PETRI NET PLAN COORDINATION FOR ROBOCUP TEAMS
Georgios D. Kontes, Michail G. Lagoudakis · 2008
For more than a decade, the international RoboCup competition promotes scientiflc research in robotics through soccer games between teams of autonomous robots. Efiective coordination between the players of a team is a crucial ability for the success of a soccer team. Unfortunately, research in non-simulated RoboCup leagues has focused mostly on single player skills, demonstrating only limited results in coordinated team play. Team coordination in a human soccer game boils down to various team formations, tactics, and strategies. In our work, we adapt and transfer formations, tactics, and strategies from human soccer to our four-legged RoboCup team. We deflne roles for each player in each tactic and we implement these roles using Petri Net Plans (PNP). The choice of tactic and the assignment of the corresponding roles to players is performed dynamically during the game depending on the current game state using a simple broadcast-based communication scheme and a flnite state machine. Our approach is implemented and tested on our four-legged RoboCup team Kouretes. The proposed coordination scheme can be easily expanded to include additional tactics and can be adapted to RoboCup teams of various size. We believe that Petri Net Plan Coordination can be useful in various robot team applications beyond robotic soccer, such as planetary exploration and search-and-rescue missions.