RE-HOBI362 Rachel Hoffman-Bice, PhD
Project

Slalom Course Robot

A team robot that ran a slalom, stripped a rack and cleared seven tennis balls without knocking anything over

Org UF
Period Summer 2012
Location Gainesville, FL

A team robot that had to run a slalom course, pull a bucket off a rack, lift seven tennis balls off a three-tiered stand, and drop them all in the bucket — without knocking anything over.

Built from a stockroom of 80/20 extrusion, PVC pipe and small gearmotors, this was where I first went through a full, formal design process: run a weighted decision matrix before committing to a concept, build a prototype, watch it fail in a way you didn’t expect, and fix it without starting over from scratch.

Working in a five-person team (Team 1A), I mostly owned budget tracking and helped run the concept-selection process, on top of the general hands-on build work everyone did.

Concept selection

Before anyone touched a hacksaw, we scored six candidate drivetrain combinations against speed (30%), controllability (40%), size (20%), and modularity (10%) — controllability got the highest weight on purpose, since knocking over a single tennis ball or bucket wiped out any time bonus. The matrix pointed us toward 13.6” drive wheels paired with 44-RPM right-angle gearmotors: fastest of everything we scored (~1.95 ft/sec) while still controllable enough with differential steering.

The manipulators

The ball-pickup mechanism is a length of 3” PVC pipe with a slot cut into the underside, mounted on a motor so it swings down onto a tennis ball, forces it up through the slot, and lets it roll down into a bucket the robot is carrying on its front.

PVC ball manipulator and drive wheel assembly, close up

The bucket manipulator is what actually taught me something about iteration. Original plan was plywood gripper arms — simple enough, until we realized there was no good way to attach a wooden shaft to the drive motor without splitting the wood or stripping a set screw over a few runs. We scrapped it before it was ever fabricated and rebuilt the whole manipulator out of 80/20 extrusion with shaft couplings on each end instead.

Assembling the ball manipulator arm in the lab

Here’s the route we planned around the ball stand and release bucket:

Course route diagram from Appendix F

On competition day the robot grabbed all seven tennis balls and made it around the bucket without knocking anything over.

Competition run.