=Paper=
{{Paper
|id=Vol-1419/section0015
|storemode=property
|title=None
|pdfUrl=https://ceur-ws.org/Vol-1419/section0015.pdf
|volume=Vol-1419
}}
==None==
https://ceur-ws.org/Vol-1419/section0015.pdf
Cognitive Development and Architectures for Cognitive Robotics
Chairpersons
Alessandra Sciutti
Robotics, Brain and Cognitive Sciences Department, Istituto Italiano di Tecnologia, Genoa, Italy
alessandra.sciutti@iit.it
Tomoyuki Yamamoto
CiNet, National Institute of Information and Communications Technology, Osaka, Japan
yamamoto@irl.sys.es.osaka-u.ac.jp
Minoru Asada
Graduate School of Engineering, Osaka University, Osaka, Japan
asada@ams.eng.osaka-u.ac.jp
Giulio Sandini
Robotics, Brain and Cognitive Sciences Department, Istituto Italiano di Tecnologia, Genoa, Italy
giulio.sandini@iit.it
Speakers
David Vernon
Interaction Lab, School of Informatics, University of Skövde, Sweden david@vernon.eu
Jochen Steil
Research Institute for Cognition and Robotics (CoR-Lab), Bielefeld University, Bielefeld, Germany
jsteil@cor-lab.uni-bielefeld.de
Yukie Nagai
Graduate School of Engineering, Osaka University, Osaka, Japan
yukie@ams.eng.osaka-u.ac.jp
A key feature of humans is the ability to entertain models different aspects of cognitive robotics with a focus on
of other agents, to anticipate what they are going to do and learning and development by selected speakers as Professor
to plan accordingly a collaborative action. Analogously the Yukie Nagai (Osaka University) and Professor Jochen Steil
focus of cognitive robotics is on predictive capabilities: (Bielefeld University). Professor Giulio Sandini (Istituto
being able to view the world from someone else's Italiano di Tecnologia) and Professor Minoru Asada (Osaka
perspective, a cognitive robot can anticipate that person's University) will chair the session providing an introduction
intended actions and needs. Hence, a fundamental aspect of and a link between the different perspectives of
cognition, both natural and artificial, is about anticipating developmental cognitive robotics and discussing its
the need for action and developing the capacity to predict relevance for a multidisciplinary understanding of cognition.
the outcome of those actions. But how does this capability
develop in humans and how can it be developed in robots? This symposium is part of the CODEFROR project
(COgnitive Development for Friendly RObots and
The goal of this symposium is to address these questions Rehabilitation, https://www.codefror.eu/), which aims at
by investigating aspects of cognitive development through joining the forces and expertise of the participating partners
the development of cognitive robots. The discussion will (Italian Institute of Technology, Bielefeld University,
focus on what is a cognitive architecture, on how predictive Osaka University and Tokyo University) to help the
learning could lead to social cognition, and how bio-inspired establishment of an international community of researchers
cognitive architectures in robotics could prove fundamental that shall effectively bridge the expertise of the different
for (physical) interaction. The session will start with an disciplines as robotics and cognitive sciences in the
overview on artificial cognitive architectures by Professor investigation of cognitive development.
David Vernon (University of Skövde), followed by talks on
27
Introduction on cognitive robotics models for robots to learn to acquire cognitive functions.
Predictive learning is defined as a process to minimize a
Giulio Sandini prediction error between an actual sensory feedback and a
predicted one. For example, the prediction error of the self’s
body should become zero because the state change of the
In this talk I will introduce the symposium by presenting
self can be perfectly predicted after learning. In contrast, the
how the study of artificial intelligent systems has evolved
body of other individuals produces a prediction error due to
from addressing separately the sensory, motor and cognitive
aspects of intelligence to an integrated “embodied” the influence of a context even after learning. Infants
approach. Using examples from our current studies of therefore can discriminate the self from others based on the
prediction error. Social behaviors such as imitation and
visually driven human-robot interaction and human
cooperation may emerge through the process of minimizing
sensorimotor development I will explain how this change of
the prediction error. A failure in others’ action induces a
perspective is now building a new scientific community
larger prediction error and thus triggers the execution of
composed of robotics engineers and neuroscientists studying
the central role of the body in mediating the integration infants’ own action to reduce the error, which results in the
between multiple sensory representations of space and accomplishment of the failed action. My talk will present
our robotics studies investigating how infants acquire the
actions.
ability of self/other cognition, goal-directed action, and
altruistic behavior. Furthermore, a potential of our
Artificial cognitive systems
hypothesis to understand the mechanism of autism spectrum
David Vernon disorders (ASD) will be explained.
This talk offers a review on the emerging field of artificial Biomorphic control as key for cognitive soft robotics
cognitive systems. Cognition, both natural and artificial, is
Jochen Steil
about anticipating the need for action and developing the
capacity to predict the outcome of those actions. Drawing
on artificial intelligence, developmental psychology, and The new scientist describes it as: " I AM in Jochen Steil's
lab, grasping a segmented, whiplashing tentacle that resists
cognitive neuroscience, the field of artificial cognitive
and tries to push me away. It feels strangely alive, as though
systems has as its ultimate goal the creation of computer-
I am trying to throttle a giant alien maggot. In fact, I am
based systems that can interact with humans and serve
training a bionic elephant's trunk to do real-world jobs like
society in a variety of ways. In this talk a working
definition of cognitive systems will be provided—broad picking apples or replacing light bulbs – something non-
enough to encompass multiple views of the subject and deep experts haven't been able to do until now." (Paul Marks,
13.03.2014).
enough to help in the formulation of theories and models.
The talk presents the scientific work behind this
Moreover, a brief survey of the different paradigms of
experience with the futuristic bionic handling assistant
cognitive science - the cognitivist, emergent, and hybrid
(BHA) soft robot, which indeed was modelled after an
paradigms will follow. These definitions will enable a
discussion of the broad range of existing cognitive elephant's trunk by its producer FESTO. The BHA is a
architectures, which represent the effective blueprints for large-scale pneumatically operated, flexible, soft and
compliant continuum robot, which is inherently safe to
implementing cognitive systems. The aim of the talk is to
interact with. We show how to use advanced learning
provide an understanding of the scope of the domain, the
methods to establish a proper mixture of adaptive controller
different perspectives and their underlying differences, to
approximations, and autonomous exploration for this
gain an idea of the issues that need to be addressed when
attempting to design an artificial cognitive system challenging platform. We discuss how by means of a
coherent software architecture this biomorphic control is
enabled and how it can be a blueprint for more general
Predictive learning as a key for cognitive development:
cognitive architectures in softer robots for performing actual
New insights from developmental robotics
grasping tasks, kinesthetic teaching, and accelerated online-
Yukie Nagai learning in physical interaction.
Human infants acquire various cognitive abilities such as Conclusion: Affective and cognitive developmental
self/other cognition, imitation, and cooperation in the first robotics
few years of life. Although developmental studies have
Minoru Asada
revealed behavioral changes in infants, underlying
mechanisms for the development are not fully understood
yet. We hypothesize that predictive learning of sensorimotor This talk will conclude the symposium by reviewing and
discussing the different approaches proposed. Moreover it
information plays a key role in infant development. To
will introduce the concept of Affective and Cognitive
verify the hypothesis, we have proposed computational
28
Developmental Robotics, aimed at understanding human
affective and cognitive developmental processes by
synthetic or constructive approaches. Its core idea is
"physical embodiment," and "social interaction" that enables
information structuring through interactions with the
environment. Finally, future issues involved in the
development of a more authentic form of artificial cognition
and empathy will be discussed.
Acknowledgments
This symposium is part of the European Project
CODEFROR (PIRSES-2013-612555).
29