Team 47: Perrin

Designed my Mark Tompkins '02, Raj Dandage '02, and Naveen Goela '03

We are still developing photos of our robot and hope to have them on this site soon!


Background

OK, so you're probably asking "why the name Perrin?" All three of us went to high school together at Phillips Academy, Andover where we had the pleasure of having Dr. Robert Perrin as a teacher. Dr. Perrin, an MIT alum himself, teaches the higher level math and physics classes at Andover, and is by far one of the smartest people that any of us have ever met. He is an excellent teacher and was part of what inspired us to come to MIT, so as a tribute we decided to name our robot after him.


Our single victory against King Louis in the final rounds.

Design

Perrin was originally designed to get points quickly and then possibly crash into the other robot, so we made it powerful and compact, with a very solidly braced structure that could withstand impacts. We used two servo's to steer the two back wheels, and drove each of the two front wheels independently. Driving straight wasn't a problem once we got the servo's calibrated correctly, but for turning 90 degrees we used shaft encoders to count the number of rotations on each wheel. Perrin had six touch sensors mounted on it: two on the front, two on the back, and two roller sensors on the sides to sense walls and help roll along them. Also, to collect blocks, we mounted a garage door-type mechanism in the middle of the robot so that it could roll over blocks from either direction and pull them back.

When we first started designing our robot we didn't really know where to start, so we tried to get some ideas from the robots in last year's competition. We were very impressed by Team 1: Hermes' design which was both fast and powerful, so we also put four motors on each wheel anticipating collisions with other robots. Unfortunately, speed didn't end up being a big factor in this year's contest, and most robots used a defensive strategy like ours which resulted in very few collisions, so we ended up running our motors at less than half speed most of the time to make our robot more accurate. Let this be a lesson to anybody thinking of using the winning strategy/design from past years: it doesn't always help, since every board is different!

Strategy

We spent many hours during the first week discussing our strategy, trying to keep it as simple as possible while at the same time not leaving us vulnerable to other common strategies which our opponents would be using. We came up with many ideas, including plowing down the middle and pushing 1 or 2 professors in our opponent's jail, or going to the elevated area and trying to pull professors back onto our side. The problems with these were that if an opponent implemented a similar strategy, both robots would get stuck in the middle and we wouldn't score any points, and might even inadvertently give them some. We finally decided it would be better to go for the two hackers situated next to our jail first, because if we could reliably get them both, we'd be guaranteed 8 points even if the other robot got in our way or our robot messed up while going towards the center of the board. After getting the hackers, we had our robot go down one side of the board and attempt to pull back one of the professors, scoring us a total of 11 points in a perfect run. Afterwards, we were planning on having our robot go to the opponent's side and sweep their jail, but during testing we found that our robot would often accidentally push students or a professor onto our opponent's side if we went there, so we decided to leave that routine out and focus on making our robot as reliable as possible instead.

Results

Perrin placed second in this year's competition due to its reliability and consistency, and a little bit of luck with some of the match-ups. We were also the only robot in the competition to defeat the otherwise ultra-reliable winner, King Louis, in any round, due to the fact that they accidentally missed two professors and we got both of our hackers.