Development of an automatic guiding system for builder drones

Abstract

Builder drones have the potential of aiding in construction sites by picking up and depositing masonry units. However, their accuracy is restrained due to drifting, wind and the uncertainty of positioning methods such as GPS. This thesis explores the feasibility of a vision-based guidance system for builder drones. In particular, it examines the reliability of ArUco markers as a local positioning method, as well as different control approaches to move the drone towards them. Finally, it creates a framework capable of executing a construction mission based on searching a specific brick and depositing it on its correct position. Furthermore, complementary systems are developed to help future iterations of the project.