Design of an Industrial Human-Robot System Through Participative Simulations – Tank Cleaning Case Study

Abstract : — Industrials are starting to deploy interactive robots as new solutions to improve workstations. In particular, workstations where human operators may get injured because of repetitive tasks, bad postures or heavy loads are targeted. The introduction of such interactive systems on industrial workstations brings new challenges concerning human cognitive and physical considerations to factories. We followed an approach involving operators, and ergonomics and cognitive engineering skills to improve the introduction of interactive robots in the industry. In this paper, we present the first application of our work on a pyrotechnic tank cleaning workstation. Our approach is illustrated with the design of a solution through several simulation steps involving the workstation's operators. Finally, the design of a prototype based on a teleoperated robot is introduced. I. INTRODUCTION Interactive robotics is being increasingly deployed in the industry, whether through teleoperation, collaborative robots and even exoskeletons [1]. Industrial applications range from material handling and assembly [2], to remote maintenance and inspection [3]. The recent development of the concept of Factory 4.0 increases the industrial demand for these technologies that places humans at the center of automation for better production flexibility. The deployment of interactive robots brings new challenges concerning human cognitive and physical considerations to factories. These human-robot systems also have to satisfy industrial considerations about safety, dependability, productivity, quality and cost. We faced these issues within Safran in 2012 when the first collaborative robot was introduced to improve work conditions in a foundry to manipulate hot parts of various size and weight. Some important portions of operators' work activity were missed at design stage, resulting in a physical interface not adapted to operators' needs during operations. The system deployment on the workstation lacked integration of operators to qualify it on every work situations. As a consequence some parts could not be handled using the system. Moreover, the training phase was too short for the appropriation of the interactive robot by operators. They were disturbed by virtual guides, failing to understand when they were activated or not. As a result operators found the system too constraining, and global productivity decreased on the workstation. In the end, the system was only used to handle the heaviest parts.
Type de document :
Communication dans un congrès
IEEE/SICE International Symposium on System Integration, Dec 2017, Taïwan, China
Liste complète des métadonnées

Littérature citée [32 références]  Voir  Masquer  Télécharger

https://hal.archives-ouvertes.fr/hal-01667727
Contributeur : Jean-Marc Salotti <>
Soumis le : jeudi 21 décembre 2017 - 13:05:33
Dernière modification le : jeudi 7 février 2019 - 15:48:28

Identifiants

  • HAL Id : hal-01667727, version 1

Citation

David Bitonneau, Théo Moulières-Seban, Julie Dumora, Olivier Ly, Jean-François Thibault, et al.. Design of an Industrial Human-Robot System Through Participative Simulations – Tank Cleaning Case Study. IEEE/SICE International Symposium on System Integration, Dec 2017, Taïwan, China. 〈hal-01667727〉

Partager

Métriques

Consultations de la notice

326

Téléchargements de fichiers

227