AIsle0

Landing Page
Flight Crew Role Selection
Shifts Scraped from Workday (AA API)
3D Plane Model + Map
Role Specific Inspection Checkoffs
AI Hazard Tagging Audit Log
Entire Team's Inspection Tasks

✈️ Inspiration

AIsle0 was inspired by a simple but persistent gap in airline operations: workers assigned to the same flight often have no shared, flight-specific way to communicate. Ground crews, flight crews, and service teams rely on fragmented tools and assumptions, which can lead to delays and miscoordination.

We set out to build a platform that improves communication during aircraft turnarounds, making work smoother, safer, and a little less stressful for the people who keep flights moving.

🚀 What It Does

AIsle0 is a web and mobile platform that connects airline ground and flight crews by flight number, providing real-time visibility into pre- and post-flight task status.

By sharing live updates (such as fueling in progress, completed, or temporarily restricted), the platform helps teams:

Coordinate safely
Avoid equipment conflicts
Sequence work more efficiently during aircraft turnaround

🛠️ How We Built It

AIsle0 is a real-time web and mobile platform designed around flight-specific coordination during aircraft turnaround.

Architecture Overview

Frontend

Web: React, TypeScript, Vite
Mobile: React Native, Expo

Backend

Supabase (PostgreSQL, Auth, WebSockets, Edge Functions)
Render (API hosting)

AI Layer

OpenAI for analysis, planning, and risk tagging

Core Database Tables

`inspection_sessions`

Prop	Type	Default	Description
`id`	`uuid`	-	Inspection session identifier
`status`	`text`	`pending`	Overall inspection state
`progress_pct`	`int4`	`0`	Completion percentage
`flight_number`	`text`	-	Associated flight
`updated_at`	`timestamptz`	`now()`	Last update time

`inspection_step_instances`

Prop	Type	Default	Description
`id`	`uuid`	-	Step instance identifier
`session_id`	`uuid`	-	Parent inspection session
`status`	`text`	`pending`	Step status
`photo_path`	`text`	-	Uploaded image reference
`ai_severity`	`text`	-	AI risk level

`user_flights`

Prop	Type	Default	Description
`user_id`	`uuid`	-	Assigned user
`flight_number`	`text`	-	Active flight
`selected_at`	`timestamptz`	`now()`	Selection time

🔄 Real-Time Updates

Task progress and AI analysis are synchronized instantly across web and mobile clients using Supabase WebSockets, ensuring a shared operational view during aircraft turnaround.

⚠️ Challenges We Ran Into

User research at TAMUHack 2026 was challenging at first, as airline operations are a specialized domain with limited direct access to end users. After iterating on several ideas, we converged on a concept that balanced realism with feasibility.

On the technical side, implementing real-time updates and socket-based communication was difficult due to the structure and constraints of Supabase’s real-time features.

🏆 Accomplishments We’re Proud Of

Designed custom drawings and GIFs to clearly communicate complex workflows
Built a clean, scalable structure that connects web and mobile users
Successfully implemented real-time synchronization between airline workers

📚 What We Learned

Building AIsle0 taught us how complex real-world operational systems can be, especially when multiple teams work in parallel under strict safety constraints.

We also learned how to design and implement real-time systems under tight hackathon timelines.