In the last decades, micro aerial multi-rotors have proven themselves suitable for a wide variety of applications, such as search and rescue, crop monitoring in agriculture, and load transportation in logistics or constructions, just to name a few.
Several commercial aerial robots are available on the market, which, however, are only partly programmable and/or limited to indoor flight. Among the main challenges when transitioning from an indoor to an outdoor setup are non-controlled environmental conditions and robot localization, which cannot rely on motion-capture systems anymore but is typically achieved using GPS-based and/or vision-based methods.
The main question that this thesis will answer is what the influence is of different environmental conditions (wind, flying over water, in the woods, etc.) on the vision-based outdoor localization of an unmanned aerial robot.
The overarching question will be achieved through the following steps:
- Building a quadrotor from a hardware and software point of view, using the open-hardware and open-source software released at https://github.com/uzh-rpg/agilicious, under the supervision of an expert in the domain
- Extend the software and hardware package in the previous point by installing a GPS receiver on the robot and coding a GPS driver that uses the data coming from the GPS for localization.
- Carrying out outdoor flight tests with the built quadrotor using GPS-based localization
- Carrying out outdoor flight tests with the built quadrotor using vision-based localization in multiple environmental conditions and comparing the results among them and with those obtained in the previous point.
- Philipp Foehn, Elia Kaufmann, Angel Romero, Robert Penicka, Sihao Sun, Leonard Bauersfeld, Thomas Laengle, Giovanni Cioffi, Yunlong Song, Antonio Loquercio, Davide Scaramuzza, "Agilicious: Open-Source and Open-Hardware Agile Quadrotor for Vision-Based Flight", AAAS Science Robotics, 2022, https://rpg.ifi.uzh.ch/docs/ScienceRobotics22_Foehn.pdf