Echo - The Mini Sumo Robot

The UW Robotics team is holding a school wide mini sumo robot competition. For $50, they provide most of the construction materials (they gave us a PCB which we didn't use, sheet metal, and all the sensors) and lessons on programming, Solidworks, and soldering. The prize for winning the competition is $250.

We are a team of four first year students. Our sumo robot uses two Faulhaber motors and Banebots wheels for its drive train. The brain consists of a ATMEGA168 microcontroller. Toshiba TA8050P chips are used to drive the motors. Sharp IR range sensors are used for object detection, and Fairchild IR emitter/phototransistor pairs are used for line detection. Everything is powered by two 9V batteries wired in parallel. The chassis is constructed from sheet aluminum.

Echo is longer than the size restrictions so it starts up in a standing position, a quick reverse movement causes it to drop down.

Software is writen in AVR-GCC using AVR Studio. Our strategy is very aggressive. The two object sensors both look forward but 30 degrees to each side and an inch into the robot for optimal detection.

Competition - We Lost and Won

Team Circle of Current placed 5th in the fighting part, we needed to be in 4th to be in the finals. Still not bad considering all the problems we had.

We won the judges award, 3 professors inspected our robot and listened to our presentation, and our team was the favorite. We're getting some sort of award at a dinner later.

We had terrible luck as soon as the morning started. We found out the hard way that Faulhaber motors are not meant for robotics at all. The gears are attached to D shafts with 0.9 mm hex grub screws, and nobody sells 0.9 mm allen wrenches unless you special order them online. Our first motor's gear vibrated loose, and lacking any options, we replaced it with our remaining 2 motors, there goes $8. Then during more testing, the gearbox (not the right angled one, I mean the gearbox that's enclosed) on the other side completely failed, another motor gone. Finally, during the competition, another grub screw came loose, this one was accessible with a normal screw driver so we retightened after every match, but then later, the inaccessible screw was loose, and now the robot is basically crippled.

Don't use these motors unless you have a 0.9 mm allen wrench.

The event organizers also used different rings for the actual competition part, with a thinner boarder, which threw us off for one match before a quick firmware change had to be done to adjust for it.

A ton of other stupid problems also occured. One of the motor drivers would apply power to the motors when the AVR microcontroller wasn't even powered, which never happened before.