The development and improvement of new technologies has led to a raise in the use of automated solutions inside industrial settings. In recent years an exponential increase in the use of collaborative robotics has been observed, which allows the automation of industrial process in shared environments, where humans and robots work together. Some of the reasons behind this large diffusion are due to the flexibility and versatility of collaborative manipulators. Indeed, they can be used without the need to install physical barriers, making them suitable for small and medium-sized enterprises. However, large companies still tend to automate their processes with traditional robotic solutions, despite their greater complexity and installation costs. The main reason behind this choice is related to the fact that the performance of collaborative robotics is not yet comparable to that of traditional robotics. This is because, to date, collaborative robotics applications do not fully exploit all the potential offered by this new technology. In particular, there are two key aspects on which it is important to focus: task allocation and safety. In fact, to improve the performances, it is necessary to define an approach that makes the most of the differences between operator and robot, creating a synergy between these two actors. Through collaboration, the human-robot team can perform tasks that neither the operator nor the robot would be able to perform individually. Concerning the safety, the close contact between operator and robot and the absence of physical barriers required to pay a lot of attention to the definition of the methodologies on how to ensure safety. Thus, regulations have been updated to formalise the different degrees of collaboration and to assess safety in each collaborative mode. The mere application of these regulations, however, results in a conservative approach: safety is treated as a barrier in front of which the robot can only slow down until stop. It would be better to manage safety not as a constraint but as a variable in the production process. In this way, it would always be possible to have optimal collaboration while ensuring safety. The work in this thesis aims to address these two issues to change the paradigm of collaborative robotics. In this way, it will no longer be necessary to make a choice between performance and safety, but it will be possible to have both characteristics, bringing collaborative robotics ever closer to traditional robotic solutions. Initially, the context of the work of this thesis and the European project ``RObot enhanced SenSing, INtelligence and actuation to Improve job quality in manufacturing'' (ROSSINI) will be detailed. Subsequently, the two problems in question will be addressed. A first part will focus on the task allocation and scheduling, showing a framework that optimise collaboration and adapt it to the needs of the production process. The following part will focus on trajectory planning, showing solutions that can dynamically manage both safety and robot behaviour. Subsequently, the final result of the ROSSINI project will be presented, i.e. a modular and flexible architecture capable of increasing the performance of collaborative robotics without violating safety regulations. The architecture was implemented and validated on case studies proposed by the project partners, demonstrating how the proposed architecture is so general that it can be used in very different applications and hardware independent. Finally, some extensions of the architecture implemented within the ROSSINI project framework will be presented. These extensions are intended both to improve previously validated solutions and to generate new strategies to support them, making them more complete.

Lo sviluppo ed il miglioramento delle nuove tecnologie hanno portato ad un incremento dell'impiego di soluzioni automatizzate all'interno degli ambienti industriali. Negli ultimi anni si è assistito ad un aumento esponenziale dell'impiego della robotica collaborativa, che permette di automatizzare processi di lavorazione in ambienti di lavoro condivisi. Alcuni dei motivi dietro a questa ampia diffusione sono dovuti alla flessibilità e alla versatilità dei robot collaborativi. Infatti, possono essere utilizzati senza la necessità di installare barriere fisiche, risultando adatti alle piccole e medie imprese. Le grandi imprese, invece, tendono ad automatizzare i propri processi con soluzioni robotiche tradizionali, nonostante la loro maggiore complessità e costi di installazione. Il principale motivo dietro a questa scelta è legato al fatto che le prestazioni della robotica collaborativa non sono ancora paragonabili a quelle della robotica tradizionale. Questo perché, ad oggi, le applicazioni di robotica collaborativa non sfruttano appieno tutte le potenzialità offerte da questa nuova tecnologia. In particolare, ci sono due aspetti chiave su cui è importante focalizzare l'attenzione: suddivisione dei compiti e sicurezza. Per migliorare le prestazioni, è necessario definire un approccio che permetta di sfruttare al meglio le differenze tra operatore e robot creando una sinergia tra questi due agenti. Grazie alla collaborazione, il team uomo-robot può eseguire compiti che né l'operatore né il robot da soli sarebbero in grado di svolgere. Per quanto riguarda la sicurezza, lo stretto contatto tra operatore e robot e l'assenza di barriere fisiche ha richiesto di porre molta attenzione alla definizione di alcune metodologie su come garantire la sicurezza. Pertanto, le normative sono state aggiornate al fine di andare a formalizzare i differenti gradi di collaborazione. La mera applicazione di queste normative si traduce in un approccio molto conservativo: la sicurezza viene trattata come una barriera di fronte alla quale il robot può solo rallentare fino a fermarsi. Sarebbe meglio gestire la sicurezza non come un vincolo ma come una variabile del processo produttivo, in modo da ottenere una collaborazione ottima e sicura. Il lavoro di questa tesi si pone come obiettivo quello di affrontare queste due problematiche al fine di modificare il paradigma della robotica collaborativa. In questo modo, non sarà più necessario effettuare una scelta tra prestazioni e sicurezza, ma sarà possibile avere entrambe le caratteristiche, portando la robotica collaborativa sempre più vicina alle soluzioni robotiche tradizionali. Inizialmente, verranno presentati il contesto in cui si colloca il lavoro di questa tesi ed il progetto europeo "RObot enhanced SenSing, INtelligence and actuation to Improve job quality in manufacturing" (ROSSINI). In seguito, verranno affrontati i due problemi in questione. Una prima parte si concentrerà sulla suddivisione dei compiti, mostrando una soluzione in grado di ottimizzare la collaborazione e di adattarla alle esigenze del processo produttivo. La parte seguente, invece, verterà sulla pianificazione delle traiettorie, mostrando un'architettura in grado di gestire dinamicamente sia la sicurezza che il movimento. In seguito, verrà presentato il risultato finale del progetto ROSSINI: un'architettura modulare e flessibile in grado di incrementare le prestazioni della robotica collaborativa. L'architettura è stata validata sui casi di studio proposti dai partners del progetto, dimostrandone la generalità. Infine, verranno presentate alcune estensioni dell'architettura. Queste estensioni hanno lo scopo sia di migliorare le soluzioni precedentemente convalidate che di generare nuove strategie per renderle più complete.

Tecniche di Controllo Incentrate sulla Sicurezza per una Collaborazione Uomo-Robot Avanzata / Andrea Pupa , 2023 Mar 23. 35. ciclo, Anno Accademico 2021/2022.

Tecniche di Controllo Incentrate sulla Sicurezza per una Collaborazione Uomo-Robot Avanzata

PUPA, ANDREA
2023

Abstract

The development and improvement of new technologies has led to a raise in the use of automated solutions inside industrial settings. In recent years an exponential increase in the use of collaborative robotics has been observed, which allows the automation of industrial process in shared environments, where humans and robots work together. Some of the reasons behind this large diffusion are due to the flexibility and versatility of collaborative manipulators. Indeed, they can be used without the need to install physical barriers, making them suitable for small and medium-sized enterprises. However, large companies still tend to automate their processes with traditional robotic solutions, despite their greater complexity and installation costs. The main reason behind this choice is related to the fact that the performance of collaborative robotics is not yet comparable to that of traditional robotics. This is because, to date, collaborative robotics applications do not fully exploit all the potential offered by this new technology. In particular, there are two key aspects on which it is important to focus: task allocation and safety. In fact, to improve the performances, it is necessary to define an approach that makes the most of the differences between operator and robot, creating a synergy between these two actors. Through collaboration, the human-robot team can perform tasks that neither the operator nor the robot would be able to perform individually. Concerning the safety, the close contact between operator and robot and the absence of physical barriers required to pay a lot of attention to the definition of the methodologies on how to ensure safety. Thus, regulations have been updated to formalise the different degrees of collaboration and to assess safety in each collaborative mode. The mere application of these regulations, however, results in a conservative approach: safety is treated as a barrier in front of which the robot can only slow down until stop. It would be better to manage safety not as a constraint but as a variable in the production process. In this way, it would always be possible to have optimal collaboration while ensuring safety. The work in this thesis aims to address these two issues to change the paradigm of collaborative robotics. In this way, it will no longer be necessary to make a choice between performance and safety, but it will be possible to have both characteristics, bringing collaborative robotics ever closer to traditional robotic solutions. Initially, the context of the work of this thesis and the European project ``RObot enhanced SenSing, INtelligence and actuation to Improve job quality in manufacturing'' (ROSSINI) will be detailed. Subsequently, the two problems in question will be addressed. A first part will focus on the task allocation and scheduling, showing a framework that optimise collaboration and adapt it to the needs of the production process. The following part will focus on trajectory planning, showing solutions that can dynamically manage both safety and robot behaviour. Subsequently, the final result of the ROSSINI project will be presented, i.e. a modular and flexible architecture capable of increasing the performance of collaborative robotics without violating safety regulations. The architecture was implemented and validated on case studies proposed by the project partners, demonstrating how the proposed architecture is so general that it can be used in very different applications and hardware independent. Finally, some extensions of the architecture implemented within the ROSSINI project framework will be presented. These extensions are intended both to improve previously validated solutions and to generate new strategies to support them, making them more complete.
Safety-Aware Control Techniques for Advanced Human-Robot Collaboration
23-mar-2023
SECCHI, Cristian
File in questo prodotto:
File Dimensione Formato  
PhD_Thesis.pdf

embargo fino al 22/03/2026

Descrizione: Tesi definitiva Pupa Andrea
Tipologia: Tesi di dottorato
Dimensione 93.83 MB
Formato Adobe PDF
93.83 MB Adobe PDF   Visualizza/Apri   Richiedi una copia
Pubblicazioni consigliate

Licenza Creative Commons
I metadati presenti in IRIS UNIMORE sono rilasciati con licenza Creative Commons CC0 1.0 Universal, mentre i file delle pubblicazioni sono rilasciati con licenza Attribuzione 4.0 Internazionale (CC BY 4.0), salvo diversa indicazione.
In caso di violazione di copyright, contattare Supporto Iris

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11380/1300912
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus ND
  • ???jsp.display-item.citation.isi??? ND
social impact