‘Skywalker’ flight tests

The project was based on Katja Rinck’s bachelor’s thesis, which was written in collaboration with AIRlabs Austria. Her scientific work laid the methodological and technical foundation for the first flight tests and the further development of the technologies used. The focus was on the question of whether a system could be developed that was as cost-effective and easy to operate as possible – ideally with a standard smartphone camera as the payload. The idea behind it: smartphones are also widely available in remote regions such as Namibia, which makes them much easier to use, purchase and maintain locally.

Katja’s special connection to Namibia – where she has family roots and once lived – strongly influenced the project’s focus. Her aim was to find practical solutions that could actually be used on the ground and make a tangible contribution to the protection of endangered animal species.

Currently, wildlife monitoring in Africa is typically conducted using helicopters, aeroplanes or on foot – a process that is not only costly but also time-consuming. The use of drones offers great potential here: they could perform these tasks more efficiently and sustainably. Supported by artificial intelligence (AI), the recorded images could be analysed in order to identify drought zones or animal herds, for example, and take targeted measures.

The ‘Skywalker’ fixed-wing aircraft model was used for the tests. Compared to multicopters, it can be flown over longer distances – a significant advantage when monitoring large areas.

A precise flight mission was planned in advance for the aerial survey. During the test flights in Austria, take-off and landing were still carried out manually by the remote pilot, while the actual surveillance flight was carried out automatically by the autopilot. The long-term plan is to also automate take-off and landing in order to further simplify operations.

The test flights took place in Austria, including at the flight day in Thannhausen, where the team tested the ‘Skywalker’ fixed-wing drone under real conditions for the first time. Particular attention was paid to optimising technical processes – such as the ideal setting of the Pixhawk flight controller, camera integration and safe and stable flight operations.

To evaluate image quality, a commercially available smartphone equipped with a 64 MP telephoto lens was used as the payload. Automated image capture was carried out via a simple app at short time intervals. In order to assess the image sharpness and object identification, the team laid out reference objects on the ground. Tests were carried out at different flight altitudes (80 m, 60 m, 40 m) and at two different flight speeds (20 m/s and 10 m/s) in order to analyse the effects on image quality and flight stability.

The tests showed that the image quality of the smartphone camera was sufficient for simple monitoring tasks. The optimal flight conditions were identified as an altitude of 60 meters and a speed of 10 m/s. At higher speeds, vibrations – presumably caused by natural frequencies – resulted in blurred images. These findings will help to optimise future missions.

Another lesson learnt: manual take-off and landing procedures require practice. With a fully automated system, this effort could be avoided in future.

It was also shown that AI models for detecting wild animals already exist – based on the YOLO framework, for example. For practical use in the field, however, these algorithms would still need to be integrated into an app or software in a user-friendly way.