Talking tech with AWS – how machine learning is transforming airline operations
When airlines were asked about their top business goals for 2022-2023 for the Skift 2022 Digital Transformation Report, two in five (41%) said 'reducing operational costs' and two in three (68%) said 'improving customer service'.
Accurately forecasting demand, reducing flight delays and cancellations, and improving on-time performance are key to achieving both goals. That's why some leading airlines are turning to solutions based on machine-learning or artificial intelligence to optimise core operations.
In fact, all of the airlines surveyed by Skift said machine learning and artificial intelligence would be very important, or somewhat important, for driving value over the next three years.
- Airlines prioritize reducing operational costs and improving customer service for 2022-2023.
- Accurate demand forecasting and on-time performance are crucial for achieving these goals.
- Machine learning and artificial intelligence are seen as essential for driving value in the next three years.
- Solutions like Passenger Traffic Forecasting and TCS Aviana are helping airlines optimize operations.
- Azur Airlines and iFlight showcase the benefits of gaining situational awareness and reacting quickly to disruptions.
- Travellers' eagerness to travel highlights the importance of offering exceptional operational experiences and customer service.
By employing machine-learning, airlines can make traffic forecasting "not only better but more consistent," explains Farhan Mohammad, who leads the Travel & Hospitality Solution Portfolio for Amazon Web Services (AWS).
Forecasting conventionally uses mathematical models that rely on historical data to predict capacity, demand, and pricing, but advanced solutions are now helping airlines minimise delays and anticipate disruptions in near-real-time.
"Ultimately, they're creating more opportunities to accelerate revenue growth and increase customer satisfaction", he says.
Enhancing forecasting with machine-learning
The AWS Travel and Hospitality Solutions Library contains curated solutions for common-use cases for the Travel and Hospitality industry, and AWS Marketplace - the place to find, test, buy, and deploy software that runs on AWS - offers many vetted, purpose-built industry solutions, highlights Mr Mohammad.
One example is Passenger Traffic Forecasting from 'Mphasis', a solution that provides 30 weeks of forecast of passengers who are expected to travel using historical weekly passenger traffic data.
This tool "helps airlines and passenger rail and ground services companies predict the number of expected passengers more accurately", says the AWS executive, using ensemble machine-learning algorithms with automatic model-selection algorithms.
"As soon as an accurate forecast is in place, travel companies can develop pricing strategies to sell seats at a price point and quantity that will maximise revenue", he adds.
So, how much of a difference can it make?
According to Mr Mohammad, when used to predict the monthly domestic passenger traffic at John F Kennedy International Airport in New York, the solution outperformed algorithms by other companies, pre- and post-COVID-19 pandemic.
Mphasis Passenger Traffic Forecasting of Passenger Traffic at New York's JFK International Airport
Improving on-time performance, starting with real-time data
Airline operations are complex.
They involve a broad set of time-critical processes that span ground and flight operations, above- and below-the-wing tasks, flight and cabin crew operations, and customer experience management.
"That's why United Airlines modernised its operations with TCS Aviana, an intelligent airline operations solution built on AWS", explains Mr Mohammad.
"Its installation means that United can now detect operational deviations and mitigate disruptions faster, resulting in greater opportunities for hassle-free travel."
On-time operations is one of the most valuable metrics for United Airlines and an essential promise it makes to customers.
But with large, complex operations involving over 5,000 daily departures across more than 300 airports and involving more than 3,000 operations personnel, achieving that goal requires an extraordinary coordination and effort.
"The airline realised that it needed both seamless collaboration and interoperability among business units and solution partners, and to provide frontline staff with near-real-time information to make accurate decisions faster," says Mr Mohammad.
United addressed these challenges by connecting its siloed systems using TCS Aviana.
The solution identifies deviations in operations and flags severity based on thresholds set in the airline's operational plan.
"Early visibility into critical issues, such as missed passenger connections and baggage delays, makes it possible for United to take proactive action and resolve problems more quickly", says Mr Mohammad.
TCS Aviana's impact has been significant; it configured over 500 operations anomalies and flagged nearly 400 business events per second for immediate remediation.
"TCS Aviana helps orchestrate harmony across aviation operations, activating an empowered workforce and resilient operations and delivering a superior traveller experience", explains Sreedhar Gudla, Product Head of TCS Aviana.
Gaining situational awareness for increased profitability
Another example of an airline deploying new solutions is Azur Airlines, an expanding Eastern European airline with more than 30 aircraft.
"As it grew, operating, managing, and optimising its fleet became increasingly challenging tasks", says Mr Mohammad.
With severe weather, geopolitical incidents, and natural disasters on the rise, the airline needed the means to obtain vital information (including fuel and GPS data and aircraft block times) in near-real-time to "visualise" its fleet, Mr Mohammad explains.
Azur turned to FLYHT Aerospace Solutions - which helps airlines gain full visibility over assets and the movement of resources and inventory - and began using its AFIRS 228 system and UpTime Cloud Aircraft Situational Display, an AWS-hosted application.
The airline installed AFIRS 228, a data acquisition unit, connecting it to aircraft interfaces to obtain pertinent data, such as aircraft position, block times, and fuel onboard.
The unit sends the data from the aircraft to the AWS infrastructure on the ground.
The UpTime Cloud application then displays the data, facilitates near-real-time fleet monitoring, and activates automation through alerts - all through a simple user experience.
"Now, Azur gets all the information that it needs in one platform, as well as weekly, automated flight summary reports, directly from the Uptime Cloud AWS platform", says Mr Mohammad.
What are the benefits of achieving situational awareness?
"A maximised fleet and improved profitability", Mr Mohammad says.
Reacting quickly when disruptions occur
Though airlines can't control the weather, they can use machine-learning and artificial intelligence to improve how they react to disruptions, weather-related or otherwise, more quickly and effectively.
Mr Mohammad highlights iFlight from IBS Software as a solution, a highly configurable platform for aircraft, flight, crew, and hub planning, management, and optimisation.
"Its artificial intelligence-driven tools highlight operational issues early, making it possible for airline teams to execute and communicate optimal recovery solutions quickly", according to the AWS executive.
By indicating down-line impacts, iFlight helps minimise both the direct short term cost of disruption (for example, revenue dilution, passenger compensation, and hotel accommodation) and the long term costs of passenger dissatisfaction.
One airline group turned to iFlight after multiple attempts to integrate end-to-end airline operations and crew management IT systems failed, notes Mr Mohammad.
"It needed a better way to make collaborative decisions, from crew planning and management to process standardisation and proactive disruption management."
Using iFlight gave it "a single view of truth", leading to USD3million in savings each year, he adds.
Travellers are still eager to meet in person, sightsee, and explore
The spike in travel demand in 2022 made it clear that travellers are still eager to meet in person, sightsee, and explore.
To build loyalty and retain travellers, airlines must offer both the operational experience and customer service that today's travellers expect, says Mr Mohammad.
About Farhan Mohammad...
Farhan Mohammad leads the Travel & Hospitality Solution Portfolio for Amazon Web Services (AWS), where he manages and brings to market a portfolio of solutions to address the industry's most pressing needs. Farhan has 20 years of experience in go-to-market strategy and digital transformation.
As an avid traveller and consumer of the industry, with millions of flight miles logged and thousands of hotel nights booked, Mr Mohammad offers a unique perspective on the needs of the traveller.
He has worked with the world's largest airlines, hotels, and hospitality companies to reimagine the future of travel.
About AWS Travel and Hospitality…
AWS Travel and Hospitality is the global industry practice for Amazon Web Services (AWS), with a charter to support customers as they accelerate cloud adoption.
Companies around the world, across every segment of the travel and hospitality industry - and of every size - run on AWS.
This includes industry leaders like Airbnb, Avis Budget Group, Best Western, Booking.com, Choice Hotels, DoorDash, Dunkin' Brands, Expedia Group, Korean Air, McDonald's, Ryanair, SiteMinder, Sysco, Toast, United Airlines and Wyndham Hotels.
These companies, and many others, are transforming their businesses by leveraging technology to enhance customer experiences and increase operational efficiency.
For more information visit: AWS Travel and Hospitality