// the find
mkyharry/Warehouse-Inventory-Management-Robot
A university-project AGV that navigates a fixed warehouse map using line following, ArUco marker detection, and a dual-controller architecture (Raspberry Pi 3B + Parallax Propeller). It counts defective vs non-defective widgets at predefined stations. This is a one-off academic build, not a general-purpose library or framework.
The dual-controller split is sensible — offloading real-time motor/sensor control to the Propeller while Python+OpenCV handles vision on the Pi is the right call for latency reasons. ArUco markers are a practical choice for inventory identification: cheap to print, fast to detect, and OpenCV handles them well. The project scope is well-defined and the hardware list is specific enough that someone could actually reproduce it.
Zero stars, one contributor, last touched March 2024 — this is a submitted assignment, not an active project. The only Propeller code artifact is a pre-compiled binary plus source; there's no build script, no dependency pinning, and the setup instructions are basically 'clone it and good luck.' The Pi code is a single flat Python file with no mention of how the Pi and Propeller actually communicate (serial? GPIO?). No tests, no simulation mode, no way to run or verify any of this without the exact physical hardware.