|
|
|
|
|
|
|
Systems become increasingly complex. Their decomposition into smaller units is the usual way to overcome the problem of complexity. This has historically led to the development of atomized structures consisting of a limited number of autonomous subsystems that decide about their own information input and output requirements, i.e. can be characterized by what is called an information closure. Autonomous subsystems can still be interrelated and embedded in larger systems, as autonomy and independence are not equivalent concepts. These ideas are gaining very strong interest in both academia and industry, and the atomized approach to information flow modelling and evaluation is an idea whose time has certainly come. This presentation discusses some modelling and evaluation issues and challenges existing in the exciting area of information flow management for autonomous subsystems.
Biography
Professor Edward Szczerbicki , MSc (Gdansk), PhD (Gdansk), DSc (Szczecin) has had very extensive experience in the theory of information, autonomous systems analysis, and decision support systems development over an uninterrupted 25 year period, 13 years of which he spent in the top systems research centers in the USA, UK, Germany and Australia. He has published 170 refereed papers, over 100 of which appeared in international journals covering the area of systems science, decision support, and autonomous systems modeling and simulation. His DSc degree (1993), was gained in the area of the theory of information flow for autonomous systems. His PhD (1983) was gained in uncertainty modeling for design and MSc (1976) in engineering management.
He is now with The University of Newcastle, Newcastle, Australia.
Associate Professor Edward Szczerbicki
Assistant Dean, Faculty of Engineering
Co-ordinator, Master of Engineering Management Program
Department of Mechanical Engineering
The University of Newcastle
Newcastle NSW 2308, Australia
Telephone: +61 (02) 49 21 6209
Fax: +61 (02) 49 21 6946
Web: www.eng.newcastle.edu.au/me/staff/szczerbicki.html
This presentation tries to identify current critical issues of autonomous systems operating as physical entities in the real world. Although there are many computer systems in close connection to the physical world already (embedded systems), we wonder how much and what kind of autonomy we should wish for or be afraid of. In some environments basic autonomy is not just an option, but required - like in many underwater or planetary exploration applications. In most other environments we need to understand the possible or potential degrees of autonomy before declaring a possibly multiple ton vehicle or fast and powerful machine ready to interact with the physical world or even with biological species. Real-Time and synchrony are the further key topics in understanding dynamical systems in physical environments, while reliability and robustness are the essential constraints on the road towards practical systems. Moreover, interesting systems need to be adaptive to changing environments and should perhaps even implement life-long learning. How close or how far we are from integrating all these hard and challenging topics, will be investigated based on a set of historic and recent examples.
Human and robot skills are synergistic and complementary. Humans provide as yet unmatched capabilities to perceive, think, and act when faced with anomalies and unforeseen circumstances, but there are huge potential risks to human safety involved in getting these benefits. Robots provide complementary skills in being able to work in extremely risky environments, but their ability to perceive, think, and act by themselves is currently not error-free, although these capabilities are continually improving with the emergence of new technologies. There is substantial historical evidence to validate these generally qualitative notions. For example, N. Armstrong’s celebrated terminal descent maneuver to the lunar surface could most probably not be done reliably, even now, several decades after the event occurred. In contrast, robots have survived at Venus, Jupiter and other very extreme environments, not likely to be endured by humans in the foreseeable future. There are myriad similar historical anecdotes, both on Earth and in space, that suggest the relative strengths of humans and robots. This evidence is undoubtedly all accurate. However, such evidence must now be augmented with a more rigorous analytical framework that enables systematic, quantitative evaluation of human and robot roles, in order to optimize the design and performance of human-robot system architectures using well-defined performance evaluation metrics.
This talk summarizes a new analytical method to conduct such quantitative evaluations. Ideas from several disciplines, including robotics, human factors, and information and system theory are combined into a unified analytical framework for performance evaluation, optimization and forecasting. The method is generically applicable to all scenarios, including past, current and future missions. The talk discusses how to compare quantitatively the performance of a large variety of human-robot systems, how to optimize the allocation of roles to humans and robots, and how to forecast future performance. The results of 2 representative case studies, one in surface science exploration and the other in space assembly of large telescopes, are used throughout to illustrate the application of the method.
Charles R. Weisbin
Dr. Weisbin was Associate Professor of Computer Science, teaching artificial intelligence at the University of Tennessee. He has served as associate editor for IEEE Expert, and as member of the editorial board of IEEE Robotics and the International Journal of Applied Intelligence. He has also served as a proposal referee for the National Science Foundation, Department of Energy, and the National Aeronautics and Space Administration.
Dr. Weisbin was program chairman for the IEEE Second International Conference on Artificial Intelligence Applications. He has served as member of the Joint Technology Panel on Robotics (a group advising the military services of all branches regarding technology status and needs in robotics) and as co-chairman of the Robotics and Telepresence Subcommittee of the Space Technology Interdependency Group (STIG).
Dr. Weisbin was the co-chairman of the NASA Telerobotics Intercenter Working Group for seven years and received the 1993 NASA Exceptional Service Medal for formulation and development of the NASA Telerobotics Program . He is a recipient of the 1998 Thomas O Paine Award for Advancement of Human Exploration to Mars and a IEEE Computer Society Golden Core Member award. He also received the 1999 NASA Exceptional Service Medal for development and infusion of robotics from basic research into NASA flight missions and missions of other agencies. Finally, he has received the award for Outstanding Leadership in Surface Robotics for the year 2000, the Decadal Planning Team achievement award in 2001, and the NASA Group Achievement Award in 2001 for contributions to the NASA Decadal Planning Team.
