Fifteen years ago, it was acceptable for airlines to focus purely on pushing planes through the skies. Today, they are pushing terabytes of data, too. An airline’s core competency has extended to shifting—and sifting—data from a constellation of different sources. Big data in aviation is changing everything.
Imagine this: You run a major airline, and a storm is hitting the East Coast today, meaning that several flights will be delayed. On each flight are passengers with their own loyalty and revenue profiles, and many need to make a connecting flight with your airline.
You must determine whether to hold each connecting flight long enough to accommodate late passengers. You must also consider baggage transfer times, the number of transferring passengers, the flights they are coming from, and the length of time it takes them to get from one gate to another, along with many other variables related to the profitability of each outgoing flight. Where do you start?
This is one example of how big data in aviation can help make business more efficient. These transportation companies live and die by data, and they are producing more of it each day in every part of their operations. A Boeing 787, for example, creates half a terabyte of data on average for each flight it takes.
Combined with information on weather, customer contact center interactions, ticketing information, and airport performance times, this ocean of data can offer some significant business insights for companies operating in a highly competitive industry.
Take fuel usage, for example. At 17 percent of overall operating costs, fuel is the second-highest expense for airlines after labor, making fuel efficiency a critical metric. Airlines are now using big data in aviation to help squeeze new efficiencies into their fuel usage.
Computing power has developed to the point where airlines can gather and process the vast amounts of data they need to analyze fuel usage on a per-trip basis. Southwestern Airlines has collected data directly from sensors embedded in its aircraft, including information on wind speeds, ambient temperatures, plane weight, and thrust. It feeds this into an analytics engine and marries it with operational data around fuel, passenger, and cargo loads, along with weather data, to search for patterns in trip profitability.
The airline hopes that this data mining will produce actionable intelligence around decisions such as adding or subtracting flights to routes, setting fuel loads for each aircraft, and selling additional passenger tickets. They can also give this intelligence to pilots in the air. If turbulence creates the need to adjust an aircraft’s height in flight, big data can now give Southwestern’s pilots a detailed analysis of the extra fuel burn associated with a specific altitude and its associated cost.
Qantas Airways has also started using big data to provide similar decision support to its own pilots as part of its FlightPulse aviation analytics system.
Data from aircraft-based sensors can also create insights beyond fuel efficiency. Boeing uses analytics to look at 2 million conditions each day across 4,000 airplanes as part of its Airplane Health Management (AHM) system. This data, which includes in-flight measurements, mechanic write-ups, and shop findings, helps the company to plan equipment maintenance with minimal disruption to flights.
For example, data analytics predicted the failure of an integrated drive generator, allowing it to investigate and correct the issue before it became a problem. That saved up to $300,000 in service delays and repair costs.
By amassing data from flight incidents, regulators are also hoping to improve safety across the entire industry. The European Aviation Safety Agency (EASA) launched Data4Safety, a data collection and analysis program to detect risks using a combination of safety reports, in-flight telemetry data, air traffic surveillance information, and weather data.
The program will enable regulators to identify the biggest safety risks and determine whether industry stakeholders are taking the right actions to mitigate them. By combing through terabytes of data, it hopes to find the weak links in the aviation chain.
While much of the data collected by airlines focuses on what happens in and around planes, there are equally big gains to be made on the other side of the gate. United Airlines used big data to switch from an aggregated view of its customers to focused, individual profiles.
Instead of simply identifying its most successful products, the airline uses big data to explore each customer’s buying habits. By analyzing more than 150 variables about each customer, including prior purchases and destinations, it predicts their likely actions and dynamically generates personalized offers. Using big data in this way increased its revenue from nonticket sources, such as baggage fees and onboard food and services, by 15 percent.
Use cases such as these demonstrate how airlines are more than just transportation companies these days—they are information technology and data science companies. The use of big data in aviation is piloting airlines toward a new future.
Learn more about how airlines can use big data and machine learning to predict flight delays.