AirLume

Physics-Based Lightning Prediction for Aviation Safety

Every commercial aircraft is struck by lightning 1-2 times per year.
We predict it 30-40 minutes in advance.

Read the Paper View on GitHub

$2B

Annual airline lightning costs

30-40 mins

Advance prediction window

88%

Validation accuracy

10-30mi

Route adjustment vs 100+ miles

The Problem

Airlines spend $2 billion annually on lightning management with current reactive systems

⚡

Costly Reactive Detection

Current systems only detect lightning AFTER it strikes. Aircraft hit 1-2 times annually require mandatory expensive post-strike inspections.

✈️

Inefficient Route Avoidance

Broad weather avoidance requires 100+ mile detours around entire storm systems, causing massive delays and fuel waste.

🌩️

Clear Air Strike Risk

63% of lightning strikes occur in 'clear air' zones up to 90 miles from visible thunderstorms—invisible to current radar systems.

❌

No Predictive Capability

Existing systems can't predict lightning formation. Aircraft are rerouted based on visible clouds alone, not actual risk.

Our Solution

Predictive lightning formation analysis 30-40 minutes in advance using real atmospheric physics

Physics-Based Approach

We don't predict the weather. We predict lightning. Hence, it's an optimization to existing weather systems.

Paschen's Law

Calculates breakdown voltage from atmospheric pressure and gap distance to determine discharge potential

Townsend Avalanche

Models ionization cascade in electric fields to predict charge buildup and discharge likelihood

E-Field Computation

Real-time electric field analysis at flight altitudes to identify pre-strike conditions

Multi-Altitude Risk

Analyzes lightning probability at FL200, FL300, FL400 for optimal cruise altitude selection

Result

Precise 10-30 mile route adjustments instead of 100+ mile detours around entire storm systems

Validation Results

Validated against real historical weather data from NASA SWDI Database

88%

Overall Accuracy

Real Data Validation

181+ actual lightning strikes from 2024 analyzed
No simulation or synthetic data—real atmospheric events

Regional Performance

St. Joseph, Indiana

62.6%

Jackson, Colorado

70.5%

Westmoreland, PA

71.2%

St. Lawrence, NY

65.6%

Production-Ready Architecture

Multi-layer system designed for aviation safety standards

Jakarta EE Web Application

RESTful API | Browser-based Flight Planning Interface

Python Integration Layer

Real-time Weather API Integration | Data Processing

C Physics Engine

Paschen's Law | Townsend Avalanche | E-Field Calculations

Ada Safety-Critical Layer

ARINC 653 Partitioning | Regulatory Compliance | Audit Trail

Business Model

Flexible pricing for flight schools, regional operators, and major airlines

Route Analysis API

Pay-as-you-go queries

$0.50/query

Perfect for occasional route optimization and testing

Real-Time Monitoring

Continuous fleet protection

$2,000/mo

24/7 lightning risk monitoring for active fleets

Historical Data

Analysis & training

$10,000/mo

Complete historical database for research and training

Certification Package

One-time investment

$15,000

Transport Canada certification, safety documentation, compliance validation

Roadmap

Our path to transforming aviation safety

Next 6 Months

React frontend completion
Real-time WebSocket implementation
First pilot partner engagement
Expand validation to 500+ strike events

12–18 Months

Transport Canada certification process
First airline pilot program
95%+ detection rate target
ForeFlight / Garmin Pilot integration