Project WARP: Teaching a Robot to Paint an Army

I have apparently decided that painting miniatures by hand isn't enough of a challenge.

The latest project on my workbench is Project WARP — Wargaming Automated Robotics Painter.

The idea is relatively simple: build a robot that can base-coat wargaming miniatures.

Not paint eyes. Not paint buttons. Not produce competition-level display miniatures. At least not yet.

The first goal is much more modest: take a 25mm miniature, put it on a painting station, and have a machine apply a consistent base coat.

Why? Because I paint armies, not individual figures. There is a certain point where painting the 47th infantryman in exactly the same way as the previous 46 stops being particularly interesting.

And that makes it an excellent automation problem.

Start With Vision

The first version of WARP may not even have a robot.

I already have a computer, Python, and a webcam. That's enough to start teaching the software what a miniature looks like.

The first experiment will be incredibly simple:

"There is a miniature here."

From there, WARP needs to determine where the figure is, how large it is, and eventually which direction it is facing.

Once we can reliably locate a miniature in a camera image, we can start translating those pixels into physical coordinates.

That is the beginning of the robot.

Then We Add Motion

The eventual machine will probably use a small CoreXY-style gantry, similar in concept to a 3D printer, but instead of carrying an extruder it will carry an airbrush.

The miniature itself will sit on a small rotary painting platform.

The basic idea is:

Airbrush moves. Figure rotates. Paint gets applied.

Keeping the figure stationary during painting should make the system considerably easier to control. A separate loading mechanism can eventually move figures from a storage tray to the painting station.

Why 25mm?

Because I'm not crazy enough to start with 10mm.

At 25mm, there's enough surface area to experiment with coverage, spray distance, pressure, rotation, and positioning without immediately running into the microscopic problems involved in painting a 10mm figure.

If WARP can reliably paint a 25mm infantryman, we can make things smaller later.

The Bigger Idea

I'm deliberately not trying to build the ultimate miniature-painting robot on day one.

WARP will be developed incrementally:

Vision → Motion → Airbrush → Painting → Automation

Each stage should produce something useful and teach us something about the next.

Eventually, I'd like to see whether WARP can take a tray of unpainted miniatures and turn them into uniformly base-coated figures with minimal human intervention.

And if it works?

Well, then we can start asking the more interesting question:

How much of an entire wargaming army can a robot paint?

That's probably where this project gets dangerous.

In the best possible way.