PREFACE

RoboCup is an international joint project to promote Artificial Intelligence (AI) and Intelligent Robotics, and related fields. It is an attempt to promote AI and intelligent robotics research by providing a standard problem where wide range of technologies may be integrated and examined. RoboCup chose to use soccer as a central topic of research because this is a very complex and attractive collective game and in order to be able to play a soccer game, a huge amount of technologies must be involved. However the RoboCup federation is also aiming at innovations to be applied for socially significant problems and industries.

In order for a robot team to actually perform a soccer game, various technologies must be incorporated including: autonomous agents, multi-agent systems, strategy acquisition, knowledge representation, teamwork, coordination and cooperation, communication, real-time reasoning, distributed planning, learning, opponent modeling, image processing and analysis, computer vision, robotics, electromechanical design, power and digital electronics, applied control theory, intelligent perception and sensor-fusion.

The ultimate goal of the RoboCup project is “By 2050, develop a team of fully autonomous humanoid robots that can win against the human world champion team in soccer”. By concentrating on a small number of interrelated well-defined problems, research groups from all around the world cooperate and compete with each other chasing this final goal.

Robotic soccer competitions are organized each year in order to analyze the improvements made by the different research groups. Four major leagues, e.g. simulation, small-size, middle-size and legged league are the basis of RoboCup international competitions. Official competitions started in July 1997, in Nagoya, Japan. Followed by Paris 1998, Stockholm 1999, Melbourne 2000 and Seattle 2001, the annual events attracted many participants from all around the world. Following previous conferences and games, the 6th RoboCup will be held in Fukuoka, Japan in cooperation with Busan, Korea in June 2002. This RoboCup will coincide with the"2002 Soccer World Cup Korea/Japan". Today, more than 4000 researchers from more than 40 countries participate in several projects associated with RoboCup such as international games, conferences, research and educational programs. More than 600 participants and thousands of spectators are expected in RoboCup 2002 venue: Fukuoka dome stadium.

In addition to the simulation and robotic events, RoboCup International symposium provides a forum for the exchange of ideas and experiences between people involved in research in different leagues. Each year, the RoboCup federation grants two awards, e.g. the scientific and engineering awards, to papers presented at this symposium that demonstrate most potential to advance their respective fields of research.

Robotic Soccer simulation league is a very rich domain for the study of multi-agent real-time planning, coordination and communication and to study machine learning and opponent modeling. The soccer server is an excellent simulator of the soccer domain, in which software agents act autonomously, with very limited perception, limited action abilities, limited energy and a single unreliable communication channel with low-bandwidth for communicating. Two teams composed of eleven autonomous agents (plus a coach) need to collaborate in a real-time, noisy, adversarial environment to achieve two opposite goals: score goals in the opponent’s goal and prevent the opponent team from scoring. Coordination between the agents must be achieved before the game start through common knowledge and during the game through communication and reactive reasoning to the sensed information.

The Robotic leagues offer several different environments to study the robotic aspects of RoboCup. In the small-size league, two teams composed by five robots play on a green carpet field of the size of a ping-pong table. The rules permit that a camera may be placed on top of the field giving the robots a global vision of the field. During the game, a central computer receives vision information and uses wireless communication to control the robots. No human intervention is allowed during the game. Small-size league games are very exciting to watch because the robots and the ball move very fast on the field. This need for speed and accurate control of the robots gave this league a reputation as the engineering league.

The middle-size league poses a unique combination of research problems. Teams composed of four completely autonomous robots play in a green field with 9x5 meters, containing two goals with different colors (blue and yellow) and surrounded by white walls. The robots must have a black body and follow rigid constraints regarding its shape, height and size. All sensors and actuators of the robots must be carried on board and no global sensing system is allowed. Problems that are trivial in the small-size league like self-localization, finding the ball or localizing teammates and opponents are very complex to solve on this league.

There are several applications for RoboCup research and technologies. These include battlefield scenarios, rescue scenarios, real/simulated sport competitions, mine clearance, land exploration, control of hospital robots, public transports coordination, and implementation of AI opponents for simulation games. One of the major applications of RoboCup technologies is a search and rescue in large-scale disaster. This way, the RoboCup Federation initiated RoboCup Rescue project to specifically promote research in socially significant issues. RoboCup Rescue is also divided into a simulation and a robotic league. Disaster rescue is one of the most serious social issues, which involves a very large number of heterogeneous agents in a hostile environment. The intention of the RoboCup Rescue project is to promote research and development in this socially significant domain at several levels involving multi-agent team work coordination, physical robotic agents for search and rescue, information infrastructures, personal digital assistants, a standard simulator and decision support systems, evaluation benchmarks for rescue strategies and robotic systems.

Portuguese research groups are between the most successful in RoboCup international competition. Starting with a third place in Paris 98 small-size league accomplished by 5dpo (from FEUP), the success of Portuguese teams continued with two Europeans (2000 and 2001) and a world championship (Melbourne 2000) for FC Portugal (Univ. Porto and Aveiro) in the simulation league. Also researchers from the IsocRob team (from IST) achieved the engineering challenge award in Melbourne 2000 and 5dpo achieved 1^st place in the small-size league and 3^rd place in the middle-size league in the German Open 2001.

The tutorial “The RoboCup International Project: AI Meets Robotics to Create RoboSoccer Teams” aims, by grouping the most notable Portuguese RoboCup researchers in a single forum, at promoting the research in AI and Intelligent Robotics using RoboCup as the test bench. Different research perspectives will be shown by Portuguese teams together with videos, demonstrations and a simulation league demo tournament, in order to give a global perspective of the RoboCup international initiative.

December, 2001

Pages maintained by Luis Paulo Reis
For problems or questions regarding the Tutorial contact [Luis Paulo Reis].
Last updated: Dezembro 17, 2001.