Interaction control in human, for robot and with robot

This is a tutorial organized for 2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2021). Since IROS will be held virtually this year, this tutorial will be online.

Main Organizer

Yanan Li, University of Sussex, UK, email: yl557@sussex.ac.uk

Co-organizers

Atsushi Takagi, NTT Communication Science Laboratories, Japan, email: atsushi.takagi.yx@hco.ntt.co.jp

Jonathan Eden, Imperial College of Science, Technology and Medicine, London UK. email: j.eden@ic.ac.uk

Etienne Burdet, Imperial College of Science, Technology and Medicine, London UK. email: eburdet@ic.ac.uk

Goal

The last decades have seen tremendous developments in robot sensing supported by accurate and affordable wearable sensors, however, robots are primarily conceived to interact and exchange energy with their environment. Despite extensive studies over several decades examining the force control of a robot in contact with its environment (such as hybrid force position control [1]), in many aspects, robots are far from approaching human interaction capabilities. Additionally, as robots are increasingly used in contact with human operators (in collaborative robots for manufacturing, rehabilitation robots and robotic physical trainers, etc.), existing control techniques must be modified to ensure safe and skillful interaction.

Arguably, understanding human motor control may help us develop more skillful robot controllers to interact with unknown environments and humans. In particular, impedance control [2] has appeared as a theory to both explain human motor behaviour (with the experimental evidence we provided in [3]), and develop robotic interaction control. Using an approach inspired by the results of [2], we could develop a biomimetic adaptive impedance controller enabling a robot to compensate for the force and impedance arising from the interaction with a rigid or soft environment [4,5].

In recent years we have observed that interacting humans exchange haptic information [6], and have deciphered the underlying control mechanism of haptic communication [7,8,9]. We have started exploiting this mechanism for human-robot interaction. Through haptic communication, a robot may understand the motion intention of the human operator and use this sensory augmentation to better assist them [7,10]. Game theory of sensorimotor interaction may also support the robot to best complement the human operator according to a suitable role from assistance to education and competition [11,12].

Based on the related literature including above results, the tutorial will give attendees a presentation in the contemporary computational neuroscience and robotics aspects of human-robot interaction, necessary for developing efficient robots to work in physical contact with their environment and humans. The target audience includes researchers and postgraduate students who are interested in these topics or who are developing robots interacting with humans in applications ranging from physical training (for sport and neurorehabilitation), to shared driving and cobots for manufacturing. It will also be useful for psychologists who want to learn the robotic modelling for this fundamental aspect of human motor control.

Tutorial outline

In this tutorial, we will firstly introduce the background of how humans interact with their environment, and how generalised adaptive impedance control for robots can be implemented. We will then present recent findings of the haptic information exchange between interacting humans, and establish computational models to explain the underlying control mechanism of haptic communication. Finally, we will elaborate on how we use this knowledge to develop robotic interaction control strategies. The tutorial will be structured in six lectures as follows:

1. Motor control in human-environment interaction, by Etienne Burdet

2. Controller design for robot-environment interaction, by Yanan Li

3. Haptic communication between humans, by Atsushi Takagi

4. Sensory augmentation in human-robot interaction, by Jonathan Eden

5. Continuous role adaptation and effort sharing in human-robot interaction, by Yanan Li

6. Towards applications in human-robot interaction, by Atsushi Takagi and Jonathan Eden

Prerequisite knowledge

The target audience would benefit from knowledge of robot control and signal processing. However, the course will be designed in a self-contained way so that attendees with graduate level background of mathematics will be able to benefit from it and learn the underlying computational neuroscience and robotics.

Pre-recorded lectures

A pre-recorded version of tutorial will be made available to registered participants, at Hall 13 through the link: https://iros2021.gcon.me/programme.

Registration link: https://www.iros2021.org/registration

Live interactive session

The live tutorial session is planned for 14:30-18:00, Monday, 27th September 2021 (Prague time). The live session will be streamed via the IROS conference platform (https://gcon.cz/en/).

Supplementary materials

Some content from this tutorial is described in the following references.

References

[1] O Khatib (1987), A unified approach for motion and force control of robot manipulators: The operational space formulation. IEEE Journal on Robotics and Automation 3(1): 43-53.

[2] N Hogan (1985), Impedance control: An approach to manipulation. ASME Journal of Dynamic Systems, Measurement, and Control 107(1): 1-24.

[3] E Burdet, R Osu, D Franklin, T Milner and M Kawato (2001), The central nervous system stabilizes unstable dynamics by learning optimal impedance. Nature 414(6862): 446-9.

[4] C Yang, G Ganesh, S Haddadin, S Parusel, A Albu-Schaeffer and E Burdet (2011), Human-like adaptation of force and impedance in stable and unstable interactions. IEEE Transactions on Robotics 21(5): 918-30.

[5] Y Li, G Ganesh, N Jarrassé, S Haddadin, A Albu-Schaeffer and E Burdet (2018), Force, impedance, and trajectory learning for contact tooling and haptic identification. IEEE Transactions on Robotics 34(5): 1170-82.

[6] G Ganesh, A Takagi, R Osu, T Yoshioka, M Kawato and E Burdet (2014), Two is better than one: Physical interactions improve motor performance in humans. Scientific Reports 4: 3824.

[7] A Takagi, G Ganesh, T Yoshioka, M Kawato and E Burdet (2017), Physically interacting individuals estimate the partner’s goal to enhance their movements. Nature Human Behaviour 1: 54.

[8] A Takagi, F Usai, G Ganesh, V Sanguineti and E Burdet (2018), Haptic communication between humans is tuned by the hard or soft mechanics of interaction. PLOS Computational Biology 14(3): e1005971.

[9] A Takagi, M Hirashima, D Nozaki and E Burdet (2019), Individuals physically interacting in a group rapidly coordinate their movement by estimating the collective goal. eLife 12(8): e41328.

[10] A Takagi, Y Li and E Burdet (2020), Flexible assimilation of human's target for versatile human-robot physical interaction. IEEE Transactions on Haptics doi: 10.1109/TOH.2020.3039725.

[11] N Jarrassé, T Charalambous and E Burdet (2012), A framework to describe, analyze and generate interactive motor behaviors. PLoS ONE 7: e49945.

[12] Y Li, G Carboni, F Gonzalez, D Campolo and E Burdet (2019), Differential game theory for versatile physical human-robot interaction. Nature Machine Intelligence 1: 36-43.