The aviation technology landscape continues to evolve apace.
As Amadeus builds a modern retailing environment alongside our airline partners, the offer to passengers will become more personalized and efficient as we work to boost sustainable growth. At the same time, new technologies, such as the cloud and artificial intelligence (AI), will accelerate a drive toward efficient simplicity.
Integral to these changes is network planning, less visible to passengers but a particularly significant area for driving airline profits.
Through accurate forecasting, decisions made at this stage of the journey - often many months before a passenger sets foot onboard an aircraft - can meaningfully impact on airline profitability.
As such, Amadeus is leading innovation in this area to drive improvement, by offering a suite of tools to airline partners as they work to build simplicity and increased efficiency into everything they do.
Today, I would like to explore this topic in a little more depth, examining how improved network forecasting can drive a better passenger experience.
Right on time
The fundamentals of airline scheduling are easy to state – but exceptionally difficult to execute.
In essence, carriers are aiming to produce a flight schedule that is profitable for them, but also operationally feasible. If a schedule is too optimistic, with aircraft scheduled to depart too soon, it may become unreliable, with insufficient time for an effective turnaround of planes at the airport. Too few departures, on the other hand, and airlines could leave money on the table.
At the same time, when it comes to forecasting, airlines struggle to build detailed scenarios which allow them to understand the impact of a single facet on their entire network, for example entering an alliance or a competitor launching a new route.
The key is to balance these various demands, predicting iteratively and accurately - and this is where effective technology can help.
Best in class
The best solutions are able to build multiple versions of a schedule, iteratively, in search of the best outcome. To do so, they adhere to a set of key principles.
Firstly, they must be specific to a particular airline during a precise future schedule period, such as the upcoming summer season, while also reflecting the market position of a carrier. A good schedule will respond to those of competitors.
At the same time, effective scheduling will consider aggregate demand for airline travel, including the projected volume of travel between airports and the revenue contribution passengers will provide. It will also consider the cost structure of an airline, using metrics such as fleet utilization, passenger counts and block hours flown.
Integration with revenue management practices must also be considered when generating the forecast. That is, when confronted with a demand forecast of different passenger types and limited aircraft capacity, the schedule forecasting solution should closely mimic the behavior of a revenue management solution when estimating which passengers can be accommodated.
Finally, but most importantly, preferences of the passenger must be considered.
If we know when a passenger would like to fly, their preferences when it comes to different carriers or aircraft types, their interest in different sorts of travel, non or single-stop, for example, then airlines can create a schedule to suit their needs. This information can also vary by region, or season.
If these principles are followed, airlines can avoid delays, create an efficient schedule, boost the bottom line and serve the needs of passengers.
In practice
Following ouracquisition of Sky suite, Amadeus has been building these capabilities into ourSkyCAST solution. The technology now helps airlines make more profitable decisions - with accurate forecasts for passenger demand, revenue and cost.
Today, it can be used to compare two manually constructed schedules. For example, schedulers and planners could create two scenarios, each adding a different new route and then use the forecasting solution to choose between them. Another example would be to share the forecasting logic with other optimization tools within the airline.
Forecasting solutions must be flexible enough to support both kinds of applications - standalone and integrated - with consistency between them.
It is also vital for a forecasting solution to be accurate. The more accurate the forecasts, the better the associated scheduling decisions and the better the experience for the passenger.
The most recent innovation in this area is to calculate schedule profitability for a historical period and compare the resulting forecasts with real world statistics. An analysis of the errors at different levels of aggregation will provide a sense of how closely the forecast reflects reality - thus providing guidelines on how the solution should be used.
This analysis will also provide a guide on how to make improvements.
Best practice is to perform these measurements frequently, so schedulers and planners can have the assurance that the solution they use is always accurate.
Taking off
Modern solutions allow us to create iterative schedules, constantly adapting multiple metrics to create a network that serves the needs of both the airline and the passenger. Developments, such as the growing role of AI and Amadeus’ move to the cloud, are already supporting this technology and will allow us to accelerate the benefits further.
Passengers may be unaware of much of what goes on behind the scenes, but they are set to benefit as we work to integrate effective network planning with the wider operation of airlines.
TO TOP