Looking at daily flight pattern, introducing our new statistics function. ... more
A first attempt at using data to analyse flight distances.
How plane age and seat capacity determine aircraft utilisation. ... more
Where airlines dominate airports and how this could explain high US ticket prices. ... more
Planes are fast. With a bit of tailwind, planes can reach peak speeds above 740mph (1,200 kmh), though ground speeds of 500-550mph (800-900kmh) are more common at cruising altitude. And while most planes today have very similar cruising speeds, the time they can fly at that speed heavily depends on the duration of the flight. Average speeds for a long-distance flight can easily top 800kmh more than twice as much as some shorter flights
So how far do planes fly in their lifetime? Most answers found online vary widely in their answers and are based on rough estimates without using any actual data. Flightera's dataset doesn't allow (yet) to track planes over their whole lifetime but just the data of a few months can deliver good insights into the dynamics at play. Three factors play a big role in how far a plane will fly in its lifetime:
Just looking at December 2017 data, I retrieved the average monthly distance per plane (which is on our website for each active plane) to create the table below.
|CN-ROE||Royal Air Maroc||B738||489,460|
Looking at a 20 planes is hardly representative and time zones aren't all to explain plane distances. Therefore, I went for the whole dataset of ~23,000 active planes to explain more general trends. The plane size (which also determines the range and sort of route it's used for) expressed in seat capacity is plotted below and shows a clear relationship to monthly distance flown. This is mostly long-haul vs. short-haul but some linear relationship appears to exist ( more on that later).
As newer planes are more fuel efficient and have longer ranges, it's also reasonable to assume that planes will fly longer distances as they're younger. And indeed, that's what we find looking at data, as the chart below shows.
But how important are these factors? Running a regression on both of them, it appears that the age and seat capacity can explain ~30% of the variation in distance flown. According to the regression, each additional seat increases monthly distance by 480km while each year reduces it by 2,500km each month. For a long-distance plane (e.g. B777-300) with a 25 year life-span, this leads to a life-time distance of around 90 million kilometers (~50 million miles), or around 20,000 times around the earth.
This is a very rough estimate, only using the two simplest factors, and adding more airline-specific factors could probably increase accuracy significantly. Better factors and more robust methods, this will be part of the upcoming series on using machine learning on flight data. Stay tuned! In the meantime, follow us on Twitter to receive updates or to send us ideas for future articles.