Physics grades every shortcut.
Starting from “how do you make a robot?” is not a clean spec. The builds needed reliable mechanics, electronics, and enough control logic to make motion predictable.
Evidence file 06 / robot bay
SCIOLY Robotics
Science Olympiad robots built for real competition: mechanical design, electronics, and autonomous control across electric vehicle, Robot Tour, and hovercraft work.
What it was actually solving.
The messy version of engineering: robots that had to move fast, straight, accurately, and on schedule while surviving real competition conditions.
Prototype in public.
Build, fail, adjust, repeat. A robot can be ninety percent tape and still be good if it performs the task better than the polished machine next to it.
The machine exposes hand-waving.
In software, a bug can hide behind an abstraction. In hardware, the robot just veers left in front of everyone. Work compounds. Time without work does not.
Illustrative route map / not live telemetry
Code meets floor.
Every run closes the loop between a control idea and physical evidence: measure the miss, change one thing, and put the machine back down.
More from the bench.
Projects where inputs, outputs, and failure modes just happen to live inside a computer.