The rise of Industry 4.0 is radically modifying the industrial scenario. The fast evolution of production processes, aided by rapid technological development, is leading to a transition from mass production to mass customization. In this framework, the application of collaborative robots, which provide high flexibility and industrial productivity, represents one of the fundamental pillars required by the paradigm shift. In the current work, a programming methodology based on Skill-Based Programming is developed, in order to reduce the programming effort and to preserve the criteria of generality, reusability, robustness and intuitiveness, even with a limited hardware availability. The architecture of the adopted solution is based on a macro-differentiation between Action Skill and Check Skill, proposing a trade-off between force control and classical position control of the manipulators, to guarantee an adaptive motion also in the most critical phases. The proposed strategy is validated and applied to a complex real case study: the automation of an assembly operation consisting in the piston-jaw-lever coupling, which is part of the assembly process of different types of pneumatic grippers manufactured by Camozzi. In the work, a library of Primitives/Skills pre-implemented with a low-level robotics programming language is presented, from which a State Machine capable of carrying out the operation is composed. The design of flexible robotic fingers, able of adapting to different gripper models and ensuring the correct execution of the aforementioned assembly, is also described. The proposed methodology is validated through experiments performed at the MERLIN laboratory of Politecnico di Milano, with the use of a UR5e collaborative robot. The results obtained were compared with those derived from classical robotic programming methods.
L’avvento dell’industria 4.0 sta modificando radicalmente l'assetto industriale. La rapida evoluzione dei processi produttivi, coadiuvata da un veloce sviluppo tecnologico, comporta uno spostamento dalla produzione di massa verso la personalizzazione di massa. In questo contesto, l'applicazione di robot collaborativi, i quali garantiscono un'elevata flessibilità e produttività industriale, rappresenta uno dei pilastri fondamentali richiesti dal cambio di paradigma. Nel presente lavoro è sviluppata una metodologia di programmazione basata sulla Skill-Based Programming, in grado di alleggerire lo sforzo di programmazione e di preservarne i criteri di generalità, riusabilità, robustezza e intuitività, pur con una disponibilità hardware limitata. L'architettura della soluzione adottata è basata su una macro-differenziazione tra Action Skill e Check Skill, proponendo un compromesso tra il controllo di forza ed il classico controllo di posizione dei manipolatori, per garantire un movimento adattivo anche nelle fasi più critiche. La strategia proposta è validata ed applicata ad un complesso caso di studio reale: l’automatizzazione di un'operazione di assemblaggio consistente nell’accoppiamento pistone-griffa-leva e facente parte del processo di assemblaggio di diverse tipologie di pinze pneumatiche di produzione Camozzi. Nell'elaborato viene proposta una libreria di Primitive/Skills pre-implementate con un linguaggio di programmazione robotica di basso livello, partendo dalla quale viene composta una Macchina a Stati in grado di portare a compimento l'operazione. Viene inoltre descritta la progettazione di dita robotiche flessibili, in grado di adattarsi a diversi modelli di pinze e di garantire la corretta esecuzione del suddetto assemblaggio. La metodologia proposta è validata mediante esperimenti condotti presso il laboratorio MERLIN del Politecnico di Milano, attraverso l'utilizzo di un robot collaborativo UR5e. I risultati ottenuti sono confrontati con quelli derivanti dai classici metodi di programmazione robotica.
Robust and flexible skill-based programming for robotic assembly tasks
VILLANOVA, ANDREA;ZAMPELLI, GABRIELE
2019/2020
Abstract
The rise of Industry 4.0 is radically modifying the industrial scenario. The fast evolution of production processes, aided by rapid technological development, is leading to a transition from mass production to mass customization. In this framework, the application of collaborative robots, which provide high flexibility and industrial productivity, represents one of the fundamental pillars required by the paradigm shift. In the current work, a programming methodology based on Skill-Based Programming is developed, in order to reduce the programming effort and to preserve the criteria of generality, reusability, robustness and intuitiveness, even with a limited hardware availability. The architecture of the adopted solution is based on a macro-differentiation between Action Skill and Check Skill, proposing a trade-off between force control and classical position control of the manipulators, to guarantee an adaptive motion also in the most critical phases. The proposed strategy is validated and applied to a complex real case study: the automation of an assembly operation consisting in the piston-jaw-lever coupling, which is part of the assembly process of different types of pneumatic grippers manufactured by Camozzi. In the work, a library of Primitives/Skills pre-implemented with a low-level robotics programming language is presented, from which a State Machine capable of carrying out the operation is composed. The design of flexible robotic fingers, able of adapting to different gripper models and ensuring the correct execution of the aforementioned assembly, is also described. The proposed methodology is validated through experiments performed at the MERLIN laboratory of Politecnico di Milano, with the use of a UR5e collaborative robot. The results obtained were compared with those derived from classical robotic programming methods.| File | Dimensione | Formato | |
|---|---|---|---|
|
2021_06_Villanova_Zampelli.pdf
non accessibile
Descrizione: Robust and flexible Skill-Based Programming for robotic assembly tasks
Dimensione
72.29 MB
Formato
Adobe PDF
|
72.29 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/175021