Rerouted

Rerouted is a parallel tabu search-based optimization engine for rapid airline crew recovery during operational disruptions (flight delays, cancellations, etc). It leverages realistic mock flight data from the Flight Engine API and provides a robust modeling of airline operations, FAA-compliant rules, and end-to-end real-time tools for schedulers, crew, and operations managers.

Inspiration

In large-scale disruptions, reassigning airline crew is a major bottleneck. Crew schedulers must comply with complex FAA duty rules, location/certification constraints, and need to minimize cascading cancellations. Today, this process is manual and slow, often taking hours, worsening the passenger/operator impact.

Rerouted automates this process with parallel optimization, fast simulation, and a modern end-user UI for both web and mobile.

---

Overview

• Parallel Tabu Search: Multiple scenarios explored in parallel for fast, high-quality crew reassignments.
• Real-time updates: Leveraging WebSockets for scheduler/crew/ops dashboards.
• FAA-compliance engine: 14-hour max duties, 10-hour rest, aircraft certifications, and more.
• Rich mock data: Real AA flight data, synthetic crew/rotation/pairing info, realistic disruption modeling.
• Cross-platform: Next.js web/React Native mobile apps.

---

Tech Stack

Frontend:

• Next.js 14 (App Router, SSR)
• React, React Native (Expo)
• Tailwind CSS, Shadcn for UI
• Expo Notifications
• React Navigation

Backend:

• Node.js + Express for API server
• PostgreSQL for flight/crew/assignment data
• Redis for caching & real-time crew availability
• WebSockets (Socket.io)
• American Airlines Flight Engine API integration

---

Key Features

• Parallel Tabu Search optimization for crew reassignment
• CPU-efficient simulations of large hub-and-spoke airline networks
• Full modeling of real airline constraints (duty, connections, rest, certifications, pairing logic)
• Handles real-world disruptions – delays, cascading effects, cancellations
• Mobile & web dashboard with real-time crew status/notifications

---

Data Model

Aircraft

Column	Type
aircraftId	string
model	string
homeBase	string
rotation	string[]

---

Crew

Column	Type
crewId	string
role	enum
name	string
homeBase	string
certifications	string[]
seniorityScore	int
status	enum
currentDutyStart	datetime| null
flightTimeToday	int
dutyTimeToday	int
consecutiveDutyDays	int
lastRestEnd	datetime
currentLocation	string

---

Crew Pairings

Column	Type
pairingId	string
crewId	string
flights	string[]
dutyStart	time
dutyEnd	time
totalFlightTime	int
totalDutyTime	int
returnsToBase	boolean

---

Flights

Column	Type
flightNumber	string
origin	string
destination	string
aircraft	string
scheduledDeparture	time
scheduledArrival	time
duration	int
distance	int
typical_passenger_count	int
aircraftId	string
legIndex	int

---

Disruptions

Column	Type
flightNumber	string
type	enum
delayMinutes	int | null
cause	enum
originalDeparture	time
newDeparture	time | null
isCascade	boolean
cascadeSource	string | null

---

Affected Crew

Column	Type
crewId	string
originalPairing	string
currentLocation	string
state	enum

---

Summary

Column	Type
totalFlights	int
uniqueRoutes	int
airportsServed	int
aircraftTypes	string[]
totalAircraft	int
totalCrew	int
pilots	int
flightAttendants	int
crewByBase	map<string, int>
totalPairings	int
avgLegsPerPairing	float
crewReturningToBase	int
unassignedFlights	int
totalDisruptions	int
delays	int
cancellations	int
cascadingDelays	int
crewAffected	int
seed	int
generatedAt	datetime

---

Violations

Column	Type
violations	list
MCT	int
DUTY_OVERLIMIT	int
TYPE_RATING	int
REST	int
TOTAL	int

---

API Usage / Flight Engine

This repo provides a mock flight scheduling API for development/testing. Deploy it on Render, Heroku, or locally.

See endpoints:

• /airports?code=<IATA-CODE>
• /airports/all
• /flights?date=YYYY-MM-DD
• /flights?date=YYYY-MM-DD&origin=<IATA>
• /flights?date=YYYY-MM-DD&destination=<IATA>
• /flights?date=YYYY-MM-DD&flightNumber=<num>

Responses are JSON-formatted, see examples in original Flight Engine README.

---

Challenges

• Adhering to full FAA regulations and practical airline SOPs
• Efficient parallelization of crew recovery
• UI/UX complexities for real-time ops
• Simulating realistic disruptions and cascading effects

---

Accomplishments & Takeaways

TODO: Add accomplishments, takeaways, and demo outcomes here.

---

What’s Next

• More robust mobile features for on-duty crew
• Improved disruption prediction/simulation
• Airline-scale Gantt visualizations
• Open data export and reporting

---

Testing

Run tests with:

npm run test

Additional scripts:

• test — all tests
• test:changed — only tests for changed files

---

Contributing

Interested in contributing? See CONTRIBUTING.md.

---