A carregar...
Publicação
Self Localization and Mapping using a Magabot
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 20.35 MB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
Autonomous navigation is becoming an increasingly relevant topic in industry due to the rise
in autonomous cars and factory robots. Companies invest in complex sensor setups, like LiDARs,
to achieve excellent autonomous navigation results. These setups can be costly, making them
less affordable for educational purposes. In this dissertation an IDMind Magabot with five sonar
sensors and a laptop camera are used to replace the LiDAR. This dissertation method offers an
affordable setup that can be easily implemented in educational settings, providing students with
a hands-on experience in autonomous navigation technologies.
We can create a simple and affordable hazard detection and map-making setup using a
laptop camera and sonar sensors. Sonar sensors measure distances, while the camera provides
information about the robot’s relative position within its environment, enabling the creation of
corridor maps.
We tested this mapping method in the corridors of the C6 building, 3rd floor, at FCUL.
We evaluated the shapes and distances of the maps created using this method. This method
produced mixed results, while the distances on the map were good, with an error smaller than
5%, the shapes of the maps were not accurate with reality. Despite these results, this method
can work in an educational setting with enough refining.
Descrição
Tese de mestrado, Engenharia Física, 2025, Universidade de Lisboa, Faculdade de Ciências
Palavras-chave
Navegação autónoma Configurações de sensores deteção de perigos criação de mapas vii Teses de mestrado - 2025