The growing use of Autonomous Mobile Robots (AMRs) in industrial environments is revolutionising the automation of complex tasks. Among the many future uses, large-scale additive construction represents a particularly promising application scenario, in which fleets of robots are required to transport materials and tools in a coordinated, autonomous, and scalable manner. This thesis investigates a Leader-Follower approach as a potential solution for such an environment, seeking to provide an open-source and reproducible prototype model for real-world deployment. While the theoretical foundations of AMR coordination have been well studied, existing literature falls short of offering practical guidelines for implementation within industrial environments where ease of operation and system robustness are essential. This work aims to bridge this gap by conceptualising, implementing, and experimentally testing a decentralised control framework for AMRs using the Robot Operating System (ROS) with simplicity, modularity, and reproducibility as guiding principles. The system was validated in simulated and real environments. Simulations allowed for sequential calibration of control parameters and logic, while experiments conducted in the real world demonstrated that the follower robot attained correct spatial alignment with the leader even with abrupt changes in direction and variation in speed. The results showed consistent tracking performance and stability, validating the proposed architecture. The system design allows for easy use by inexperienced operators, offering a functional link between theoretical models and practical application. In addition, this research revealed possible avenues for further improvement, such as the need for quantitative performance metrics, increased modularity, and better parameter adaptation. In conclusion, this thesis provides an experimentally guide to the use of Leader-Follower control in AMRs, enabling further advances in Autonomous Mobile Robots in industrial distribution and manufacturing processes.
Il crescente utilizzo di Robot Mobili Autonomi (AMR) negli ambienti industriali sta rivoluzionando l'automazione di compiti complessi. Tra i possibili utilizzi futuri, la costruzione additiva su larga scala rappresenta uno scenario applicativo particolarmente promettente, in cui flotte di robot devono trasportare materiali e strumenti in modo coordinato, autonomo e scalabile. Questa tesi studia un approccio Leader-Follower come potenziale soluzione per un ambiente di questo tipo, cercando di fornire un prototipo di modello open source e riproducibile per l'impiego nel mondo reale. Mentre le basi teoriche del coordinamento degli AMR sono state ben studiate, la letteratura esistente non offre linee guida pratiche per l'implementazione in ambienti industriali in cui la facilità di funzionamento e la robustezza del sistema sono essenziali. Questo lavoro ha cercato di colmare questa lacuna concettualizzando, implementando e testando sperimentalmente un framework di controllo decentralizzato per AMR utilizzando il Robot Operating System (ROS) con semplicità, modularità e riproducibilità come principi guida. Il sistema è stato convalidato in ambienti simulati e reali. Le simulazioni hanno consentito la calibrazione sequenziale dei parametri e della logica di controllo, mentre gli esperimenti condotti nel mondo reale hanno dimostrato che il robot follower raggiungesse un corretto allineamento spaziale con il leader anche con bruschi cambiamenti di direzione e variazioni di velocità. I risultati hanno evidenziato prestazioni di inseguimento e stabilità costanti, e convalidando l'architettura proposta. Il design del sistema consente un facile utilizzo da parte di operatori inesperti, offrendo un collegamento funzionale tra i modelli teorici e le applicazioni pratica. Inoltre, questa ricerca ha rivelato possibili vie di ulteriore miglioramento, come la necessità di metriche quantitative delle prestazioni, maggiore modularità ed un migliore adattamento dei parametri. In conclusione, questa tesi fornisce una guida sperimentale all'uso del controllo Leader-Follower negli AMR, consentendo ulteriori progressi nella Robotica Mobile Autonoma nei processi di distribuzione e produzione industriale.
Progettazione e Valutazione di una Strategia Leader-Follower per il Controllo di Robot Mobili Autonomi
STARACE, CHRISTIAN
2024/2025
Abstract
The growing use of Autonomous Mobile Robots (AMRs) in industrial environments is revolutionising the automation of complex tasks. Among the many future uses, large-scale additive construction represents a particularly promising application scenario, in which fleets of robots are required to transport materials and tools in a coordinated, autonomous, and scalable manner. This thesis investigates a Leader-Follower approach as a potential solution for such an environment, seeking to provide an open-source and reproducible prototype model for real-world deployment. While the theoretical foundations of AMR coordination have been well studied, existing literature falls short of offering practical guidelines for implementation within industrial environments where ease of operation and system robustness are essential. This work aims to bridge this gap by conceptualising, implementing, and experimentally testing a decentralised control framework for AMRs using the Robot Operating System (ROS) with simplicity, modularity, and reproducibility as guiding principles. The system was validated in simulated and real environments. Simulations allowed for sequential calibration of control parameters and logic, while experiments conducted in the real world demonstrated that the follower robot attained correct spatial alignment with the leader even with abrupt changes in direction and variation in speed. The results showed consistent tracking performance and stability, validating the proposed architecture. The system design allows for easy use by inexperienced operators, offering a functional link between theoretical models and practical application. In addition, this research revealed possible avenues for further improvement, such as the need for quantitative performance metrics, increased modularity, and better parameter adaptation. In conclusion, this thesis provides an experimentally guide to the use of Leader-Follower control in AMRs, enabling further advances in Autonomous Mobile Robots in industrial distribution and manufacturing processes.File | Dimensione | Formato | |
---|---|---|---|
Tesi Christian Starace.pdf
non disponibili
Dimensione
1.49 MB
Formato
Adobe PDF
|
1.49 MB | Adobe PDF |
I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/20.500.12075/22729