Keep flight delays at bay.

AVIATION INTELLIGENCE · DELAY ANALYTICS

An interactive R Shiny application exploring historical departure delays at New York’s three major airports.

The idea: turn flight records into an accessible tool for exploring delay patterns and estimated risk.

Delays disrupt journeys

• Passengers need better ways to understand historical delay patterns.
• Departure delays may vary by airport and scheduled departure hour.
• Historical data can reveal patterns worth investigating.

Our objective

Build an interactive dashboard that lets users explore flight scenarios and understand the limitations of delay estimates.

From flight records to an estimate

##                     Measure                Value
## 1 Historical flight records              328,521
## 2        Departure airports        JFK, LGA, EWR
## 3              Dataset year                 2013
## 4           Delay threshold More than 15 minutes

Data source: nycflights13, a dataset of New York-area flights from 2013.

Application inputs - Departure airport - Scheduled departure hour - Flight distance in miles

Model: logistic regression estimates the probability of a departure delay exceeding 15 minutes using the selected airport, hour and distance.

Historical departure delays

In the Shiny app, users can select an airport, adjust departure hour and distance, and run the model to explore an estimated delay probability.

What can we conclude?

• Historical delay rates can be compared across airports.
• Departure-hour patterns can be explored visually.
• Interactive inputs make the model accessible to non-technical users.

Limitations - The dataset represents 2013, not live operations. - Historical associations do not prove causation. - The model is educational and has not been validated for operational forecasting.

Flights at Bay

Understand delays. Make informed decisions.

Built with R, Shiny, nycflights13, dplyr and ggplot2.