Punctuality is a continuing problem for airlines. Since 2004 the 35 largest scheduled operators at UK airports have gradually slipped from an average of 74.9% of all departures within 15 minutes of STD to 70.74% for the period January to June 2007. In the US, JFK is set to be taking drastic action to improve punctuality. And Chinese and Indian airlines are struggling with airport capacity and operational controls.
There are some legendary examples of airlines with poor punctuality performance turning into on-time performers. SAS under Jan Carlson and Australian Airlines under James Strong are two such examples of dramatic improvement in punctuality in the 1980s and 1990s. Of course things are more complex today: The sheer scale of aviation has changed. There are more flights with more people on board, and the airways and airports are more crowded.
But, achieving a high level of punctuality was unusual then as it is today. Punctuality was not achieved by padding the schedule, providing contingency aircraft, employing more people or buying more equipment. It was about effective planning, operating control, responding effectively to unexpected events and staff attitude.
To this day many of us still have our watches with the inscription “on time every time, customers have a choice” that we were each given, yes every employee at Australian Airlines, when we had achieved the habit of departing on time. |
The Cost of Delay
Aircraft delays are costly. Every unplanned minute on the ground incurs a cost for someone:
- For airlines: whose customers have a choice and judge them every time they fly, whose customers miss connections and their booked seats and require re-booking work. And who in extreme cases have the cost and disruption of crews going out of hours and have operations control centres constantly monitoring and managing the effects of disruption.
- For airport operators: who need extra gates and stands to accommodate delayed departures aircraft on the ground, whose terminal facilities need to be designed to accommodate more passengers than the schedule would indicate and who suffer negative publicity by association.
- For ground handlers: who pay the penalties of not meeting their Service Level Agreements with airlines, who try to plan staff rosters and equipment requirements to be efficient while dealing with fluctuations in performance and whose staff stay beyond normal shift times to handle delayed aircraft.
- For passengers: who are late for business appointments at their destination airport, who miss connections, who miss valuable hours of their long-awaited holidays and who have the concern of meeting people at destination.
- For caterers: who in extreme cases supply additional food on the ground or need to re-cater or re-ice if meals are going “out of time”.
And for the many other parties who lose money for every minute of delay
A late departure is rarely an isolated problem. Late occupancy of a stand often delays the arrival of the next flight allocated to that stand and causes unproductive time for the third parties waiting to prepare the next aircraft planned for the stand.
Turnrounds happen in series. If the time is not recovered, the result may be rolling delays across the operational day, which may affect a whole network or several networks if the knock-on affects other airlines, adding cost, rework and inconvenience to all stakeholders’ operations.
Airline punctuality has been a much discussed topic for many years and continues to be so, even at the White House now. Punctuality is clearly a key performance measure for airlines; but in many cases not enough progress is being made to improve it. In the meantime, aviation is ever more competitive and making a profit more difficult. The cost of delays is therefore ever more important for carriers and other stakeholders.
There are, of course, many reasons why the departure of aircraft may be delayed. The most prevalent reasons quoted when we fly - which we do very frequently - are: air traffic control, poor weather, late inbound aircraft, missing passengers, late baggage, check-in system failures, aircraft technical issues, airport queues or general ground handling delays.
We will explore these handling delays in greater depth in this paper and will discuss the key considerations in identifying the root causes of ground service delays and taking actions to provide for a consistently ‘punctual’ turnround process.
Turnround Delays: Why is the problem so hard to fix?
The objective of the aircraft turnround process is quite simple: to convert the aircraft from an arrival to a departure – on-time. Airlines, handlers and their staff carry out thousands of turnrounds every year so why does it not work smoothly every time?
The simple answer is that an aircraft turnround is a highly complex process made up of multiple activities involving numerous parties, pieces of equipment and infrastructure, many of which are dependent on one another. There is a high chance that, without effective planning and control, one of the activities will go wrong. When it does, it is likely to affect other parts of the turnround process. This makes the challenge of identifying and resolving the root cause of failure and controlling outcomes time after time more difficult. The aircraft turnround process is even more complex at busy hub airports where resources are assigned to carry out many turnrounds rather than being dedicated to few turnrounds.
In production terms, the activities that constitute an end-to-end process, such as a turnround, are referred to as:
- Processes on the critical path: a set of dependent activities that must be carried out in sequence, and where the total duration of all these activities determines the time required to complete the turnround. An example of activities on the critical path might be once the engines are shut down the doors cannot be opened before chocks are in place and stairs or jet bridge positioned.
- Parallel processes: processes that occur in parallel to the critical path that must be carried out within the total turnround time but do not depend on other events – such as the toilet truck. The toilet truck can carry out its tasks at any point in the turnround, is dependent on no other events but cannot start before the aircraft is on chocks and must be completed during the planned turnround time.
- Sub-processes: processes that supply or support the critical path and must be ready by a specific time to avoid delay to the critical path, although the process itself is not on the critical path, for example producing catering or preparing cargo.
- Statistical fluctuations: fluctuations in the time required for various activities or even differences in the exact job requirements from occasion to occasion. An example might be the different drive times or control post access times for a catering vehicle to reach the stand or different times required to clean the aircraft dependent on the level of soiling in the cabin.
- Dependent events: events that are dependent on other events, where the late end of one will cause a late start on the next. These events may not be on the critical path, but a delay in the first activity might push dependent activities onto the critical path. For example, aircraft loading is not on the critical path. However a delay in loading the holds will delay dependent ramp tasks (closing hold doors, removing equipment, ground power etc), potentially pushing these activities onto the critical path.
- Buffers: these might be either materials, resources or time set aside to minimise or eliminate the impact of statistical fluctuations in dependent events. For example if the trip from catering to the stands is 15 minutes but it can vary unpredictably, taking up to 25 minutes, one might plan for 15 minutes and have a 10 minute buffer at the stand so that if catering was on the critical path it would not cause delays.
If the turnround is not broken down into and planned according to these elements, it is highly likely that the critical path is not controlled. Inconsistent performance and, ultimately departure delays ensue if the critical path of the turnround is not identified and controlled.
Improving punctuality performance is made more complicated still by the fact that both carriers and handlers are trying to minimise cost and maximise efficiency. The conflict between minimising cost and maximising efficiency can lead to organisations optimising their own parts of the process without regard to the effectiveness of the overall turnround.
Where minimising cost is the goal; the principle pursued for turnrounds would be to supply the minimum amount of handling resource and equipment, for the shortest possible length of time. This approach is typically based on the flight schedule and the planned timing and duration of activities. However, because fluctuations occur in the timing and duration of activities, the flight schedule and activity plan are not definitive guides to resource requirements.
The Effect of Delay on Critical Path Activities: An ExampleCleaning crews are often on the critical path. They are a relatively low cost input, are typically tightly scheduled to achieve maximum productivity and are therefore allocated little interval time to move between jobs in peak periods. A delay at one aircraft, or when travelling between aircraft, will extend the critical chain on the next departure which will always delay that departure. Time cannot be made up without cutting what is delivered on the critical path. This is demonstrated in the diagram below:
Figure 1: The diagram shows example activities on the critical path of a turnround and the effect of delayed cleaning activities:
|
Resolving Punctuality Problems
Finding solutions, on the ground, to poor punctuality starts with establishing the critical path within the turnround process and its associated sub-processes. Then the exploratory work starts: identifying the constraining factor(s) and finding ways to optimise and protect the critical path.
Sounds simple? In a complex chain such as the aircraft turnround process, this requires comprehensive, methodical research and analysis that explores primary and secondary information – combining observation, interview and time study techniques.
Participation from the various parties involved in the end-to-end turnround is vital, to understand their respective processes, costs, constraints, performance objectives and fluctuations in the performance of their process. Turnround problems can rarely be resolved by a single stakeholder because of the dependent tasks involved in the process.
The next step is to take action to resolve the root cause of persistent delays. This could mean many things: reducing or eliminating the individual constraint; protecting the bottleneck resource or creating additional buffers.
Protecting the Critical Path to Consistently Improve Performance
Finding and fixing the root cause of delays is a good start. To maintain punctuality performance the critical path has to be constantly protected, even at additional cost to individual supporting processes. All resources need to be in the right place at the right time, poised to make their ‘contribution’ to the turnround process on-time.
The time required to complete supporting tasks for a turnround may fluctuate. For example, catering vehicles may be required to travel through control posts to access the aircraft. Control post access times may vary greatly, and may delay the arrival of catering vehicles to the aircraft side and therefore the loading process, which forms part of the critical path. |
Fluctuations are inevitable in the airline business: sector times, landing slots, taxi times even passenger loading times do vary. Where there are statistical fluctuations affecting dependent events, to be efficient it may be necessary to accept sub-optimal performance of some operations, for the greater good of the end-to-end process – and ultimately on-time departure.
This means creating ‘feeding buffers’ - making sure operations that feed the critical path are buffered where necessary to be certain they will never delay the critical path.
Minimising or eliminating the impact of performance fluctuations, by applying such buffers, is central in making sure actions supporting the critical path are never a cause of delay. In an efficient operation, these feeding buffers are determined scientifically; calculated based on the range and probability of fluctuations in the performance of the process. In lean terms, this is relieving the constraint or bottleneck.
Today this concept is considered counter intuitive in most organisations where managers are judged by “local” rather than organisation-wide measures of efficiency for their department. Typically, however, the savings gained from predictable performance more than compensate for any additional local costs to support the critical path as long as buffer requirements are determined scientifically.
The next step is to reduce the fluctuations in processes so that buffers, and therefore costs, can in fact be minimised. Returning to the control post example for the catering trucks, if the average control post process time is 10 minutes, 90% of events take less than 15 minutes and 99% of events less than 25 minutes, a punctual organisation may allow for 25 minutes in its planning. The cost of punctuality can clearly be greatly reduced by getting to 99% in less than 20 minutes.
Once the principle of optimal total performance is accepted, it is possible to calculate the optimal resource and equipment levels by department to achieve a desired probability level of on-time departure. Following this, financial decisions can be made concerning the value versus the cost of punctuality.
Being On-Time Every Time
It is not enough to have an ‘ideal’ process in place because fluctuations in the timing and duration of events at an airport and the availability of resource mean a plan formed rigidly around the schedule is a plan to fail.
Maintaining on-time performance requires consistently well planned and managed turnrounds, where the end-to-end process is treated as an entity that unites and jointly coordinates all parties, equipment and infrastructure, with appropriate buffers, and each phase of the critical path remains on-time. In an ideal world, each turnround would be as smooth as a Formula 1 pit stop.
On-time performance is a job for managers but they need support from the operational research and industrial engineering teams to deliver an appropriately designed and planned turnround process.
back