Determine which AI solution to use for print failed detection
Closed this issue · 0 comments
ipepe commented
Determine which AI solution to use for print failed detection.
Preferably we can have multiple adapters. If multiple are specified then we can have these strategies to pause print:
- Fail fast - if any of AIs determine failed print
- unanimous vote - if all AIs agree that print failed
- majority vote - if more than half of AI agree that print failed
- tiered - this is mostly for performance/cost saving reason, idea is that we use local AI to determine failure with very low threshold and ask "smarter" (probably cloud AI) to evaluate if that's the case. Only if higher tier AI agrees that print failed, we pause it.
Suggested solutions:
- https://github.com/printpal-io/OctoPrint-PrintWatch
- https://github.com/Manicben/3DPrintSaviour
- https://cloud3dprint.com/ai-failure-detection/
- https://github.com/kennethjiang/TheSpaghettiDetective
- https://obico.io/
- https://github.com/fiberpunk1/OctoPrint-Nexus-AI
- https://github.com/mikulash/OctoPrint-Detector2
- https://octoeverywhere.com/
- https://github.com/QuinnDamerell/OctoPrint-OctoEverywhere
- https://github.com/antirez/failed-3d-prints-bot