Airlines in Australia have been widely criticised recently for supposedly price gouging by limiting the number of flights they offer.
In reality, it’s more complicated than that. As the industry has emerged from covid there is a real a resourcing issue because in the wake of the pandemic the airlines can’t get enough staff to put on extra flights.
The airlines are working to put on more staff but it’s not going to happen immediately. Demand for flights will continue to outstrip supply for the foreseeable future, so the challenge for airlines is to make better uses of the resources they already have.
This is where data analytics has a crucial role to play, using demand and resource optimisation.
A demand forecast can identify where unmet demand is strongest and the routes where a lower number of flights wouldn’t have such a strong impact. Resource optimisation can identify those available resources and how to use them to best match the demand forecast.
Airlines are struggling to meet the current high demand, but this varies across routes and time of day – this is where the opportunity lies for airlines.
One flight schedule change causes many more
Airlines can shift flights from routes and times of day when demand is lower and aircraft are less full to routes and times where demand is stronger.
For instance, perhaps an airline has 30 per cent of its flights on one route with surplus capacity and other routes where it can’t meet demand. All it has to do is to reorganise its flight schedule!
This is easy to say but very difficult to do in practice.
There are a huge number of constraints which limit the options for dropping a flight on one route and redeploying it on another and each change can ripple through the entire system.
The first is obviously the aircraft itself. Will it be at the right airport for the new flight? Do there need to be changes to previous flights to ensure an aircraft is available? And how will the change affect aircraft availability for the flights that come later?
For instance, if a Sydney-Melbourne flight is switched for Sydney-Brisbane, that will result in a plane being in Queensland at the end of the flight when it might be needed for another flight from the Victorian capital. And doing it the other way around – switching a Melbourne-Sydney flight for a Brisbane-Sydney flight, for example, raises even more logistical difficulties. Similar constraints apply to the cabin crew.
Then there are external constraints. Will airports have take-off and landing slots for the new flight times? Will there be enough check-in and security staff at the terminals?
Taking a system-wide view of aviation
These are much bigger problems than can be worked out on a spreadsheet. Only sophisticated data modelling will be able to account for the many possible flight schedule combinations and identify the best use of airline resources that will see the largest number of passengers get to their destinations on time.
The Australian aviation space is essentially one giant system (which is part of an even larger global system), encompassing all the moving parts of airlines, airports and support services such as air traffic control and customs.
Ultimately it may be possible to build a digital twin, or model, of the system and see how one small change to one part will ripple through the rest of the system.
That’s something for the future. In the meantime, airlines have the opportunity to maximise the use of their own resources to get more passengers on flights.