Contributions published at Artificial Intelligence Lab

This thesis summarizes our contribution to science and technology education by developing a novel robotic toolkit (EmbedIT). Our aim was to enable students to explore in an easy and fun way interdisciplinary topics such as biologically inspired robotics, embodied artificial intelligence, electronics, computer science, neuroscience, psychology, arts etc. The need for a new toolkit emerged based on our experiences in teaching robotics to students with different backgrounds. The current commercially available robotic kits did not fully meet our needs. Therefore, we proposed a new design approach which includes modularity, compatibility with other toolkits, compliance with industrial standards, distributed control, versatility, and open-source licensing. We emphasized a practical approach closely related to the target audience: the secondary school students. We designed a number of example exercises using EmbedIT which train a variety of skills that young people need for a successful professional life. These exercises were conducted in real class environments with a large number of students. We addressed a number of research questions on the conceptual, technical, and practical level in the context of this thesis. Pedagogues, engineers as well as teachers benefit from the insights gained, the tools developed, and the example exercises designed. This work further contributes to educational robotics research and to the dissemination of biologically inspired robotics and embodied artificial intelligence. We are sharing EmbedIT with members of the open-source community who may customize the platform to their own needs and contribute to exploit the toolkit’s full potential. Im Rahmen dieser Doktorarbeit wurde ein neuartiger Roboterbausatz (EmbedIT) entwickelt, der als Lehrmaterial den Unterricht technischer und wissenschaftlicher Schulfächer unterstützen soll. Das Ziel ist es, mit Hilfe dieses Bausatzes verschiedene interdisziplinäre Gebiete auf spielerische und einfache Art zu entdecken, zum Beispiel biologisch inspirierte Robotik, künstliche Intelligenz, Elektronik, Informatik, Neurowissenschaften, Psychologie, Kunst. Das Bedürfnis, einen neuen Bausatz zu entwickeln, entstand aufgrund unserer Erfahrungen als Lehrkräfte im Robotikunterricht mit Studenten aus verschiedenen Studienrichtungen. Die herkömmlichen, auf dem Markt erhältlichen Roboterbausätze wurden unseren Ansprüchen nicht gerecht. Wir entschieden uns für ein Design, welches die Eigenschaften Modularität, Kompatibilität mit anderen Bausätzen, Einhaltung von Industriestandards, dezentrale Steuerung, Vielseitigkeit und Open-Source-Lizenzierung aufweist. Ein hoher Stellenwert hatte die Praxisbezogenheit und Nähe zum Zielpublikum, den Schülern auf Gymnasialstufe. Wir entwickelten mit dem Bausatz einige Besipielübungen, mit denen Kompetenzen trainiert werden können, die junge Menschen heutzutage im Berufsleben benötigen. Die Übungen wurden im Unterricht an verschiedenen Gymnasien mit einer grossen Zahl von Schülern durchgeführt. In dieser Doktorarbeit befassen wir uns mit Forschungsfragen auf konzeptueller, technischer und praktischer Stufe. Von den Erkenntnissen, dem entwickelten Bausatz und den Beispielübungen profitieren Pädagogen, Ingenieure und Lehrer zugleich. Die Ergebnisse tragen zur Forschungsrichtung «educational robotics» bei und zur Popularisierung von biologisch inspirierter Robotik. Der entwickelte Roboterbausatz EmbedIT wird unter einer Open-Source-Lizenz veröffentlicht, um zu ermöglichen, dass andere die Plattform zu ihren eigenen Zwecken und für ihre eigenen Bedürfnisse anpassen und weiterentwickeln können.

null

Several bivalve species burrow into sandy sediments to reach their living position. There are many hypotheses concerning the functional morphology of the bivalve shell for burrowing. Observational studies are limited and often qualitative and should be complemented by a synthetic approach mimicking the burrowing process using a robotic emulation. In this paper we present a simple mechatronic set-up to mimic the burrowing behaviour of bivalves. As environment we used water and quartz sand contained in a glass tank. Bivalve shells were mathematically modelled on the computer and then materialized using a 3D printer. The burrowing motion of the shells was induced by two external linear motors. Preliminary experiments did not expose any artefacts introduced to the burrowing process by the set-up. We tested effects of shell size, shape and surface sculpturing on the burrowing performance. Neither the typical bivalve shape nor surface sculpture did have a clear positive effect on burrowing depth in the performed experiments. We argue that the presented method is a valid and promising approach to investigate the functional morphology of bivalve shells and should be improved and extended in future studies. In contrast to the observation of living bivalves, our approach offers complete control over the parameters defining shell morphology and motion pattern. The technical set-up allows the systematic variation of all parameters to quantify their effects. The major drawback of the built set-up was that the reliability and significance of the results was limited by the lack of an optimal technique to standardize the sediment state before experiments.

Self-organization is a phenomenon found in biomolecular self-assembly by which proteins are spontaneously driven to assemble and attain various functionalities. This study reports on self-organized behavior in which distributed centimeter-sized modules stochastically aggregate and exhibit a translational wheeling motion. The system consists of two types of centimeter-sized water-floating modules: a triangular-shaped module that is equipped with a vibration motor and a permanent magnet (termed the active module), which can quasi-randomly rove around; and circular modules that are equipped with permanent magnets (termed passive modules). In its quasi-random movement in water, the active module picks up passive modules through magnetic attraction. The contacts between the modules induce a torque transfer from the active module to the passive modules. This results in rotational motion of the passive modules. As a consequence of the shape difference between the triangular module and the circular module, the passive modules rotate like wheels, being kept on the same edges as the active module. The motion of the active module is examined, as well as the characteristics and behavior of the self-organization process.

Morphological computation can be loosely defined as the exploitation of the shape, material properties, and physical dynamics of a physical system to improve the efficiency of a computation. Morphological control is the application of morphological computing to a control task. In its theoretical part, this article sharpens and extends these definitions by suggesting new formalized definitions and identifying areas in which the definitions we propose are still inadequate. We go on to describe three ongoing studies, in which we are applying morphological control to problems in medicine and in chemistry. The first involves an inflatable support system for patients with impaired movement, and is based on macroscopic physics and concepts already tested in robotics. The two other case studies (self-assembly of chemical micro-reactors; models of induced cell repair in radio-oncology) describe processes and devices on the micrometer scale, in which the emergent dynamics of the underlying physical system (e.g., phase transitions) are dominated by stochastic processes such as diffusion.

Abstract Anthropomimetic robotics differs from conventional approaches by capitalizing on the replication of the inner structures of the human body, such as muscles, tendons, bones, and joints. Here we present our results of more than three years of research in constructing, simulating, and, most importantly, controlling anthropomimetic robots. We manufactured four physical torsos, each more complex than its predecessor, and developed the tools required to simulate their behavior. Furthermore, six different control approaches, inspired by classical control theory, machine learning, and neuroscience, have been developed and evaluated via these simulations or in small-scale setups. While the obtained results are encouraging, we are aware that we have barely exploited the potential of the anthropomimetic design so far. But, with the tools developed, we are confident that this novel approach will contribute to our understanding of morphological computation and human motor control in the future.

Animals and humans engage in an enormous variety of behaviors which are orchestrated through a complex interaction of physical and informational processes: The physical interaction of the bodies with the environment is intimately coupled with informational processes in the animal's brain. A crucial step toward the mastery of all these behaviors seems to be to understand the flows of information in the sensorimotor networks. In this study, we have performed a quantitative analysis in an artificial agent — a running quadruped robot with multiple sensory modalities — using tools from information theory (transfer entropy). Starting from very little prior knowledge, through systematic variation of control signals and environment, we show how the agent can discover the structure of its sensorimotor space, identify proprioceptive and exteroceptive sensory modalities, and acquire a primitive body schema. In summary, we show how the analysis of directed information flows in an agent's sensorimotor networks can be used to bootstrap its perception and development.Read More: http://www.worldscientific.com/doi/abs/10.1142/S0219525912500786

null