Inspiration
Every minute an aircraft is grounded (AOG) costs airlines thousands of dollars. From frontline experience in aviation maintenance and receiving inspection, we know that many delays aren't caused by complex engineering issues, but by supply chain bottlenecks—missing critical parts, misplaced tools, or delayed orders. We wanted to bring autonomous AI out of the chatbox and onto the hangar floor.
What it does
AeroNode is an autonomous, Gemini-powered MRO (Maintenance, Repair, and Overhaul) agent. It continuously monitors mock hangar inventory and flight schedules. When it predicts a critical shortage (e.g., O-rings, filters), it doesn't just alert the user; it actively prepares a procurement order and executes it via external APIs, utilizing a secure "Human-in-the-Loop" approval system.
How we built it
- The Brain: Google's Gemini models handle the reasoning and prediction logic.
- The Data Bridge: We use Fivetran (MCP) to ingest real-time mock inventory data and supplier lead times.
- The Audit Trail: In aviation, safety and traceability are paramount. We use Arize Phoenix (MCP) to trace every decision the agent makes, ensuring zero hallucinations in procurement.
- The Action: GitLab (MCP) and webhooks handle the automation of ticketing and the final order execution.
What's next
This is currently a prototype to demonstrate the architecture, but the end goal is full integration with standard aviation ERPs.
Log in or sign up for Devpost to join the conversation.