A dark circular minimap render showing a rotated floor plan in blue grey and cream, with a small green triangle marking the viewer's position and heading

Personal systems projectPart 2 of 3Drawing 15.2 of 37

Finding people

Four ways to find a person through a wall, tracking where the gun is pointing in the dark, and a map that doesn't turn a person standing still into a wall.

StatusIn development, 2026
WrittenAbout 129,000 lines of hand-written C++17 across 15 subsystems
Desktop tests105 run, 105 pass, on a laptop with no sensor attached
Not run on hardwareZero lines. Every performance figure below is a budget
OwnedThe helmet display. Everything else is still unbought
TargetJetson Orin Nano 8GB worn on the body, Raspberry Pi 5 on the robot

Sheet 02 of 04/Through walls

Four ways to find a person through a wall

I find people through walls four separate ways: ultra-wideband ranging, acoustic direction finding, radar and passive radio. Relying on just one of them would mean a single failure mode.

Sheet 03 of 04/Gun tracking

Tracking the gun

Four infrared LEDs on the gun and a filtered camera on the helmet give a drift free read on where the gun is pointing in total darkness. When the LEDs are blocked, the gun's own motion sensor backs it up. The HUD shows the aim as a cone that grows with distance instead of a dot, because the system doesn't know exactly where the gun is aimed, and a dot would pretend it did.

Sheet 04 of 04/Mapping

Mapping the arena

As the rig moves, it builds an occupancy grid of the arena and merges it with what teammates are mapping. If a person stands still, the map keeps treating them as a person and doesn't add them as a permanent wall.

An occupancy grid render on a near black background, with white walls, dark blue grey floor regions and a thick magenta shape running across the middle — *PL 05*The same grid without the fog. Magenta is everything the map has decided is moving. It won't freeze any of that into a wall, no matter how solid it looks right now.