# BoardPilot

- **Event:** [Meta Global AI Developer Hackathon](https://cerebralvalley.ai/e/meta-ai-developer-hackathon)
- **When:** Oct 26, 12:00 PM – Nov 4, 9:00 PM (EST)
- **Where:** Online
- **Team:** [Mauricio Tagliaferro](https://cerebralvalley.ai/u/mostro)
- **GitHub:** https://github.com/mostro3000/BoardPilot/
- **Demo video:** https://www.youtube.com/watch?v=QmhhvENuv38
- **Gallery:** https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery
- **Page:** https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/852

BoardPilot is a native Android copilot that helps technicians repair electronic boards at the component level instead of throwing them  away. It is aimed at field technicians and small repair shops, starting in Latin America, where boards from power supplies, TVs and   appliances are often discarded for lack of a diagnosis.
  
  It is a multi-step diagnosis agent, not a photo analyzer. The technician photographs the board and describes the symptom. SAM 3.1 maps the components; Muse Spark scans for visible damage, ranks competing fault hypotheses with probabilities and plans only the multimeter measurements that can change the decision. The app guides them one at a time, with a magnified view of the part, the meter setup and AI-suggested probe points, and updates the probabilities deterministically with every reading. When the evidence is not enough, it asks for another measurement or photo instead of guessing. When a part is confirmed, it suggests the replacement, Muse Image renders an illustrative reference, and the app checks whether the unit works, re-planning if it does not.
 
  Safety is decided by the code, not the model: unpowered measurements always come first, live ones are withheld while there is visible damage or an unrepaired fault, and every measurement sits behind a safety gate. Everything runs on the phone with the user's own Meta Model API key, with no backend of its own.

## More from Meta Global AI Developer Hackathon

- [AKINOLA OLATUNJI](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/849)
- [Andrea Conklin](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/850)
- [Chiheb](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/851)
- [Probe & Patch](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/853)
- [GTM Shortlist](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/854)
- [Daniel's Team](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/855)

---

Markdown version of https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/852. Site index for agents: https://cerebralvalley.ai/llms.txt · full text: https://cerebralvalley.ai/llms-full.txt