Objectives
From ICRA Workshop on Automated Diagnosis
Gerald Steinbauer (Talk | contribs)
(New page: Technical systems in particular robot systems get more complex as they work on more advanced tasks in more realistic environments. The increase of complexity is true for the hardware and s...)
Next diff →
Current revision
Technical systems in particular robot systems get more complex as they work on more advanced tasks in more realistic environments. The increase of complexity is true for the hardware and software of the robot systems as well as for the algorithms and methods uses in the system. This complexity is hard to control and lead to unexpected behaviors, faults and mission failures. These characteristics are diametrical to the goal of reliable, dependable, autonomous robot systems.
In the last two decades a lot of research had been conducted in the field of fault-tolerant, self-aware and robust autonomous systems. The attempts to develop such intelligent systems can roughly be divided into two directions. One direction is basically driven by the development and application of robust robot systems and origin from the Robotics, Electrical and Mechanical Engineering community. The focus lies here for instance on the development of safe development processes, redundant systems and highly technical solutions in order to cope with the issue.
The other direction followed the development of intelligent monitoring, reasoning and diagnosis algorithms and has been driven by people from Computer Science and Artificial Intelligence. The main interests are here is a fundamental understanding of how technical systems can be modeled accurately, how faults or unexpected behaviors can be detected efficiently and how plausible explanations and solutions for a problem can be found automatically.
Both directions went a long way and achieved fantastic results. But we are convinced that both direction can learn a lot from each other. Therefore, the main objective of the workshop is to bring together researcher from these various directions in the context of one of the largest Robotics and Automation conferences. We expect an intense exchange of knowledge, experiences and ideas between these groups. This exchange will lead to an stimulating cross-fertilization of these areas. In particular researchers which develop and build intelligent robot systems will learn new methods to cope with unexpected situations and researchers from the diagnosis community will see new potential application domains and learn where are the real problems in an autonomous system acting in a real-world environment.
