Drone: Dynamic positioning in relation to a given object

Abstract

This report deals with the bachelor thesis "Drone: Dynamic positioning in relation to a given object", given by Rolls-Royce Marine and NTNU Ålesund. The goal of this task is to develop a framework for a drone autopilot and finding usage scenarios of drones in maritime industry. Main focus for this task is dynamic positioning in relation to an object in movement og autonomous landing and autonomous take-off.

The result is an autopilot that has three main modes. The first mode is supposed to be a simple takeoff mode, this will initialize, arm and move the drone to a given altitude. The second mode is the balloon finder mode, this will look for a target (represented by balloon) the pivot towards the target and maintain a set distance. The third mode is a simple landing mode, this will bring the drone back to its starting position while using the SIFT algorithm to recognise to find the correct landing spot which is qr-code.

The report has the needed background theory, methods used, the results that have been achieved, and a discussion of the methods and results and at last a conclusion. In appendix you also find source code for the autopilot, meeting reports, progress reports, laws and regulation for drones by lovdata and the pre-project report.