Team Convenience: Little Eyes-Eyes
Tim Mickel, Elaina Chai, Chris Tam
About • Strategy • Setbacks • The Competition • Lessons Learned
» About us
Team members:
- Tim Mickel
- Class of 2014
- Course 6-2
- Elaina Chai
- Class of 2012
- Course 8, 6-1
- Chris Tam
- Class of 2014
- Course 8
» Strategy
The goal of this contest was to push as many golf balls (1 point) and tennis balls (3 points) across the barrier as possible. We originally intended to tackle this challenge as follows:
Initial strategy
- Elevated robot
- Robot drives over balls, golf balls collect into the center of the robot
- Robot drives to the rift, expels golf balls using a moving wall powered by a servo
- Robot finds tennis balls and lifts them with a scoop
- Robot might find the rift bomb using the same mechanism as the tennis balls
New strategy
- Very light, fast-moving robot
- Only focuses on golf balls, since this is 8 points and other teams were struggling to get tennis balls
- Get as many golf balls across as possible as quickly as possible, ignore everything else
» Setbacks along the way
- Burnt-out motor: Soldering too long caused one of our motors to be burnt out in the first week. It was replaced.
- Mechanical strength of wires/connections: The given wires were sometimes weak and flimsy; we used stronger wires, more hot glue, and shrink tubing to keep our wires attached to our components.
- Keeping a strong structure throughout the month: As we added components over the course of the month, our lego structure became messy and clumsy. We had to rebuild much of our structure before the competition.
- The pattern: the vision positioning system was very sensitive to minor variations in the fiducial pattern's placement, coloring, surrounding objects, etc. We created a solid system by mounting our pattern on a hinge.
- Gears: We discovered that we needed a much lighter robot or else the gears locked up. We also had to be very careful about the placement of the encoders or else they got stuck in the gears. The gears also shifted over time, and so had to be readjusted frequently.
- Being careful with received data: We faced some issues such as misinterpretation of data, e.g., the gyroscope data was being sent in values ranging from -2048 to 2048, not -180 to 180.
- Faulty, but believable, data: Early in the competition, we had a cold solder joint which led to our program being fed poor gyroscope data that seemed to be just poor calibration. Switching ports changed this.
- Testing time: Lack of time at the end of the competition meant our navigation code was buggy and failed frequently.
» The Competition
Before the competition
- During one of our last software updates on the morning of the competition, the HappyBoard USB connector broke off our HappyBoard.
- Unable to upload code, we were given a spare board with a faulty radio.
- We were eventually given an old board that worked, although this board did not seem to perform as well as our original board (e.g. the motors moved slower)
- As a result of this incident, our team missed the seeding round and we were seeded last.
- Little Eyes-Eyes consistently scored points in every round.
- We won our first match!
- We were knocked into the loser's bracket after facing a particularly good robot, The Outlaws, which scored tennis balls (they later won 3rd place).
- In the loser's bracket, we still knocked out several teams, scoring 3-4 points each round with just golf balls.
- Eventually, we lost to Ballzooka, who eventually placed 4th.
» Lessons Learned
- Get your mechanical design solid fast, because getting controls to work is the most important and hardest part of robot-building.
- Always check if the problem is a physical problem, not a software problem.
- Debug, debug, debug.
- Always underestimate the degree of accuracy of your robot.
- Robustness is more important than cleverness.
- Robot-building can be a lot of fun!
Massachusetts Institute of Technology | 6.270 | © 2011 Tim Mickel, Chris Tam, Elaina Chai