The evolution of the industrial production processes, which is guided by the emerging paradigm of Industry 4.0, calls for new forms of interaction between humans and machines. In this context, the introduction of collaborative robotics represents one of the pillars of the fourth industrial revolution. Within this framework, this work presents a safe motion control system for industrial manipulators, designed to maximize the robot productivity and to meet the safety requirements of the operator. The use of an RGB-D camera and an algorithm able to predict the kinematic configuration of the human worker, allows to obtain a real-time estimate of the future space occupied by the operator. This estimation is represented by a series of convex swept-volumes considered by the control system as obstacles to be avoided. The real-time solution of an optimization problem allows to control the robot's motion in accordance with pre-programmed trajectories, respecting both the kinematic constraints and those imposed by the adopted collision avoidance strategy. Subsequently, a safety perception system is presented. It aims to increase the operator's psychological safety and prevent potentially dangerous movements, through the digital visualization of the future robot motion. The implementation of a software application for Microsoft Hololens and the development of a real-time motion prediction algorithm provide the operator with a continuous display of the robot's motion intentions on the basis of updated data. Finally, a control strategy based on the human's field of view is presented. It aims at increasing the safety perceived by the operator through the real-time modification of the trajectory performed by the robot, such that the end-effector tries to remain as close as possible to the center of the user's field of view during the desired robot operation. An approximation of the human's field of view is obtained thanks to the use of Microsoft Hololens. The proposed systems have been validated through simulated experiments on the industrial manipulator ABB IRB 140 and the obtained results are reported and described.
I cambiamenti nei processi produttivi industriali, guidati dal paradigma emergente di Industria 4.0, rendono necessarie nuove forme di interazione tra esseri umani e macchine. In tale contesto, l'introduzione di una robotica collaborativa rappresenta uno dei pilasti della quarta rivoluzione industriale. In questo quadro, questo lavoro presenta un sistema di controllo del moto per manipolatori industriali, concepito per massimizzarne la produttività e per rispettare i requisiti di sicurezza dell'operatore. L'impiego di una RGB-D camera e di un algoritmo di previsione della configurazione cinematica dell'essere umano, permette di ottenere una stima in tempo reale dello spazio futuro occupato dall'operatore. Tale stima è rappresentata da una serie di swept-volumes, considerati dal sistema di controllo come ostacoli da evitare. La soluzione in tempo reale di un problema di ottimizzazione permette di controllare il moto del robot in accordo con traiettorie pre-programmate, rispettando sia i vincoli cinematici che quelli imposti dalla strategia di anticollisione adottata. Successivamente, viene presentato un sistema di percezione della sicurezza atto ad incrementare la sicurezza psicologica dell'operatore e prevenire movimenti potenzialmente pericolosi, attraverso la visualizzazione digitale del moto futuro del robot. L'implementazione di un'applicazione per Microsoft Hololens e lo sviluppo di un algoritmo real-time di previsione del moto forniscono all'operatore una visualizzazione continua delle intenzioni di moto del robot sulla base di dati aggiornati. Infine viene presentata una strategia di controllo basata sul campo visivo dell'essere umano, finalizzata ad incrementare la sicurezza percepita dall'operatore attraverso la modifica in tempo reale della traiettoria eseguita dal robot, tale per cui l'end-effector rimanga il più possibile al centro del campo visivo dell'utente durante l'esecuzione dell'operazione. Un'approssimazione del campo visivo dell'essere umano è ottenuta grazie all'utilizzo di Microsoft Hololens. I sistemi proposti sono stati validati attraverso esperimenti simulati sul manipolatore industriale ABB IRB 140, i cui risultati sono riportati.
Physical and psychological safety in human-robot collaboration
MAIOCCHI, MARCO
2018/2019
Abstract
The evolution of the industrial production processes, which is guided by the emerging paradigm of Industry 4.0, calls for new forms of interaction between humans and machines. In this context, the introduction of collaborative robotics represents one of the pillars of the fourth industrial revolution. Within this framework, this work presents a safe motion control system for industrial manipulators, designed to maximize the robot productivity and to meet the safety requirements of the operator. The use of an RGB-D camera and an algorithm able to predict the kinematic configuration of the human worker, allows to obtain a real-time estimate of the future space occupied by the operator. This estimation is represented by a series of convex swept-volumes considered by the control system as obstacles to be avoided. The real-time solution of an optimization problem allows to control the robot's motion in accordance with pre-programmed trajectories, respecting both the kinematic constraints and those imposed by the adopted collision avoidance strategy. Subsequently, a safety perception system is presented. It aims to increase the operator's psychological safety and prevent potentially dangerous movements, through the digital visualization of the future robot motion. The implementation of a software application for Microsoft Hololens and the development of a real-time motion prediction algorithm provide the operator with a continuous display of the robot's motion intentions on the basis of updated data. Finally, a control strategy based on the human's field of view is presented. It aims at increasing the safety perceived by the operator through the real-time modification of the trajectory performed by the robot, such that the end-effector tries to remain as close as possible to the center of the user's field of view during the desired robot operation. An approximation of the human's field of view is obtained thanks to the use of Microsoft Hololens. The proposed systems have been validated through simulated experiments on the industrial manipulator ABB IRB 140 and the obtained results are reported and described.File | Dimensione | Formato | |
---|---|---|---|
2020_06_Maiocchi.pdf
Open Access dal 18/05/2021
Descrizione: Testo della tesi
Dimensione
35.39 MB
Formato
Adobe PDF
|
35.39 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/10589/154370