Interacting with Humans – A Robot’s View

Gerhard Sagerer, Cor-Lab and CITec, Bielefeld University

One common goal of an increasing number of projects is the development of robots interacting with humans in various and changing environments. Starting from paradigms of situated communication, cooperative construction and optimization of behaviors we are focusing on interacting and learning based on the idea that robots are enabled to understand the human as communication partner. This requires the perception and interpretation of different modalities like speech, gestures, actions, etc. In this approach, a robot is viewed as an agent that acts in the real world and learns to understand its environment through communication and interaction with a human tutor.

In my talk I will present our approach of communication driven interaction and learning which we interpret as a two-fold process: modeling attention towards the communication partner on the one hand and focusing attention on the semantics as highlighted by the human in order to make sense of the otherwise unstructured sensory information. We argue that by taking the human’s structuring behavior into account a robot can learn to assign meaning to presented actions and objects such as goals and means or targets and obstacles. Only by understanding the meaning of actions and objects the system will be able to generalize to unknown situations and interact in the right manner.

From wireless networks to sensor networks and onward to networked embedded control

P. R. Kumar

We address the issue of organizing principles and architecture for three different types of emerging systems, which may possibly be at the cusp of a take-off: wireless networks, sensor networks, and networked control. In wireless networks, there is no a priori notion of links: transmitting nodes simply radiate energy, and receiving nodes hear a superposition of all concurrent transmissions. We address the question of what should be the architecture of wireless networks. Should they be organized like wired networks or in an entirely different way? Also, can we do away completely with wired networks and replace them with wireless networks? Or are there limits to what wireless networks can provide? Sensor networks are comprised of nodes equipped with sensors for monitoring their environment, as well as computational and wireless communication capabilities. The entire network comprises a computational-cum-communication-cum sensing system. We address the issue of how information should be processed in-network within such systems. Finally, we turn to the problem of networked control, where nodes can also act on their environment, besides sensing the environment. What are the appropriate abstractions and architectures for such systems that can interact with the physical world? We provide an overview of efforts in the Convergence Lab at the University of Illinois to elucidate this issue.

 (Joint work with many graduate students).

Medical surgical robots and its evaluation system

Toshio Fukuda

There are many medical robots research and developments but not many available in the market. It is because there need more medical data to validate the safety and reliability to human. Medical doctors also need more practice for the surgical operations to improve their skills. Therefore, it is more important to make use of the simulators for this purpose and evaluate the performance of the doctors. In this presentation, an intravascular surgery robot robot will be introduced and shown the effective methodologies. To validate these applications, the integrated simulator system is made for skill evaluation. This simulator can be used for the medical students to learn the basic human surgery, medical doctors to have simulation before the operation as tailored medicine and doctors to get the proper licenses for the operation. Some examples will be shown in this line.