Data Driven

Then came VisiCalc, the world’s first computer spreadsheet invented by Dan Bricklin and Bob Frankston in 1979. Suddenly, companies could model any number of potential scenarios – or play the “what if” game – at the touch of a button or click of a mouse.

This at-the-time revolutionary piece of software unlocked the potential held in data and that process of extracting more insights from the numbers continues to this day.

Especially in aviation, which generates terabytes of data daily, if not hourly, from fuel burn rates to fleet utilisation and much more.

The responsibility for analysing what the data says or how it can help improve operations and therefore the bottom line is inevitably falling not on the shoulders of men and women at a desk in front of a spreadsheet, but on algorithms or software that can crunch through the numbers in a matter of seconds.

And software company IFS is hoping its new Tail Planning, Optimisation and Assignment product will prove to be the right solution for an airline looking to better manage its fleet.

New Product Aims To Improve Fuel Efficiency

Although global oil prices may be low now compared with a few years ago, it still represents about a third of an airline’s operating expenses.

As a result, having the right aircraft operating the right route for the given payload goes a long way to helping keep costs down.

Alongside fuel, there are other considerations – does the route need a widebody for cargo? What is the maintenance capability at the outport? There are also hard constraints. For example, runway length or legal restrictions on operating particular aircraft in certain countries.

And soft constraints, such as a dedicated crew rest area for long-haul flights, which do not prevent an aircraft from operating a particular route, have to be accounted for as well. If an aircraft does not have one, could blocking off a few seats in business class suffice for the 12-hour flight across the Pacific? The IFS Tail Planning, Optimisation and Assignment product, launched in October, has been designed to give airlines a leg up in the way they fly their aircraft.

IFS Middle East solutions manager Andrew Stimpson said the program integrates an airline’s commercial planning, operations planning and management, and engineering and maintenance units to deliver the optimum solution amid the hundreds of potential constraints and considerations.

“We basically see it as sort of one part of a set of intermeshed cogs where all the different cogs impact on each other. They have to work smoothly together,” Stimpson told reporters invited to a media briefing about the product in Bangkok on October 25.

“I suspect quite a lot of airlines actually do that job manually at the moment. They have a tool that can assign tails to flights but it is manual click-and-drag.

“Conceptually that all seems quite simple but there’s a lot of different factors that actually make this a very difficult problem to solve in an optimal way. ” IFS currently works with airlines, airports and maintenance, repair and overhaul (MRO) facilities, offering enterprise software that helps them run their operations, manage their supply chains and keep their technical documentation up to date, among other tasks.

Its aviation clients in this part of the world include Hawker Pacific and TAE, according to a slide presentation accompanying the product launch briefing in Bangkok.

Further afield, IFS also lists Emirates, Dassault Falcon, GE Aviation, Alitalia Maintenance Systems and the US Federal Aviation Administration as customers.

However, the new product is part of the global company’s foray into the fleet scheduling aspect of an airline’s operations.

IFS believed the product would be best suited to airlines with 40 or more aircraft and was applicable to both low-cost and full-service carriers.

“The same business problem still exists. You need to manage your constraints, try and allocate your best aircraft to your flights and integrate maintenance,” Stimpson said.

“LCCs have a greater commercial imperative to do that well because they are trying to keep their cost base low. ” The program takes advantage of the fact that airlines rank their aircraft based on fuel efficiency, with the list updated regularly. That list is used to inform the airline which is the optimum aircraft for a particular route.

“All the airlines will have engine monitoring tools that say how efficiently they are operating,” Stimpson said.

“At the same time, they will be monitoring the total weight of the aircraft and the aerodynamic efficiency of the aircraft so they can then give a ranking in terms of how much payload they lose out with running Aircraft A on a certain route versus Aircraft B, so they know what the most efficient aircraft is.

“They publish those and they update those ranking tables often because as you change the airline configuration the ranking table changes.

“Basically, what we are saying is if you can consistently get the best aircraft onto the longest flight you will be able to achieve a greater amount of payload for a reduced amount of fuel. ” Asked what savings an airline might achieve, Stimpson declined to offer an estimate on the potential financial benefit, preferring instead to talk in terms of payload improvements based on initial modelling.

“From what we have seen so far, on payload impact when we are looking at fuel efficiency, we were somewhere about 15 to 20 per cent better than the previous manual process,” Stimpson said.

“Exactly how that correlates into dollars I’m not 100 per cent sure but in terms of percentage improvement on payload loss, 15-20 per cent is what we have been seeing. ” The program also determines the best window for aircraft to have their regular maintenance checks, and helps to get the schedule back on track after disruption due to weather or technical issues.

Qantas Taps Amadeus For Disruption Management Tool

The ‘Defining the Future of Travel Through Intelligence’ report from travel technology provider Amadeus, published earlier in 2016, noted there were more than 600 million passengers affected by disruption in 2014 and getting operations back to normal is a difficult task requiring input from many different parts of the airline.

To that end, Amadeus developed a Schedule Recovery product in partnership with Qantas that uses data analytics tools to help airlines “swiftly identify the most critical issues, and act upon them”. “Operational recovery from disruption is complex, as the airline must balance costs, customer service, and constraints such as staffing,” the report said.

“The volume of information and the decision-making required can be overwhelming, and it is very difficult for manual decision-making to take in the full operational and traveller picture.

“Schedule Recovery reduces manual effort and improves the quality of decisions while closely controlling the strain on the airline operating costs and minimising the overall impact on passengers. ” The Amadeus report said Qantas was able to remove up to 300 minutes of flight delays from the initial air traffic control (ATC) slot allocation, thus reducing the number of flights that would have been reported as being late by 60 per cent due to ATC.

“Within the Qantas Operations Centre, we want better predictive analysis on what’s going to happen each day,” Qantas head of operations Paul Fraser said in the report.

“If we see there’s an expected weather event in Sydney – rather than wait for it and react, we can go to our customers the night before. Then they can decide whether to, say, delay their meeting or take their chances. They are taking control of that choice, with analytics putting the decision back in the customer’s hands. ” Fraser said Qantas’s long-term goal was to utilise these analytics to become more pro-active in addressing issues before they impact on the network.

“Today, if there’s an operational problem, then we react to this,” Fraser said.

“We’d like to get to a predictive position, where we are using predictive analytics to look at probabilities around forecast conditions, our fleet, customer load and so on. ” Amadeus executive vice president of research and development Hervé Couturier said data and “machine learning” was expected to drive many of the innovations to emerge in the period ahead.

“We don’t teach the machine, and we don’t give it rules,” Coutourier said.

“We feed the machine data, and we let it discover the pattern, with the computing power facilitating the work. This is the move from rules-based learning to pattern-based learning, which allows for a much higher ratio of confidence in output. ”

There’s a lot of different factors that actually make this a very difficult problem to solve.