Airlines face many challenges when creating their schedules. The art of scheduling is all about knowing where, when and what to fly, but airlines often struggle to find the most profitable schedule that is also operationally feasible.
These challenges include the sheer range of scheduling possibilities – for example, an airline that flies to 100 airports could potentially operate up to 4,950 routes, bringing with them a range of timing and aircraft selection options. These routes must be operationally feasible and will inevitably include a wide range of passenger types.
Airlines also face an uncertain environment, where schedules must be finalised and available for sale months in advance of their eventual operation, with assumptions made about competitor schedules and anticipated traffic while factoring in things like weather and climate events, strikes and other unpredictable events. Finally, the quality of a proposed schedule depends on the quality of the input data used to create it.
A traditional approach to scheduling has been in place for years may be a safe option, but technology offers the chance to access wider datasets and, therefore, more accurate forecasting models. Higher data processing power, Artificial Intelligence (AI) and machine learning, access to wider shopping and booking data sets, and new ways of viewing data are driving change. Airlines can, in turn, offer more customer choice and enjoy higher efficiencies, meet their on-time performance (OTP) targets and, ultimately, improve profitability.
The small margins matter so much in air travel and every second counts. The US Federal Aviation Administration has estimated the operating cost as $120 per delay minute for an aircraft with 150 seats. Studies show that optimised schedules can improve OTP by 3 per cent, which can save millions of dollars.
The seven breakthroughs listed below provide huge opportunities for airlines to improve the quality of their schedule, improve efficiencies, drive savings but – most importantly – improve the customer experience. Let’s take a look at them:
‘Clean Sheet’ optimisation
Modern scheduling solutions can perform route development and schedule planning steps without resorting to a ‘base’ schedule - the schedule from the same period in the previous year.
With clean sheet scheduling, airlines have the flexibility to develop various scenarios to derive the most optimal schedule without being restricted by timing, routing, market frequency or capacity allocation.
With clean sheet optimisation, airlines can:
• retime the entire network for better passenger flow
• evaluate the impact of differing frequency or flight additions and deletions
• allocate capacity to maximise passenger demand and meet operational requirements
• optimise routes factoring in crew considerations and route-based maintenance
Scheduling for reliability and profitability
With the increased focus on on-time performance, schedule reliability has become a key priority for airlines. Schedule reliability must be estimated with acceptable accuracy during the schedule development process and it must be possible to evaluate reliability alongside schedule profitability.
To marry these dual objectives, modern solutions build simulations to estimate schedule reliability. These models factor in block times, runway capacity, staff and gates at different airports, aircraft turnaround times, maintenance and airline schedule recovery policies.
Simulations can identify the flights most at risk of being delayed and creating knock-on delays to later flights.
Producing operations-friendly schedules
Modern scheduling solutions factor in the needs of operations, factoring in the needs of crews and maintenance into the process. This means the right crew is in the right place at the right time for their flight. Studies show that by optimising crew rotations, simulations can reduce the number of pilot and cabin duties required for each rotation and increase overall utilisation. This can also mean that crews spend more time at home and the airline incurs less cost putting staff up in hotels.
Improving forecast accuracy
State-of-the-art machine learning and AI algorithms will continue to play an increasingly important role in schedule forecasting. Machines will be able to create more complete and accurate predictions as they learn more about passenger behaviour and other key factors.
To improve forecasts, airlines may need new sources of data, such as passengers’ preferred time of departure.
Integrating scheduling into the commercial area
Modern scheduling relies on various departments within airlines working together and sharing information on developing, pricing and distributing offers to customers. This ‘Offer Optimisation’ approach will result in better offers, more revenue and happier customers. By combining market data, airline data and customer data, airlines can use customer choice modelling to inform their network schedule planning, as well as ancillary offers.
Optimising schedules after publication
Traditionally, airlines publish their schedules between six to 12 months before their eventual operation. They are often reluctant to change their schedule for commercial reasons after publication and may miss important revenue opportunities.
Automating and streamlining schedule development
Modern scheduling solutions anticipate and suggest schedule changes, rather than requiring others to always think through fixes and resolve feasibility problems with the schedule. Modern systems use micro-optimisers that solve simple, standard scheduling conflicts by applying business logic used by schedulers and taking advantage of new processing power to quickly evaluate many possible solutions.
* The writer is vice-president, strategic consulting, for Amadeus