During my internship I: -developed a tool to query Fotokite's flight database to receive flight-specific logs (Python); -worked on a new feature that detects if the drone's estimation is compromised, and if so, initiates an emergency landing (C++); -added features to Fotokite Log Analyzer Tools (Python); -developed a proof-of-concept IR image stitching, which involved writing framework (first in Python and then in C++) that: --sends a deterministic path to the gimbal composed of different positions; --collects an image at each gimbal position; --re-projects the images in order to stitch them together in the correct positions based on the original gimbal orientation and the camera calibration matrix.

I developed an autonomous bioinspired drone that can fold and access narrow crevices by switching its locomotion mode from flight to terrestrial, and assisted Ph.D. students in their research concerning robots with an environmental impact/goal. For the creation of this platform I: -designed the drone (SolidWorks); -selected its electronic components; -developed the high-level control to: --detect the gap (OpenCV, ROS, C++); --align with respect to the gap (ROS, C++); --estimate the width of the gap and move towards the gap with a given speed dependent on the estimated dimension (OpenCV, ROS, C++); --change locomotion strategy when close enough to the gap (OpenCV, ROS, C++); --return in hover after the narrow passageway has been traversed (OpenCV, ROS, C++).

The Autonomous River Cleanup (ARC) project is a student project at ETH Zurich. The project is led by students and supported by the Robotic Systems Lab. They work together with Ecolymer on the industrial side and are connected with several universities in Europe. The team consists of volunteers, semester project,s and master’s thesis students. The scientific work is supervised by experienced Ph.D. students.