Airlines get smarter in their planning

This story is sourced from Airline Business
Subscribe today »

A game-changing transformation of the way airlines plan and organise their schedules and operations could be on the horizon. The academic vision is already there for a smart planning system that sits at the heart of the airline and its ­business model, based on a common platform so that all departments can see the impacts of their initiatives on the business. And more importantly, this vision also has buy-in from the industry.

What is less clear-cut is how this vision will be achieved, particularly as current planningsystems are less than optimal. Today airlines solve multiple local problems in a serial fashion, taking a "rear view mirror" or historicapproach to the analysis that underpins their planning of routes and services. These processes are highly complex and time-consuming, ­undermining airlines' ability to be flexible in the way they adapt to the fast-changing ­business environment.

With computing power and mathematical algorithms improving all the time, there are now one or two vendors that are making great strides towards cracking the integration of processes, departmental systems and decision-making to create more holistic solutions. But getting smarter about planning demands more than technical innovation - the organisational and attitudinal changes that ­next-generation systems will require are equally challenging.

crows nest, rex features 
 ©Rex Features

For Professor Nawal Taneja, chairman of the Department of Aviation at Ohio State University, the schedule is "the heart and soul of an airline". His advice is to "make the airline schedule the central part of the airline operation". Taneja's vision is one of optimisation on a systems-basis or balanced approach, rather than the typical optimisation on an individual functional level. This means aligning organisational structures to increase flexibility with regard to resources, aircraft, crew and the ­ability to deal with a range of future ­circumstances, all facilitated and underpinned by a common IT platform offering greater and faster planning capabilities.

In his latest book, Flying Ahead of the Airplane (published by Ashgate at the end of 2008), Taneja sets out the case for combining all planning functions, such as scheduling of aircraft, crews, maintenance, ground facilities and manpower, with the objective of optimising profit-generating capabilities, ensuring the lowest possible costs, highest revenue and most cost-effective reliability. In this scenario, all activities would be co-ordinated through the centralised scheduling department, which would report directly to the chief executive.

The benefit would be vastly improved, more responsive decision-making. "Instead of the scheduling department saying it will take two to six months to change the schedule, you can change it in a day," says Taneja, adding that it would have been changed "in a much more holistic way as opposed to impacting on just one function".

Decisions taken using a smart planning system that integrates data and business logic from all parts of the airline would allow carriers to target resources more efficiently and target the overall benefit tothe airline, says Ulrike Gall, senior vice-president of airline management solutions at Lufthansa Systems: "Studies show that separation of the planning process into individual steps and focusing on individual resources lead to inefficiencies in plans. Thus, integrated plans can lead to direct cost savings. Shortening the planning process allows airlines to respond in a more agile and efficient way to changes in the environment in which they operate."

Taneja suggests the impact of an integrated planning and business model could be even more far reaching - helping determine everything from the number of aircraft needed to whether people are needed in one area or another. "Competition will be forced to adapt to it," he says. "It's going to take out some of the older generation companies with older generation executives. I don't mean older by age, I mean older in thinkingIt's the ability to be able to think ahead."

His vision has a lot of appeal for airlines. "It is viable as a visionand one that we have been building towards in big and small ways for a number of years," says Theresa Wise, senior vice-president and chief information officer at Delta Air Lines. "As the capability of IT and decision support modelling (via operations research) continues to advance, so does our ability to realise elements of that vision."

However, the system Taneja describes offers potential challenges as well as benefits. "Greater flexibility and finer-grained planning potentially brings more variability to the planned schedule, which in turn requires systems and processes in our day-to-day operation that match the sophistication of the planning systems and processes. Those too are still evolving," says Wise.

"Greater flexibility and finer planning potentially brings more variability to the planned schedule"
Theresa Wise - Chief information officer, Delta Air Lines

"To integrate the local problems in to one global problemthat resonates deeply with me as a planner," says John Jamotta, senior director of integrated planning at Southwest Airlines. He points out that the idea is not brand new, but people are wrestling with how to accomplish it. Current solutions are, in Jamotta's view, "sub-optimal", with airlines coming up with a fundamental solution to one local problem first - for Southwest it is the airline schedule - and all the other departments solve their problems against this benchmark, but they are not fully informed on issues such as cost. In addition, airlines are also stumbling to keep up with the pace of change in their businesses.

"There are two fundamental problems: you are not solving the global problem, so you are sub-optimal at the get-go. The second piece is the velocity of the business is so quick, you want to react ideally on a daily basis," says Jamotta.

These difficulties are compounded by the overall lack of integration within the typical airline organisation or, as Taneja describes it, airlines are misaligned. "There is no connection between brand strategy and external people like passengers and distribution, let alone internal people who are all working in silos," he observes. And the result is a lack of flexibility, which has a negative impact on both passengers and airline profits alike.

Economic and competitive pressures are demanding that airlines become more flexible and responsive in their product offering and more accurate in predicting demand. "If you are going to segment the market you have got to be able to segment the supply. If you segment the supply you need better planning," says Richard Clarke, director of Travel ­Technology Research. He believes that greater computing and mathematical capabilities suggest there will be a breakthrough in airline planning and this is vital.

So is anyone getting close to realising Taneja's vision? "A smart planning system that integrates all relevant commercial and operational aspects and solves a single holistic planning problem would still exceed the computing power available today," believes Gall at Lufthansa Systems. "However, certain related planning processes like fleet assignment and crew pairing construction can be tackled with current computer systems. As the computing power grows, it is increasingly a question of combining the different systems into a homogenous platform."

Already a couple of vendors have announced systems offering much greater levels of integration in the planning process.In the USA, Airline Intelligence Systems has developed jetEngineO/S , a business platform that enables the integration of planning, revenue management, operations and reservations functions in real-time, promising to drastically reduce the planning process.

Chief executive Stephen Johnson says: "When an airline plans routes, fleets, crew, maintenance - that process which all airlines do on a grand scale a couple of times a year -they are planning six months in advance and it's a three month process and long and arduous. We are able to do it in minutes." He adds that a handful of carriers are keenly interested in the system, which can be set up to run in parallel with an airline's existing systems. He says it works for all airline models and AI Systems only starts charging when the airline sees a benefit. "It replaces so much effort. It is a standalone system and runs on a small footprint, on a rack of servers as opposed to hundreds of servers in different cities."

Aeromexico signed up as the first jetEngineO/S customer in 2007 and as well as expecting its system to be deployed in the first half of 2009, Johnson is promising a separate "game changing" announcement in the first half on new customers and product evolution. Operationally, Johnson forecasts his solution will enable a tremendous streamlining of individual airline businesses as well as impacting on the industry overall. "It will change which airlines come and go, how airlines merge. I think right down to the micro level of changing seating and configuring cabin classes, up to how airlines interact with each other and who's going to stay and not going to stay."

 "Decisions made by departments will have to be delegated to a single central decision maker"
Ulrike Gall - Senior vice-president, Lufthansa Systems

UK aviation consultancy Decisal is using a complex mathematical algorithm invented at Imperial College London to develop ­SchedulAir, which unifies the optimisation of fleet assignment, aircraft maintenance routing (rotation) and crew scheduling (pairing). Launched at the IATA Schedules Conference in Athens last November, it too has received plenty of interest from airlines.

Nikolaos Papadakos of Imperial College, who was involved in the research of ­SchedulAir, explains: "The schedule has less cost because the unified optimisation works out the overall minimum cost. Robustness is obtained because the aircraft rotations are considered simultaneously with the crew schedules. In this manner the unified system foresees when pilots need to change aircraft and avoids such changes if they are likely to cause potential schedule bottlenecks. This was impossible with the sequential integration as the rotations were frozen by the second sub-system." The system can generate cost savings of up to $150 per flight. "It could be from $50 to $150, it all depends on the structure of the airlines and how these things are worked out," says Papadakos.

There is more to smart planning systems, however, than computing power and systems improvements: assembling the infrastructure to implement decisions takes time. "In certain circumstances, even today, we can indeed rearrange schedules and crews in minutes or hours," says Wise at Delta. "For more complicated moves such as starting service to a new airport or rearranging hubs, while the planning might be quick, the execution must take time. Employees may need to be hired, trained or relocated, contracts may need to be signed, facilities may need to be adapted."

Planning in a smarter way will also require new ways of thinking from airlines and airports. "Airlines are good at understanding where there are opportunities to be flexible, dynamic and robust and we can turn that into mathematical models that can be used in the planning process," says Cynthia Barnhart, co-director of the Operations Research Center at Massachusetts Institute of Technology. "It's not computer or mathematical tools that need to advance further at this point, it's encapsulating the knowledge airlines have so they can operate in a different way."

The greatest challenge airlines face as they move to smarter planning systems is how they structure themselves - any holistic, integrated approach to planning will mean an organisational rethink. "Decisions that are currently made and answered for by individual departments will have to be delegated to a single, central decision-maker," says Gall. "On the operations side, business processes for decision-making on day-of-ops are already quite tightly integrated at many airlines. However, during the complete planning process separation is still strong and decisions are driven by a local view on the business objectives. Historical data on the impact of schedule disruptions is rarely used in a systematic way."

Southwest Airlines is already tackling this problem. "We are trying to reorganise the company to get this global solution," says Jamotta. Last September, Southwest replaced its scheduled planning department with a cross-functional integrated planning department to which the communications and operations business functions now submit their requirements. The airline is using emerging and advanced mathematics and computing to solve local problems, with planning predictions based on analysis of historic data. "We are now at the point where we realise we have to go to next generation planning," Jamotta says. And Southwest is considering whether, having built its first two optimisers in-house, it continues alone or with a vendor partner.

Jamotta is keen to move away from simply basing decisions on historic data. "In a future state we would want to switch that and see what we predict is going to happen, what's emerging, demographics, future trends," he says. "Current state tools look at the rear view mirror, future state tools are trying to get better through some sort of modelling of the future."

But regardless of the smart planning systems on the horizon, the strategy will always be human, not modelled. "After all is said and done these decision support tools will be enormously powerful, but they don't relieve the business from making decisions about where to drive the business," says Jamotta.