|December 13th: The Return of the Light|
|The time when AI and Robotics rejoin... maybe!|
Artificial Intelligence and Robotics were strongly connected in the early days of AI, but became mostly disconnected later on. While there are several attempts at tackling them together, these remain isolated points in a landscape whose overall structure and extent is not clear. Recently, it was suggested that even the otherwise successful EC program "Cognitive systems and robotics" was not effective in putting together the two sides of "cognitive systems" and of "robotics".
One thing that seems to be lacking today is a clear perception by AI researchers that there is a great added value for them in making contact with robotics, and what this added value is. Many AI researchers would agree that applying their techniques to robotics can be a way to demonstrate the utility of these techniques, but few realize that this can also, and most importantly, generate new exciting research questions and challenges, which will help to advance the field of AI itself. As an example, applying answer set programming to robotics would generate fundamental questions like how to accommodate fluents, uncertainty and bounded computation.
Symmetrically, robotic researchers are increasingly interested in including higher level functionalities in robots, and they delve into topics which have been the object of intensive study in AI. Not all realize, however, that there is a great added value for them in becoming acquainted with the thoughts and results (and, why not, errors) from that field. As an example, many working in semantic maps in robotics have only little knowledge of the work done in the field of knowledge representation in AI.
This event will focus on this added value question, both for AI and robotics researchers, in all its facets: (1) what are the new scientific questions that can be generated in each field, (2) what is the knowledge that can be transferred between fields, (3) how can we educate the next generation of researchers able to truly put AI and robotics together, and (4) what are the new application opportunities created by this merger.
The aim of this event is to gather a group of people who, while coming from different communities, have long felt the need to combine the AI and robotic traditions, and discuss how this need can lead to a better defined and recognized research field.
In order to focus the discussion, we will consider a concrete goal that embodies the spirit of this event: define a strategy to create a new generation of researches who have a deep interest and understanding of the integration of AI and robotics. This strategy might involve steps like:
This is not a symposium, so we will not have paper presentations. In a truly "workshop" style, we want people to work together around common questions and objectives. We will ask, however, participants to present their position with respect to event's questions.
This is not about system integration. While system integration is a key issue for building robots that use AI techniques, the main emphasis of this event is on conceptual integration.
This is not about architectures. We have had architecture discussions for many years, and these have been important. But this event is more about defining a field, and we'll be happy to be architecture agnostic.
The event will be mostly a guided brainstorming. Guidance will be provided by a set of questions, which will direct bottom-up contributions, and by a set of target outputs, which will set top-down goals. Of course we'll strive to achieve a reasonable blend of goal-directed discussion and spontaneous, wild brainstorming. We might consider the use of an external professional to guide the discussion.
The set of questions will be distributed some time before the event, and each participant will be asked to provide a one-page position statement two weeks before the event addressing these questions. Questions to robotics people might include things like: "Suppose you could be given a ROS stack called AI-Tools, what would you like, dream, expect to find there?" Questions to AI people might include things like: "Consider the work on semantic maps in robotics and the work done in KR, where do you see that the latter could contribute to the former, where do you see that the latter should be extended?"