Artificial Intelligence Laboratory
Department of Informatics
University of Zurich
Tel. +41 44 635 4320/53
Fax +41 44 635 4507
The main goal of the AILab is understanding natural behaving systems and developing intelligent artificial systems. We employ a synthetic methodology, i.e. a methodology of "understanding by building". According to our approach of "New Artificial Intelligence", or "Embodied cognitive science" we use embodied systems - autonomous mobile robots - acting in real environments. In addition to neural control, morphology and the specifics of the physical systems play an essential role. All of our research projects are interdisciplinary.
Keywords: Embodied cognitive science, neural networks, synthetic modeling, artificial intelligence
1 professor, 4 postdocs, 13 PhD students
The main goal of artificial intelligence is to work out the principles underlying intelligent behavior. These principles will enable us on the one hand to understand natural forms of intelligence (humans, animals), and on the other to design and build intelligent systems (computer programs, robots, other artifacts). We use a synthetic methodology that can be characterized as "understanding by building". It consists of three steps: (1) modeling aspects of a biological system, (2) abstracting and exploring general principles of intelligence, and (3) using these principles in the design of artifacts.
All projects contribute, one way or other, to our central theme of understanding intelligence. The central concepts which form the basis of our approach include embodiment (the physical realization of agents), morphology, system-environment coupling, dynamics, and material properties. Our main research fields are biorobotics, learning and development, evolution and morphogenesis, and collective intelligence.
In the following, two examples of robotic projects are presented which are particularly relevant for neuroscience. Many more projects and latest results can be found on our webpage (www.ifi.unizh.ch/ailab). The first example is a biologically inspired robot, the AMouse robot (see figure). The AMouse has been built to model the whisker system of mice and cats. It is equipped with active artificial whiskers and an omnidirectional camera.
Figure 1: A mouse robot displying artificial whiskers while palpating an object.
Figure 2: Cartoon of the robot-cub.
Figure 3: Robotic hand learning to grasp objects.
Another example is the large European project Robot-cub (ROBotic Open Architecture Technology for Cognition, Understanding and Behaviors). Robot-cub's main scientific goals are:
We intend to construct an embodied system able to learn: i)
how to interact with the environment (mainly) by complex manipulation and through gesture production/interpretation; and ii) how to develop its perceptual, motor and communication capabilities for the purpose of performing goal directed manipulation tasks. The embodied cognitive system (iCub) will be shaped, physically and mentally, like a human child and will be designed as an "open system" to be shared by scientists as a common tool for addressing cognition and human-machine interface. This objective, therefore, will be achieved by jointly designing the mindware and the hardware: the body of Robot-cub with the goal of providing the scientific community with a set of physically instantiated tools indispensable to study cognition.
Our main task within the project is the research on adaptive learning mechanisms.
From Locomotion to Cognition: sensorimotor coordinated behavior is the basis for more advanced categorization and cognition. We want to study how cognitive systems arise from the interaction of embodied agents with their environment, and how their cognitive abilities depend on their locomotor afordances.
Embodied artificial intelligence implies that most experiments are done with real robots or physically realistic computer simulations. Besides many commercially available robots (Lego Mindstorms, Kheperas, Koala, Samurais, Aibo etc.), the lab has a mechanical workshop and equipment for creating printed circuit board to build custom-made robots. Fish robot experiments are conducted in a large water tank.
Several European projects, Swiss National Science Foundation (SNF)
Diese Website wird in älteren Versionen von Netscape ohne graphische Elemente dargestellt. Die Funktionalität der Website ist aber trotzdem gewährleistet. Wenn Sie diese Website regelmässig benutzen, empfehlen wir Ihnen, auf Ihrem Computer einen aktuellen Browser zu installieren. Weitere Informationen finden Sie auf
The content in this site is accessible to any browser or Internet device, however, some graphics will display correctly only in the newer versions of Netscape. To get the most out of our site we suggest you upgrade to a newer browser.