Autonomous Soccer Playing Robots

In college, I was on the Georgia Tech Robocup team in the Small Size league. I worked on the electronics team where I helped build a fleet of robots that autonomously play soccer. Here’s what they look like:

Normally, colored circles of paper go in those slots on the top. There are two cameras above the field that we use to track the robots, and we encode the team, robot ID, and direction the robot is facing based on the color pattern on top. The shell you see above is 3D printed. Here’s what it looks like underneath:

The electronics consists of 3 boards. The top one you can see in this image is the halo antenna for communicating over a 900 MHz wireless link. Underneath that is the control board with an FPGA and small CPU, along with the motor drivers. Under that board is the kicker board, which has 2 large capacitors for firing a solenoid to kick the ball. We design and assemble all electronics in house.

This is from competition in Istanbul, Turkey in 2011.

This is from our more recent competition in Mexico City in 2012.

This is the bottom of our control board. This side primarily holds the FETs for the motor drivers.

There is waaaay too much information to put it all here, so if you want to see more details on the boards we designed and AI we’ve coded and mechanical design, we have all of that documented on our wiki.

Here’s an interesting view of our software evolution over time: