RoboCup 2001 Seattle
RoboCup-Rescue Simulation League Competition Report

Preliminary Competitions   Final Competitions    

Congratulations!!!

YabAI (Takeshi Morimoto, The University of Electro-communications) won the first chamiponship in the RoboCup 2001 RoboCup-Rescue Simulation League, the first international competition in RoboCup.

Final Result

Ranking Team Name
1 YabAI, Japan
2 Arian, Iran
3 Rescue-ISI-JAIST, USA & Japan

Preliminary Competitions

Procedure

Each agent party tries 7 disaster situations which were created by the teams.  4 agent parties with the top total rank scores proceed to the final game.

Score Table

Top team gets the largest score in simulation of the same map.  If 7 teams are valid, the winner gets 7.

No. Team Name
(with link to its team description)
Disaster Maps Simulated & Winning Score of Each Team (Score Point V) Total Ranking
Aug. 4 (Sat) Aug. 5 (Sun) Aug. 6 (Mon)
9:00- 12:30- 15:00- 9:00- 12:30- 15:00- 9:00-
YabAI Gemini ISI-JAIST JaistR Arian NIT RMIT
1 YabAI 7 (18.2565) 6 (19.3625) 7 (29.3857) 7 (41.5956) 4 (28.5506) 2 (30.4815) 6 (13.2305) 39 2
2 Gemini-R 3 (56.8933) 5 (21.4629) 3 (45.6748) 3 (56.8082) 3 (51.8539) 4 (25.5616) 5 (15.2404) 26 5
3 Rescue-ISI-JAIST 5 (30.7056) 2 (27.7510) 4 (40.5552) 5 (43.7608) 6 (22.3071) 6 (24.4420) 2.5 (17.2266) 30.5 3
4 Arian Simulation Team 6 (30.3787) 7 (17.2783) 6 (33.4425) 4 (54.7166) 7 (22.3053) 3 (27.3619) 7 (10.2446) 40 1
5 JaistR 4 (30.7062) 3 (22.8194) 5 (38.5520) 1 (62.8445) 5 (22.3113) 7 (24.4106) 4 (17.2266) 29 4
6 NITRescue 2 (58.8915) 4 (22.4699) 2 (66.8879) 6 (43.7592) 2 (54.8634) 1 (36.7906) 2.5 (17.2266) 19.5 6
7 RMIT-ON-FIRE 1 (60.9129) 1 (36.8519) 1 (35.6395) 2 (62.8402) 1 (56.8618) 5 (24.7488) 1 (17.5042) 12 7

Note: RMIT-ON-FIRE developed only fire agents and used sample codes for the other agents.  Therefore, simple comparison is not fair.

Final Competitions

Procedure

Top 4 teams make a tournament by the score point (evaluation value V).

Semi Final A Score

August 9 (Thu) 9:00-
Preliminary
Game Ranking
Team Name Score Point (V) Winning Point
GeminiR-2 map NIT-2 map
2 YabAI 17.2792331741 20.2402455260 2
3 Rescue-ISI-JAIST 17.6728162788 20.4167472486 0

Semi Final B Score

August 9 (Thu) 11:00-
Preliminary
Game Ranking
Team Name Score Point (V) Winning Point Note
GeminiR-2 map NIT-2 map RMIT map
1 Arian 13.2451824319 20.2565185571 19.2300132388 1  *
4 JaistR 14.6632544855 16.3553287093 17.3933319876 2  **

Note *: In NIT-2 and RMIT maps, more civilian agents accidentally escaped from collapsed buildings at the first stage in JaistR trials.  This caused Arian's loss of games.
Note **: JaistR was disqualified because his code included a part which could not refute the claim that it did not follow the gentlemen's rule.  Although it was apparently a careless mistake, the committee could not help deciding the disqualification.  The detail will be explained soon.

3rd Position

Semi Final Group Team Name Final Ranking
A Rescue-ISI-JAIST 3

Final Score

Semi Final
Group
Team Name Score Point (V) Winning
Point
Final
Ranking
August 10 (Fri) 9:45-
Rescue-ISI-JAIST Committee 2 Committee 3
A YabAI 19.2646659794 56.8223911629 --- 2 1
B Arian 33.3954412371 56.8551949182 --- 0 2

Participating Teams

No. Team Name Team Members
(Affiliations)
Technical Appeal Points Prepared Agents
1 YabAI Takeshi Morimoto
(University of Electro-Communications, Japan)
Behavior of agents is switched according to the distribution of disater. Ambulance, FireBrigade, PoliceForce
2 Gemini-R Masayuki Ohta
(Tokyo Institute of Technology, Japan)
Agent strategies are determined according to the optimal order of rescue activities and allocation of man-power that has been automatically learned by simulation results. Ambulance, FireBrigade, PoliceForce
3 Rescue-ISI-JAIST Takayuki Ito, Milind Tambe, Ranjit Nair, Stacy Marsella
(Information Science Institute / University of Southern California, USA; Japan Advanced Institute of Science and Technology, Japan)
Cooperative and autonomous rescue activities are generated. Ambulance, FireBrigade, PoliceForce, Ambulance-C, FireBrigade-C, PoliceForce-C
4 Arian Simulation Team Jafar Habibi, Mazda Ahmadi, Ali Nouri
(Sharif University of Technology, Iran)
??? Ambulance, FireBrigade, PoliceForce, Ambulance-C, FireBrigade-C, PoliceForce-C
5 JaistR Kosuke Shinoda
(Japan Advanced Institute of Sience and Technology, Hokuriku, Japan)
Each Agents has learning mechanism based on Organizational Learning. Ambulance, FireBrigade, PoliceForce, Ambulance-C, FireBrigade-C, PoliceForce-C
6 NITRescue Tetsuya Esaki, Taku Sakushima, Nobuhiro Ito, Yoshiki Asai,
(Nagoya Institute of Technology, Japan)
Cooperation of multiple agents is important in a dynamic environment as the RoboCup-Rescue simulation.   Cooperative behabior of agents is defined as a group behavior of agents.  A dynamic grouping algorithm is developed. Ambulance, FireBrigade, PoliceForce, Ambulance-C, FireBrigade-C, PoliceForce-C
7 RMIT-ON-FIRE Lin Padgham, James Harland, John Thangarajah, Naveen Ruwanpura, Chandaka Fernando
(Royal Melbourne Institute of Technology University, Australia)
BDI intelligent agent system. Development primarily by final year undergraduates using JACK BDI agent development environment. FireBrigade

Download

The Tar Package of the Logs(12MB)
The Tar Package of the Maps(16KB)
The Tar Package of the All Files(560MB)