Combat Robotics

I had been designing FTC robots for two seasons and wanted a problem that punished different mistakes. In FTC, a mechanism fails by being slow or inconsistent. In combat, it fails by coming apart.

So a friend and I started a beetleweight team and entered BCRC events. Everything about the robot — plate thickness, fastener choice, how the weapon mounts — had to be argued against a 3 lb limit that leaves no room for a part that is heavier than it needs to be.

An FTC robot is designed around cycle time and repeatability. It gets one alliance partner, a known field, and no opponent actively trying to break it. The right answer is usually the mechanism that scores fastest with the fewest ways to jam.

A combat robot is designed around energy and survivability. The weapon needs enough stored energy to matter, which means mass at a radius spinning fast, and every gram spent there is a gram not spent on armour or drive. The robot also has to absorb the reaction to its own hits.

That flips the design process. In FTC I optimised for precision and speed. Here I optimised for load paths — where does the impact go, what does it pass through, and what fails first when it arrives. Fasteners want to back out, printed parts want to split at layer lines, and anything cantilevered is a liability.

It also changes what "done" means. An FTC mechanism is done when it works reliably. A combat robot is done when it can lose a fight and be repaired between matches, which pushed me toward a design where the weapon assembly and the drive could be taken apart independently.

The weapon is a spinning bar driven by a brushless motor through a belt reduction. The motor sits inboard, low and central, and drives a pulley on the weapon hub, so the heaviest rotating part is supported by bearings in the frame rather than hung off the motor shaft.

The chassis is a machined plate structure with printed internals. Putting the weapon on a belt rather than direct-driving it means the motor is isolated from the shock of an impact — the belt slips or the pulley takes the abuse before the motor bearings do.

The blue weapon bar is the sacrificial part by design. It is the piece that makes contact, so it is the piece that gets replaced.

Two fights at BCRC events, and the pair of them covers most of what combat robotics actually teaches you.

The first is our first KO. Everything worked the way it was drawn — the weapon spun up, the drive put us where we wanted to be, and the hit ended the fight.

The second is the one I got more out of. Our weapon system broke mid-match. We had about an hour before we were due back on the floor, so we redesigned the failed part, printed a new one, and iterated it right there between fights — and made it back in for the next match.

That hour changed how I think about designing anything that has to survive contact. A part is not finished when it works; it is finished when you can diagnose it, replace it, and be running again with only the tools that fit on the table you brought.