Organization
The Dagstuhl seminar will be held from 16.08.2009 to 21.08.2009 at castle Dagstuhl. For information on how to get to castle Dagstuhl, please visit this webpage. The Dagstuhl seminar will contain two parts.
During the first part of the workshop, we will ask every participating researcher to present their research work and the research issues they address in the context of a common application scenario: a humanoid robot performing household chores including setting the table, preparing meals, and cleaning up. These presentations will acquaint participants with the research areas of other participants. The presentations will also establish the state of the art in the relevant research areas neuroscience, cognitive psychology, robot control theory, computer science/artificial intelligence, cognitive vision/perception, and machine learning.
The second phase of the workshop will consist of group discussions in which the general objectives described in the previous section will be translated into specific scientific questions, competing solutions to open problems, grand challenges, and application-specific approaches. Goals of these discussions will be:
- to identify research topics for which we expect revolutionary successes, in particular through interdisciplinary synergies
- to debate how scalability and robustness can be achieved based on existing robotic systems, or if fundamental changes are required
- to discuss the most promising research directions for achieving robustness and scalability in robotic systems in unstructured environments
- to predict where the research area of cognition in technical systems will be in five years and in ten years
- to outline a research agenda for the field
- to identify concrete actions that can improve the cohesion of a multi-disciplinary community in the context of cognitive robotic systems
- to make suggestions for funding strategies and international collaborations
During the workshop, the list of discussion points will be grown and modified in response to the outcome of discussions.
Once the participants have agreed on a prioritized list of discussion points, suitable sub-groups will be identified for break-out sessions. During the remainder of the workshop, these break-out sessions will alternate with plenary sessions so as to advance in individual discussion points while preserving the larger context.
Each of the break-out groups will be asked to write a final-report. These final reports will be edited and merged into a single workshop report.
Outcomes
Besides the impact of the seminar on the participating researchers we plan the following additional results.
- a roadmap for research on cognitive robot manipulation systems in the presence of humans
- a final report of the workshop summarizing the most important results of the discussions
- a Springer LNAI book representing the state-of-the-art in cognition, learning and control in human-robot cohabited environments. This book will consist of a collection of papers from the seminar
Organizers
- Michael Beetz (TU München, DE)
- Oliver Brock (Univ. of Massachusetts - Amherst, US)
- Gordon Cheng (ATR - Kyoto, JP)
- Jan Peters (MPI für biologische Kybernetik - Tübingen, DE)
The Dagstuhl seminar is sponsored by the Leibniz-Gemeinschaft.
