Introduction

This is the first year in development of the JHU Robotic Soccer Team. The JHU JayBots 2001 is the first JHU entry into the RoboCup Middle Size League (F2000). The strongest aspects of the team involve the use of a color tracking system that employs XVision2, a software library used for image processing, and our robot control architecture.

The problem of teaching robots to play soccer in a competitive manner is a difficult one to solve, fusing a wide variety of disciplines within robotics research including: multiple agent cooperation, learning, situation recognition, vision, control, and strategic decision-making. The tracking system located on each robot player will process visual images which will then be used to create a world model of the game in progress. The sidelined server will determine actions based on current game situations. These actions and situations will be internalized in the robot players. This will facilitate robot learning and as playing time increases, the robot players will become more autonomous.

This paper includes a team description describing the hardware design, robot control architecture, and the game playing strategy.