For this years robot we decided to use the "bumble bot" (our name for the black and yellow robot we built last year) because it fulfilled most of our robot design goals which are listed below:
We wanted a simple design that is easy to understand. Fewer things can break and go wrong, it is easier to build and duplicate, and it requires fewer parts.
We wanted to build a robot that was compact or small in design so it would make tight turns, fit in between the wall and the solar panel and the wall and the MAV launcher, have a low center of gravity and not collide with obstacles. Also, so it would be lighter and faster.
The robot needs to be bombproof and strong because we can't afford to have pieces breaking off or requiring repair during missions if it collides with an obstacle.
The robot design needs to be reliable so we can predict it's behavior and have consistency in performance.
Our next goal was a modular robot that would be suitable for easily connecting attachments. We have one robot base and multiple attachments that use the same interface and can be quickly snapped on and off.
We wanted to be able to take attachments off in 2 seconds and put another on in 2 seconds because of the time allotment to complete missions. The transitions need to be fast.
We preferred a robot design that would use feedback orientation instead of odometry or dead reckoning because feedback is more reliable and less prone to errors. We don't want the robot to get lost. Also, we wanted to avoid requiring precise initial alignment, so we programmed the missions to allow some flexibility in starting position.
We need the robot to have an all terrain option, so we can get the robot into the crater at the far end of the table. Our original robot base was not able to get into the crater because the light sensors on the front would get stuck on the ledge.
We wanted the robot to be able to turn accurately and consistently in order to have reliable missions. If the turns are not consistent, mayhem results!
We needed to upgrade our robot in a couple ways so we made it skinner to turn through tight turns.
As we continued to test our robot and program it we needed to have our light sensors to be able to always read the same thing every time so we created a program to always read the darkest and lightest every time. We did this because we didn't want it to read something different every time because of the battery levels changing.
This year as we designed our robot, built our robot and programmed our robot to do the missions we had to repeatedly test the missions out to see if they accomplished the tasks and if we needed to change something to make it better.