What is this project?

The purpose of this project is essentially to make a robotic version of a connect four board game with a couple modifications. One, we are going to increase the resolution of the board. Two, we are going to make the top profile of the board a circle and put all the mechanisms and pieces on the inside. The pieces of this connect four board are 6mm plastic balls that are traditionally used in Airsoft.

What is the current state of the project?

I completed system validation testing and only have some software left to write.

How do the balls move through the display?

Only balls in one column are modified at a time. The entire outside of the display spins so the robot can select that column. Once the column is selected, a bunch of submodules work together to empty the column and refill it with the correct balls before continuing on.

What made version two more successful that version one?

Mostly I did this by becoming less of a perfectionist and more of a realist. Before I couldn’t see the forest for the trees. This kept me from thinking about the problem at a systems level. For version 1, I would take a month to improve a part that should not have existed or had a burdensome requirement that was misguided.

An example of this evolved thinking was deciding that the dispenser should go above the column and not be separated by the elevator. The dispenser to the column is the only portion of the whole system where the order of the balls matters and therefore they have to be output single file. As a result of this decision, I removed the need for the elevator to receive balls single file, which meant none of the other mechanisms needed to output them single file. This decision removed 5 actuators, several bits of needlessly complex fragile geometry, reduced programming complexity, and increased bandwidth of the elevator.

This is one of a few decisions that I made this time around that made the project easier to complete and made the final result better than before.