From: Silvain van Weers (scamweer@pop.students.cs.uu.nl)
Date: Mon 15 Apr 2002 - 22:01:05 GMT
Hello everyone,
We are a team of students from Utrecht University (Netherlands) doing a
project on multi-agent planning concerning the roborescue simulation
league. We're planning for participation for the WC in 2003.
We're currently in the development process of agents, and regarding the
rules of the world championship 2002, we have a few questions concerning
communication between agents. However, because only rules for 2002 exist,
we believe some rules are tentative due to 2003. After some debates in our
team we came up with the following questions:
1. The rules state that communication between agents is allowed.
Communication will consist of "simple sentences". Are there any specifics
around these "simple sentences"? How complex may they be?
2. Consider this following situation: All agents send report of a
noteworthy phenomenon to one particular agent and they it decides whether
or not to recieve it.
Thus hat agent is flooded." Is this type of flooding allowed?
3. Are the AK_SAY and AK_TELL both counted for the Maximum of 4 messages,
or just the AK_TELL?
4. "Is it possible to communicate from an agent to an agent directly (via
AK_TELL) without the existence of a center?
5. What information of an ID may be retreived when an agent receives a
certain ID? May it for instance conclude of which type it is?
6. Are agents allowed to remember they recieved but did not read a message?
if so, may they use that information for more than just the message
filtering code (i.e., the code deciding which 4 out of the n messages will
be read this turn).
7. In the rules, when spoken of a 'team' what is meant? All agents on a
citymap from a certain brigade i.e. Fire ? Or just one agent
(Fire/Ambulance/Police) itself?
8. When centers collapse is it allowed that a certain agent replaces the
functionality of the center?
Best regards,
Silvain van Weers
This archive was generated by hypermail 2.1.4 : Sun 22 Sep 2002 - 18:52:58 GMT