Tiozzo Fasiolo, Diego Autonomous Mobile Robotics For Mapping and Monitoring in Agriculture. [Tesi di dottorato]

	Documento PDF diego_tiozzofasiolo_37_COMPLETO.pdf Visibile a [TBR] Amministratori dell'archivio Download (97MB) | Richiedi una copia
	Documento PDF diego_tiozzofasiolo_37_PARZIALE.pdf Visibile a [TBR] Amministratori dell'archivio Download (52MB) | Richiedi una copia

Abstract

Autonomous mobile robots have gained increasing attention for their potential in agricultural applications, particularly in vineyard monitoring and mapping. In contrast to aerial vehicles, now a standard in precision agriculture, ground mobile robots can acquire data at much closer ranges and, when equipped with a robotic arm, can also interact directly with plants for more detailed measurements. This PhD thesis aims to develop autonomous navigation and 3D mapping systems for mobile robots in agriculture, relying on inputs from multiple sensors. A comprehensive literature review of state-of-the-art technologies supporting robotic mapping in agriculture is first conducted, followed by an experimental evaluation of Simultaneous Localization and Mapping (SLAM) algorithms for 3D reconstruction of both indoor and outdoor environments. These SLAM algorithms are tested on multiple mobile robots and also adapted to crowded environments using artificial intelligence. Then, this PhD thesis proposes a mobile robot capable of safely navigating and scanning a vineyard, retrieving geometric and multispectral data on the plants through multi-sensor fusion techniques. The results of experiments conducted in the vineyard of the University of Udine demonstrate the feasibility of the approach and contribute to the field of agricultural mobile robotics through the release of an open-source dataset. Moreover, the PhD thesis presents a perception framework, developed during a period as a visiting PhD student at the Stevens Institute of Technology, Hoboken (NJ, USA), which enhances object pose estimation for mobile manipulators and is validated through both simulations and real-world experiments. In conclusion, this work advances the field of autonomous navigation and mapping for mobile robotics, especially in the context of precision agriculture. The contributions include the application of innovative solutions for autonomous navigation, data fusion, and pose estimation to ground mobile robots designed for agricultural mapping and monitoring, establishing a robust foundation for more efficient robotics applications in precision agriculture.

Downloads

Actions (login required)

Modifica documento