Strategy

Initially, our strategy was to have Oliver go around the board in a clockwise manner to pick up all the small red balls, deposit them into our goal, and then raise the flag box. We did not plan on handling the larger blue balls because they had not arrived at lab yet. Early on, we decided to avoid dead reckoning by calculating our position through the use of distance sensors. After days of testing, we found ourselves having a lot of difficulty getting the distance sensors to read accurately enough—no matter how many times we calibrated them—to effectively use our initial strategy.

The second iteration of our strategy involved grabbing the two small red balls between our starting box and our goal, depositing those two into the goal, sweeping the board for the remainder of the small red balls, depositing those into our goal, and then going to raise the flag box. We still did not plan on going for the large blue balls because we decided that the round was too short to focus on these balls. Unfortunately, the distance sensors still did not give accurate enough readings to focus on the steps beyond putting the first two red balls in our goal as the distances required for those steps were readings which varied by too much (even though they were well within the range of values of the distance sensors).

The strategy we ended up using in the competition was grabbing the two small red balls between our starting box and our goal, depositing those two into the goal, and then blocking our opponent's goal. Sadly, this wasn't very effective as the light in 26-100 appeared to have caused the distance sensors to read different values than in lab. However, in this competition, blocking the opponent's goal was not the strategy of the winning team, but did prove to be highly effective for many teams. In hindsight, even though using LED/phototransistor pairs to read lines on the board would not have been sufficient to drive our robot the entire way through the round, the dead reckoning necessary would have worked well enough—and definitely would have worked better than using the unreliable distance sensors.