Skip to Main content Skip to Navigation
Theses

Au coeur de l’interaction humain-robot collaboratif : comment concevoir une assistance personnalisée au profil utilisateur ?

Katleen Blanchet 1, 2, 3, 4, 5
3 ACMES-SAMOVAR - Algorithmes, Composants, Modèles Et Services pour l'informatique répartie
SAMOVAR - Services répartis, Architectures, MOdélisation, Validation, Administration des Réseaux
4 LRI - Laboratoire de Robotique Interactive
DIASI - Département Intelligence Ambiante et Systèmes Interactifs : DRT/LIST/DIASI
Abstract : The transformation of production plants is accelerating, driven by advances in collaborative robotics and data science. As a result, the organisation of work is changing, directly affecting the working conditions of operators. Loss of autonomy, information overload, increased pace, operators have to change their habits and learn to collaborate with the robot. In this context, the aim of this research work is to improve the operators quality of life at work, while performing a physical collaborative task, by means of user profile-based assistance. In the literature, the assistance mainly relies on external observation devices, causes of stress, and proposes exclusively a priori-based adjustments of the robot's behaviour. Thus, these assistance do not dynamically adapt to human behaviour variations. In order to overcome these challenges, this study presents two contributions. Firstly, we propose a methodology for extracting high-level information on the user profile from the robot raw signals, which is applied to expertise. We then introduce a hybrid approach to profile-based assistance which combines human-centered reinforcement learning and symbolic logic (ontology and reasoning) to guide operators towards skill improvement. This synergy guarantees online adaptation to user needs while reducing the learning process. Then, we extend the robotic assistance with informative assistance. We have demonstrated, through simulation and experiments in real conditions on three robotic usecases, the consistency of our profile as well as the positive effect of the assistance on the skills acquisition. We thereby create a more favourable environment for professional satisfaction by reducing the mental workload.
Complete list of metadata

https://tel.archives-ouvertes.fr/tel-03145245
Contributor : Abes Star :  Contact
Submitted on : Thursday, February 18, 2021 - 11:14:11 AM
Last modification on : Saturday, February 20, 2021 - 3:29:23 AM

File

84568_BLANCHET_2021.pdf
Version validated by the jury (STAR)

Identifiers

  • HAL Id : tel-03145245, version 1

Citation

Katleen Blanchet. Au coeur de l’interaction humain-robot collaboratif : comment concevoir une assistance personnalisée au profil utilisateur ?. Interface homme-machine [cs.HC]. Institut Polytechnique de Paris, 2021. Français. ⟨NNT : 2021IPPAS001⟩. ⟨tel-03145245⟩

Share

Metrics

Record views

189

Files downloads

65