Front-left view of the robot.
The team behind the robot "Evil Shenanigans" consisted of Craig Mielcarz, Zach Lavalley, and Scott Bradley. The building and coding tasks were divided among the team members. Design of both code and the mechanics of the robot was carried out by all of the members.
Strategy
The strategy we decided on was to first and foremost attempt to place a ball into the cup. If that failed, the robot would then resort to placing as many balls as possible on the middle of the table. Initially, we wanted to throw opponents' balls off of the table. However, after several attempts to make a throwing mechanism that would work with the heavy balls, we decided to instead make a robot that could dump enemy balls off of the table. Because of our change in treatment of enemy balls, we only had to pick up an enemy's ball to a height over two inches (the height of the wall on the table).
In the first seconds of the match, the robot would head up to its own balls and sort through at least three of them to find one of its own balls. The plan was then to have the robot line-follow down to the cup. Once in the lava pit, the robot would lower a battering ram from its back to be used in removing obstacles from the cup, like other robots or another robot's ball already in the cup.
Physical Design
"Evil Shenanigans" was designed and built around the most complicated part of its mechanical structure, the lift for the balls. The gear train driving the lift is actually located in the center of the robot under the ball examination house. A single motor drives a 125:1 gear train, which drives an axle connected to two pulleys that raise and lower the lift. To sense when the lift has reached the top or bottom and avoid overdriving the lift, switches were used along with elastics to slow the lift at the bottom and top of its motions.
To capture balls, a series of wheels spin at the front of the robot in order to scoop balls onto the lift and keep them on the lift. By cutting a triangle at the back of the lift, we assured that no balls would get stuck on the lift since they would naturally fall into the ball examination chamber. We found that by leaving the lift only about half-way up, we could easily carry two balls at the same time, which allowed for greater examination speed. One ball would be in the chamber, the other waiting on the lift.
The turning mechanism was accomplished using zero-radius turns (that is, the drive wheels were used to turn the robot, elminating the need for a differential). The front wheels on the robot are castors, much like those found on shopping carts. The drive wheels were each driven by two motors on a 75:1 gear train. The medium gear ratio provided speed, but we used two motors on each wheel to give greater torque for moving the heavy body (plus potentially two balls) of the robot.
Because of size requirements (fitting the robot into a 1 foot cube), we made the battering ram retractable. A servo is geared up in order to allow a pulley to raise and lower the arm of the battering ram. The battering ram is sized so that when up or down, a game ball can fit through it. This meant that the battering ram was useful for stabilizing balls on the middle plateau or guarding a ball we placed in the cup.
Finally, the gate mechanism to control releasing balls from the ball examination chamber was achieved by simply attaching an axle to a servo to be raised or lowered.