Description du sujet de thèse

1. Context and objectives
In order to carry out most of the complex tasks of the robotics of the future, a continuous physical interaction between the human and the robot, commonly called pHRI (physical Human-Robot Interaction) [1], [2], is necessary. This type of robotics can concern the co-manipulation of objects between the human and the robot [3], [4], assistance and rehabilitation robotics [5], e-learning robotics based on human gestures [6] or active exoskeletons [7]. The aim of this thesis is therefore to understand human-human collaboration strategies in order to develop a coherent robotic collaboration strategy. Indeed, when we are looking to generate object co-manipulation tasks between a human and a robot in a way that is both safe and efficient, one promising approach is to draw inspiration from human-human co-ordination strategies and apply them in a human-robot collaborative context. The hypothesis here is that human-robot coordination will be improved by taking into account human-human collaboration techniques, which appear to be fluid and intuitive [8].
The originality of this thesis lies in the fact that it studies the human-human coordination strategy with a view to adapting it to robotic control. This represents a major challenge in terms of experiments and simulations that should make it possible to improve control and anticipate the behaviour of collaborative robots by making them more proactive [9] during collaboration. This thesis is funded by the Agence Nationale de la Recherche (CALL project: Cobotic AppLication for handover and Load transport). It forms part of the biomechanics activities of the Gepetto team at LAAS-CNRS. It is interdisciplinary in nature, and will be carried out in close collaboration with researchers in collaborative robotics (cobotics), vision and data sciences.
Ultimately, the project will help to improve the handover of an object and its co-manipulation with a robotic system.
2. Stages of the thesis
In order to achieve the objectives presented above, the proposed thesis will have to carry out at least the following stages:
a. State of the art of work on human-human physical interaction in general and particularly that aimed at bio-inspired collaborative robotic control strategies
b. Setting up biomechanical human-human load transfer and co-manipulation experiments
c. Studying and extracting human-human coordination information
d. Simulation of human behaviour
3. Bibliography
[1] A. De Santis, B. Siciliano, A. De Luca, et A. Bicchi, « An atlas of physical human – robot interaction », Mechanism and Machine Theory, vol. 43, p. 253 270, 2008, doi: 10.1016/j.mechmachtheory.2007.03.003.
[2] B. Navarro, A. Fonte, P. Fraisse, G. Poisson, et A. Cherubini, « In Pursuit of Safety: An Open-Source Library for Physical Human-Robot Interaction », IEEE Robotics & Automation Magazine, vol. 25, no 2, p. 39 50, 2018, doi: 10.1109/MRA.2018.2810098.
[3] L. Peternel, N. Tsagarakis, et A. Ajoudani, « Towards multi-modal intention interfaces for human-robot co-manipulation », in 2016 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), oct. 2016, p. 2663 2669. doi: 10.1109/IROS.2016.7759414.
[4] M. Mujica, M. Crespo, M. Benoussaad, S. Junco, et J.-Y. Fourquet, « Robust variable admittance control for human–robot co-manipulation of objects with unknown load », Robotics and Computer-Integrated Manufacturing, vol. 79, p. 102408, 2023, doi: https://doi.org/10.1016/j.rcim.2022.102408.
[5] C. Lauretti, F. Cordella, E. Guglielmelli, et L. Zollo, « Learning by Demonstration for Planning Activities of Daily Living in Rehabilitation and Assistive Robotics », IEEE ROBOTICS AND AUTOMATION LETTERS, vol. 2, no 3, p. 1375 1382, 2017.
[6] S. Ikemoto, H. Ben Amor, T. Minato, B. Jung, et H. Ishiguro, « Physical human-robot interaction: Mutual learning and adaptation », IEEE Robotics and Automation Magazine, vol. 19, no 4, p. 24 35, 2012, doi: 10.1109/MRA.2011.2181676.
[7] H. D. Lee, B. K. Lee, W. S. Kim, J. S. Han, K. S. Shin, et C. S. Han, « Human-robot cooperation control based on a dynamic model of an upper limb exoskeleton for human power amplification », Mechatronics, vol. 24, no 2, p. 168 176, 2014, doi: 10.1016/j.mechatronics.2014.01.007.
[8] I. Maroger, M. Silva, H. Pillet, N. Turpin, O. Stasse, et B. Watier, « Walking paths during collaborative carriages do not follow the simple rules observed in the locomotion of single walking subjects », Scientific Reports, vol. 12, no 1, p. 15585, 2022, doi: 10.1038/s41598-022-19853-7.
[9] J. Xia, D. Huang, Y. Li, et N. Qin, « Iterative learning of human partner's desired trajectory for proactive human–robot collaboration », International Journal of Intelligent Robotics and Applications, vol. 4, no 2, p. 229 242, 2020, doi: 10.1007/s41315-020-00132-5.

Contexte de travail

The thesis will be carried out mainly in Toulouse within the Gepetto team at LAAS-CNRS. It will be funded under the ANR CALL programme. GEPETTO is a LAAS-CNRS team specialising in the study of anthropomorphic movements, and has extensive experience in the generation of humanoid robot movements. Based in Toulouse, the GEPETTO team is developing control software suites for the HRP-2, Talos and H1 biped robots. The team is also interested in human movement and its simulation. It has a complete technical platform for movement analysis located at the Toulouse Midi-Pyrénées CREPS. The thesis will be co-supervised by Hélène Pillet from the Institut de Biomécanique Humaine Georges Charpak at Arts et Métiers ParisTech (IBHGC).
The IBHGC has recognised expertise in personalised modelling of the locomotor system. In particular, the Institute has been applying its research into the dynamic analysis of human movement for over ten years. This research has focused on defining specific protocols adapted to the study of particular situations in the laboratory, and on modelling the kinematics and dynamics of segments in healthy or pathological subjects. To carry out these analyses, the Institute uses technical resources such as 3D movement analysis systems, inertial units and force platforms. For the purposes of the ANR, it is possible to travel to the University of Sherbrooke (Canada), the ENI in Tarbes or the JRL (AIST-CNRS, Tsukuba, Japan).

Le poste se situe dans un secteur relevant de la protection du potentiel scientifique et technique (PPST), et nécessite donc, conformément à la réglementation, que votre arrivée soit autorisée par l'autorité compétente du MESR.

Contraintes et risques

Experiments on human subjects under ethical supervision are planned as part of this work.