October 14th, 2000: UC Riverside

There were four mice in the competition:

 

Dirty Rat

MicroMouse #1: "Dirty Rat", CSU Long Beach

A mouse based on the Boe-Bot kit from Parallax, Inc., it used the Basic Stamp II as its microcontroller. It had a good mechanical design which allowed it to make sharp, consistent turns and it performed well. Unfortunately, due to the limitations of the microcontroller, the algorithm it used was not very sophisticated and so it had a tough time finding the center of the maze. CSU Long Beach also showed off a second MicroMouse that it was working on. This robot, they explained, is going to have sensors that look at the maze walls from the side and use distance measuring sensors to determine how far it is from the wall.

 

 

Lego Mouse

MicroMouse #2: "Lego Mouse", UC Riverside

This mouse was based on the Lego Mindstorms kit. My friend described it best when he called it a robotic moth as it both looked and behaved like an insect. Lego Mouse was designed to be a wall follower so it would meander its way down the paths of the maze. When it lost sight of the wall, it would spin around wildly until it found another wall. This one was fun to watch and was a real crowd pleaser.

 

MicroMouse #3: Don't remember the name, UC Riverside

A more sophisticated robot from the host school, this robot was based on Motorola's 68HC11. I'm not sure if they were having hardware problems, software problems or both, but this mouse had some trouble moving throughout the maze.

 

MicroMouse #4: "Dexter", CSU Northridge

Our MicroMouse is also based on Motorola's 68HC11 and uses the flood-fill technique of solving a maze. Because our mouse actually looks like a mouse, it was also a crowd favorite. Throughout the competition, people kept referring to it as "the cute mouse".

Also in attendance were MicroMouse groups from UCLA and CSU Los Angeles. They were still designing their mice so they came to observe.

The contest started at 1:00 and consisted of two rounds.

 

Round 1:

The first mouse up was Riverside's 68HC11 based mouse. I'm not sure what happened, but it went forward 5 cells and it came to a halt.

The next mouse was Long Beach's Dirty Rat. I was very impressed by how it maneuvered throughout the maze. The mouse was able to make its way through all four quadrants of the maze. At one point, it went right by the entrance to the destination cells but it chose not to make the turn. It was in the maze a good 5 minutes before it got wedged against a wall.

Our Dexter was up next. It did its best to roll its way through the maze but it encountered a problem during its search. Riverside's maze is designed to be taken apart and consisted of 4 large sheets of particle board. The joints between these sheets, however, were not even and contained steps as large as 5mm. This was too large and our mouse' chassis got hung up.

The final mouse was Riverside's Lego Mouse. It bobbed and danced its way down the passages of the maze for 4 minutes. At one point it lost one of its Lego pieces and even managed to knock down one of the maze walls!

 

Round 2:

Because the first half of the maze contained such large steps, everyone agreed to start round 2 in the opposite corner of the maze. CSU Long Beach's entry started the second round. It made another valiant attempt at exploring the maze but its algorithm caused it to pass the destination cells once again. After 2 minutes, it started climbing a wall and had to be rescued.

Riverside's 68HC11 robot was up next. It went 3 cells before it hit a wall. A Riverside member pushed it back on its way and just when everything seemed to be going well, it started dropping screws on the maze floor! It looks like a good robot and as soon as Riverside works out the problems I'm sure it will be a good performer.

We finished up round 2. Dexter made his way through the twists and turns of the maze and found himself in a dead end right next to the destination cells. He turned around and eagerly tried to find the opening to the finish. After two more dead ends he had mapped enough of the walls to know exactly where the opening was. As he made his way closer, the audience started to cheer. Then, after 2 min. and 10 sec., he managed to cross the finish line! I never realized I could hold my breath for two minutes.

Group Picture of the Participants