# FactoryOps

- **Event:** [Zero to Agent: Vercel x Deepmind Hackathon SF](https://cerebralvalley.ai/e/zero-to-agent-sf)
- **When:** Sat, Mar 21 at 9:00 AM – 10:00 PM (PDT)
- **Where:** Shack15, San Francisco, CA
- **Team:** [ARJUN HALDIYA](https://cerebralvalley.ai/u/haldiyaarjun), [Rahul Koushik](https://cerebralvalley.ai/u/Fluffy18), [Alex Jean](https://cerebralvalley.ai/u/ajean757)
- **GitHub:** https://github.com/Fluff18/FactoryOps
- **Demo video:** https://www.youtube.com/watch?v=XzBC-LU6npU
- **Gallery:** https://cerebralvalley.ai/e/zero-to-agent-sf/hackathon/gallery
- **Page:** https://cerebralvalley.ai/e/zero-to-agent-sf/hackathon/gallery/21

FactoryOps is an AI-powered robotic assembly failure investigation platform. When a robot arm crashes, traditional debugging means manually reviewing hours of sensor logs. FactoryOps eliminates that with two phases:

Live Monitoring — A real-time factory floor dashboard synchronizes video, 6-axis force/torque readings, audio waveforms, and 7-DOF joint state data on a shared timeline. Gemini auto-generates operation segment labels and caches incident summaries per recording.

AI Investigation — When a crash is detected, FactoryOps dispatches a multi-agent system (Gemini 2.5 Flash). An Orchestrator coordinates four specialists: Vision (sends native video/mp4 clips via ffmpeg directly to Gemini for true temporal analysis), Force-Torque (6-axis spike detection), Audio (acoustic event clustering with structured CANDIDATE_WINDOWS output), and Robot State (trajectory deviation). Agents share a compressed sliding-window memory to prevent context explosion. The orchestrator cross-correlates evidence across all modalities to produce a root cause, confidence score, and recommendations — streamed live to the browser via Vercel AI SDK's Agent UI, with a ReactFlow agent graph animating in real time.

---

Markdown version of https://cerebralvalley.ai/e/zero-to-agent-sf/hackathon/gallery/21. Site index for agents: https://cerebralvalley.ai/llms.txt · full text: https://cerebralvalley.ai/llms-full.txt