-Verbindung und leitet sie lokal an die Flugsteuerung weiter.

Thirdly, the necessary software packages were installed and configured on the on-board computer. The Virtual Reality Peripheral Network (VRPN) software [1] is central to this. This receives the position of the drone calculated by the OptiTrack Motion Capture System via a WLAN connection and forwards it locally to the flight controller.

Once the preparations had been completed, the first integration steps took place on 20 November 2020. The prepared drone was integrated into the OptiTrack system. The cameras recognise the individual arrangement of the reflective marker spheres. A special identification is then assigned to this arrangement which allows the drone to be clearly recognised and localised in the three-dimensional space of the drone hall.

Finally, test flights were carried out on 11 February 2021. The drone was controlled manually and flew to predefined waypoints. The log data of a flight can be seen in Figure 2. The small deviation between the red and green lines clearly shows that the localisation of the drone worked well.

Project resources

  • 1,5 days (incl. preparations)
  • twinFOLD SCIENCE drone (provided by LSL)
  • Drone hall (provided by AAU)
  • Reflective marker spheres (purchased by LSL)

Project results and conclusions: successful execution of flights

In this project, a drone from LSL was integrated into the AAU drone hall’s OptiTrack system. Flights were successfully carried out.

The drone hangar provides a safe environment to test individual components (hardware or software) before flying outdoors. However, it only offers a limited size, which does not allow for more complex and large-scale missions.

Outlook and recommendations

For future projects, the drone hall offers an opportunity to test individual components in a safe environment. This applies in particular to drone flights that require localisation without GPS.

Publications/Literature

[1] https://github.com/aau-cns/ros_vrpn_client
[2] https://docs.optitrack.com/motive/markers
[3] https://px4.io
[4] https://github.com/aau-cns/flight_stack/wiki

Get in touch.

Inquire now