Community Research and Development Information Service - CORDIS


CogIMon Report Summary

Project ID: 644727
Funded under: H2020-EU.

Periodic Reporting for period 2 - CogIMon (Cognitive Interaction in Motion)

Reporting period: 2016-02-01 to 2017-01-31

Summary of the context and overall objectives of the project

Humans exploit compliant control and regulate their full body impedance in a variety of sophisticated skills. These include solitary actions such as soft catching and throwing, or sliding and pushing large objects. They perform joint actions such as manipulation of large-scale objects, or adapt through physical coupling in walking and joint execution of tasks. We refer to this advanced ability of organizing versatile motion under varying contact and impedance as cognitive compliant interaction in motion.

Compared to the richness and complexity of cognitive compliant interaction in humans, the flexible control of impedance and how to endow robots with it is very shallow. For developing versatile compliant robot interaction behaviour, we need to understand which information is conveyed through physical interaction by humans and how it is used in mutual adaptation.
This in turn requires to understand how models of impedance in interacting humans or robot partners can be formed, and how to predict the partners' motion behaviours from their kinematics and the observation of object movements and interaction forces.

Corresponding to the current very limited knowledge on full body impedance control, respective interaction in human-robot teams has typically been limited to a narrow set of carefully engineered scenarios. Therefore, the overarching objective of the CogIMon project is
to advance key technologies that lead to a step-change in cognitive compliant interaction in human-robot teams. It aims to integrate physical human-robot interaction, visually guided manipulation and safety integrated design in a systematic way.
To achieve this ambitious objective, cognitive modelling, compliance and impedance control, and learning need to be integrated into a coherent approach. The project is thus centered around three main experimentation scenarios dedicated to the key abilities developed in the project: Compliant catching and throwing with application in physiotherapy, compliant human-humanoid interaction for joint manipulation of larger objects, e.g. carrying a table, and multi human-robot interaction for joint manipulation in an shop floor scenario.

Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far

The project progresses along workpackes that are centered on i) advancing mechatronics to design a new compliant humanoid robot with full body impedance regulation, ii) developing a model-driven engineering approach to handle complexity of the targeted, highly integrated robotic systems both in simulation and real hardware, iii) analyzing human-human interaction in motion to identify and model human control strategies, iv) developing modular approaches for force control and impedance adaptation to facilitate human-robot interaction, v) interpretation and prediction of human motion, vi) integrating and evaluating cognitive interaction in motion in three final robotic experimentations.

In the first two years of the project, substantial progress has been made.
A novel hand design was conceived and installed on the consortiums' COMAN robots. New actuation units were developed and are included in the design of a new scaled-up COMAN robot.
In model-driven systems engineering, open-source, dynamics simulations for the target platforms COMAN and KUKA LWR-IV+ have been established that mirror the torque control interfaces of the real robots. A component-based modeling and simulation framework CoSimA utilizing Orocos as basis was developed and a modular composition architecture for domain-specific concerns has been conceived. For the model-driven software development, a dedicated toolchain was devised that complies with the requirements of the targeted robotic scenarios.

A second focus was the investigation of human-human interaction in motion. Sensorimotor control strategies for one- and two-handed catching have been investigated in experiments with humans exploiting motion capture systems. Experiments shed light on interpersonal coupling, perception of deceptive movements, and the influence of dynamic variables on the kinematics of juggling. Furthermore, coupled walking has been investigated based on novel interaction force measurement devices and respective models are developed and used to model compliant walking in contact in simulation.

New skills were developed. For soft catching, a dynamical system approach was demonstrated on a KUKA IIWA platform mounted with a dexterous hand, which
moves with the object before the hand is closed. Novel models for bi-manual coordination and catching were developed and implemented
on a pair of Kuka-LWR. First efforts were also made to model human throwing data and create respective robotic behavior. Walking under constraints, step-planning to maintain and optimize contact, and coupled walking algorithms are under development and testing. An approach derived from human data models impedance adaptation based on percieved interaction forces in cyclic tasks and has been integrated in the CoSimA simulation framework.

The project targets robotic experimentation to demonstrate the step-changes towards compliant interaction. Requirements have been assessed to meet the overall targeted technical readiness levels of the final experimentation scenarios. By means of vertical integration, the developed methods are embedded in the model-driven systems engineering approach. Substantial progress has already been made in the first two years in the implementation of bimanual catching, in bi-manual manipulation under constraints and with human intervention and on stable stepping and walking in interactions. Also, learning of manipulator dynamics and hybrid models has been focused to enhance model-based control in the absence of precise analytical formulations.

Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far)

The achieved results up to year two advance the current knowledge about human interaction in motion and the state-of-the-art in compliant control. The advancement of the humanoid robot COMAN and the scaling up towards a novel and perspectively more robust and versatile platform confirms and extends the world leading position of the European research in the development of variable impedance actuation and compliant humanoid robots. It is expected that the leading technological expertise in compliant actuation and full-body compliant robots will in the medium time-scale create new markets for novel actuation, fully compliant arms, and even full body robots.

Partners have established innovation strategies to deploy novel methods in robotics, manufacturing, healthcare and general industry e.g. through enterprise offices and dedicated transfer labs, which frequently run smaller projects with medium sized companies. Work package three targets to derive measures for naturalness and human-likeness in physical interaction, which will be an important factor for the acceptance of novel assistive and rehabilitation robotics applications. The first protocols for prospective application of throwing and catching in physiotherapy have been defined. Furthermore, the direct safe and efficient interaction with advanced robots will raise the general level of acceptance and is indispensable for applications in the private realm in the civil or health domain. The impact to the scientific community unfolds already by contributing to high-ranked journals and scientific conferences. Also, the partners’ principal researchers are all heavily involved in academic teaching, give talks, course and summer schools to educate young researchers.

Related information

Follow us on: RSS Facebook Twitter YouTube Managed by the EU Publications Office Top