Skip to content

LundstedtAdam/Gantry

Repository files navigation

Gantry

Local AI monitoring and control for 3D printers — detect failures, watch prints in progress, and pause from your phone. No subscription, no cloud, no vendor lock-in.

Gantry dashboard showing live stream and status panel

Vision: Gantry is a free, open-source app that uses computer vision to catch print failures and lets you control your printer from anywhere. The goal: give older and budget machines the exact same capabilities as the expensive ones, without any cloud lock-ins.


Download

Windows installer — the easiest way to get started.

  1. Go to the latest release and download Gantry_x64-setup.exe.
  2. Run the installer and follow the prompts.
  3. Gantry starts automatically and opens the setup wizard in your browser.

Windows SmartScreen warning: Windows may show "Windows protected your PC" because the installer is not yet code-signed. Click More info then Run anyway to proceed. The source code is fully open — you can review it before running.

Early release (v0.1.0): Gantry is new and actively developed. Rough edges exist. If something breaks or confuses you, open an issue — every report helps.


What is Gantry?

Gantry watches your printer through a camera and uses an on-device AI model to spot failed prints — the tangled "spaghetti" mess that happens when a print comes loose. When it sees trouble, it sends a push notification straight to your phone through Ntfy. It works with any printer and any camera — a USB webcam or an RTSP network camera is all you need. Everything runs on your own machine and local network: there is no cloud service, no account to create, and nothing leaves your home.

It is also a print companion. Gantry tracks the print in progress — progress percentage, time remaining, and layer count — and notifies you when a print starts, finishes, pauses, or errors. On supported printers you can pause and resume remotely, and review your print history and statistics over time. Remote access is handled by Tailscale — a free, private VPN that connects your devices directly without any cloud intermediary.

The expensive printers already do this. A flagship Bambu Lab machine ships with built-in AI failure detection and remote monitoring — but it is tied to that hardware and that company's cloud. Gantry gives the same capabilities to everyone, for free: point a camera at an older Creality, an Ender, a Prusa, a Marlin-based machine, or anything else, and it gets the same eyes-on protection and remote control the premium printers charge for — with no subscription and no vendor lock-in.


Features

Detection & alerts

  • Spaghetti detection via a custom YOLO11n model
  • Works with any RTSP or USB camera
  • Detection zones — focus on the build plate, ignore busy backgrounds
  • Ntfy push notifications to your phone or desktop

Print monitoring

  • Live camera feed from anywhere (via Tailscale)
  • Print progress, time remaining, and layer count
  • Notifications when a print starts, completes, pauses, or errors
  • Pause and resume the print remotely on supported printers
  • Print history and statistics

Printer integrations

  • Bambu Lab (LAN mode, with optional pause-on-detection)
  • Moonraker / Klipper (Creality, Voron, and others)
  • OctoPrint — Prusa, Ender, and other Marlin-based machines
  • Any printer at all via camera-only monitoring

System

  • Desktop app (Windows; macOS coming soon)
  • Self-improving AI: review queue → train → hot-swap the model
  • Timelapse recording
  • Remote access via Tailscale
  • Installable as a PWA on your phone

Quick start

For contributors and developers building from source on Windows.

Requirements: Windows 10/11, Python 3.11+, Git.

  1. Clone the repository

    git clone https://github.com/LundstedtAdam/Gantry.git
    cd Gantry
  2. Run the setup script (creates the virtual environment, installs dependencies, and seeds your config)

    .\install.ps1
  3. Start the server

    .\start.ps1
  4. Open the dashboard and follow the setup wizard to configure your camera and printer:

    http://localhost:8080
    

No detection model yet? Gantry still starts and serves the setup wizard. Model assets are published on the Releases page — drop them into ml\models\.


Desktop app

Gantry ships with a Tauri desktop wrapper (Windows; macOS coming soon) that runs the server as a background sidecar and lives in the system tray. For development:

cargo tauri dev

See docs/DESKTOP.md for building the installer and the full desktop setup.


Printer setup

Step-by-step guides live inside the app's Help center. Once the server is running, open http://localhost:8080 and go to the Help section (reachable from Settings on mobile). It includes guides for:

  • Bambu Lab (LAN mode)
  • Moonraker / Klipper
  • OctoPrint
  • Camera (RTSP)
  • Remote access (Tailscale)

Screenshots

Dashboard Review queue Statistics
Dashboard Review Stats
Training Mobile dashboard Mobile nav
Training Mobile dashboard Mobile nav
Mobile review Mobile stats Mobile training
Mobile review Mobile stats Mobile training

Contributing

git clone https://github.com/LundstedtAdam/Gantry.git
cd Gantry
.\install.ps1

# Run the test suite
.venv\Scripts\python.exe -m pytest
  • See docs/PROMPT_GUIDE.md for the development workflow.
  • Export labelled training data with python scripts/export_training_data.py.
  • More on reporting bugs and submitting changes in docs/CONTRIBUTING.md.

Configuration

Settings live in config/settings.yaml. This file is gitignored (it holds your camera URL and printer credentials) — install.ps1 creates it for you by copying config/settings.yaml.example. You can edit it by hand, but the easiest way is the Settings page in the web UI at http://localhost:8080.


Feedback & bug reports

Found a bug? Have a feature idea? Open an issue on GitHub — it is the fastest way to reach the maintainer and the most useful thing you can do for the project.


Troubleshooting

Common problems and fixes are collected in docs/GOTCHAS.md.

About

Local AI monitoring and control for 3D printers — detect failures, get phone alerts, no cloud

Topics

Resources

Contributing

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages