kazakov@cs.york.ac.uk
Date: Thu 05 Dec 2002 - 04:17:34 GMT
Dear colleagues, I would like to bring to your attention some FP5 initiatives that are quite relevant to the area of Adaptive and Learning Agents. I will also discuss some important FP6 issues - please read if you can spare a few minutes, and send feedback to the mailing list, if interested. Facts: An effort is being made in Europe to consolidate an existing body of multi-disciplinary research using the apparatus of Mathematics and Physics to study the general principles of complex systems and their various applications. To the moment, this has resulted in the creation of - Exystence, a recently launched FP5 network of excellence - RISK, a COST project, which is to start in 2003. The Future and Emerging Technology (FET) section of the soon expected FP6 call for proposals is likely to contain a pro-active call for proposals in the area of Complexity. Agents, Evolution and Adaptation will be mentioned as important topics in this context. The framework provided by Exystence is already being used to co-ordinate proposals for an FP6 NoE, and Integrated Projects. There will be at most one NoE in the area of Complexity, and only a couple of IPs are likely to be supported in that area. Comments: The above facts are of importance to the ALAD SIG and the ALAMAS proposal for the following reasons. Firstly, much of the relevant research presented at the Exystence launch meeting in Turin last month discussed topics that are central to the ALAD (ALAMAS) community (see the Exystence Web site http://www.complexityscience.org/, and check the 'Resources/Complexity Links' item in the main menu for some background reading). This FP5 Network supports collaboration, workshops, etc., and is open for new members to join. Links to this community could be useful, both in terms of direct collaboration and as a way to influence the direction the Complexity research is going to take in order to make the people involved aware of the issues we are interested in. Secondly, the above mentioned Risk COST project has Learning Agents as one of its three main work-packages. The RISK co-ordinators have been made aware of the existence of ALAD SIG and of the ALAMAS proposal, and have proposed in turn to organise joint initiatives, such as workshops and short visits, once the project has started. Thirdly, the FP6 call is expected to limit the funding for NoEs and IPs in the FET area to the pro-active mode in 2003. That means that next year, the 70% of the FP6 funding that is earmarked for FET NoEs and IPs will only be given to areas specified by the European Commission. (There will be, however, the possibility of submitting a proposal for an FP5-like thematic network in other areas to the FET-Open call.) There is a drive from the Commission to identify the relevant scientific areas and start a dialog between the potentially interested researchers. A substantial part of the Exystence launch meeting, was spent discussing FP6 proposals on the same topic. My personal perceptions of the discussions were that while a large part of them looked at issues we have listed on the ALAD page and in the ALAMAS proposal, Computer Science as a whole was under-represented, and the technology aspect of agents all but omitted. At the same time, the EU FET officer present at the meeting made it clear that any proposal should have a clear contribution to the IST (Information Society Technology). So far, the ALAMAS proposal has been designed as a NoE, and targeted the FET section of the FP6 call. If that were to remain unchanged, we should target one of the three proactive FET initiatives: (i) Beyond Robotics - which mentions self-organising, adaptive, co-operating and evolving robots (ii) Complex Systems Research (discussed in this message) (iii) The Disappearing Computer (ambient IT systems). You will be able to see the call for yourself when it is published in a week or so. While one could be tempted to consider (without excluding any other options) a proposal to the Complexity call, which sets the objective of creating "a new generation of scale-free, autonomously evolving IT systems building on…paradigms derived from complex system analysis", it is obvious that the expertise mentioned in the last part of the quote is somewhat missing from the present ALAMAS consortium. Also, competing with a proposal coming from an existing NoE that has 'complexity' in its name may be counter-productive. A better approach, and one that would match the perceived efforts of the EU Commission to consolidate all Complexity NoE effort at this early stage, would be to make the case for a significant CS presence in it and, if successful, join it. Again, this should not exclude plans for a stand-alone ALAMAS proposal to the FET-open call. If there is a sufficient interest in the Complexity FP6 NoE proposal, we should try to - Include and consolidate our areas of interest in the currently drawn concept map of the Complexity area: http://complexity.cogniview.com/MapIndex.html. (Action: contact directly the site manager). - Compile a couple of paragraphs that summarise these areas and send it to the Risk COST project mailing list - a reciprocal action from their part is expected. (Action: send me brief suggestions formulated as complete sentences). If the above two actions yield a good result, we will further discuss with Sorin Solomon, the Complexity FP6 NoE proposal co-ordinator, the role that we could play in this proposal. (Many of us will be able to meet him at AAMAS-3 where he will deliver the invited talk.) I will stop here and wait for your feedback. Best greetings, Dimitar Kazakov