Skip to Main content Skip to Navigation
Theses

Improving Motor Coordination in Human-Robot Interactions using Bio-inspired Controllers

Melanie Jouaiti 1
1 LARSEN - Lifelong Autonomy and interaction skills for Robots in a Sensing ENvironment
Inria Nancy - Grand Est, LORIA - AIS - Department of Complex Systems, Artificial Intelligence & Robotics
Abstract : Gestural communication is an important aspect of HRI in social, assistance and rehabilitation robotics. Indeed, social synchrony is a key component of interpersonal interactions which affects the interaction on the behavioural level, as well as on the social level. It is paramount for the robot to be able to adapt to its interaction partner. In rhythmic social interactions, humans experience two physical phenomena which can also be observed in oscillators: the magnet effect which entrains both systems until they are coupled and synchronised; the maintenance effect which is the struggle of each system to conserve its own intrinsic frequency. These mechanisms could play a fundamental role in physical and social interpersonal interactions. To reproduce this behaviour, bio-inspired controllers endowed with plasticity mechanisms can be employed. The main goal consists in making these interactions as natural and enjoyable as possible by integrating adaptive properties, which leads to the emergence of motor coordination and hence social synchrony. A non-negligible part of this research also consists in studying humans in HRI to understand human behaviour better and design better interactions. In the first part of this thesis, we will introduce synchrony in interpersonal interactions but also in human-robot interactions. We will also observe and endeavour to understand human behaviour in rhythmic human-robot interactions. The second part will present the bio-inspired controller. We will integrate some new plasticity mechanisms, develop its discrete functioning mode and extend the model so that it can adapt to any rhythmic or discrete movement. We also analyse the results of a comparison study with other oscillators and controllers to highlight the capabilities of this model. Moreover, we will validate the controller experimentally in two user studies on human-robot interactions with and without contact. This chapter will evaluate several aspects, such as robot power consumption, controller learning rate, human muscular effort, user perception, coordination performance and engagement. Finally, the third part will focus on the applications and perspectives of this thesis in robot- assisted therapy for children with motor deficits, notably with autistic children.
Complete list of metadata

Cited literature [425 references]  Display  Hide  Download

https://hal.univ-lorraine.fr/tel-02929729
Contributor : Thèses Ul <>
Submitted on : Thursday, September 3, 2020 - 4:56:23 PM
Last modification on : Tuesday, July 6, 2021 - 5:25:50 PM
Long-term archiving on: : Wednesday, December 2, 2020 - 7:22:17 PM

File

DDOC_T_2020_0035_JOUAITI.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : tel-02929729, version 1

Citation

Melanie Jouaiti. Improving Motor Coordination in Human-Robot Interactions using Bio-inspired Controllers. Computer Science [cs]. Université de Lorraine, 2020. English. ⟨NNT : 2020LORR0035⟩. ⟨tel-02929729⟩

Share

Metrics

Record views

168

Files downloads

197