What happened at Rome Fiumicino (FCO)?

In June this year I wrote a controversial article in my blog about whether ECTRL A-CDM improves departure punctuality or not. Using Association of European Airlines (AEA) punctuality reports I analyzed the data of A-CDM airports.

Recently I was asked about my article and I quickly updated some airports. I had to double check FCO, because since the local A-CDM go-live, the departure punctuality of the AEA carriers has dropped remarkably (see chart).

 
Does anybody know what is going wrong at FCO?

Does EUROCONTROL A-CDM improve departure punctuality?

For those who prefer executive summaries, just scroll down.

Don't get me wrong. I am a full supporter of Collaborative Decision Making (CDM) whether it is driven by EUROCONTROL, FAA, SESAR, an airport operator, a local ATC, or even a software vendor. However, CDM, as I understand it, goes beyond data exchange. It is the participation of the airport community when it comes to decisions of the airport about the development of future infrastructure and in daily operation.

According to publicly available information improvements by ECTRL A-CDM include reduced taxi times, fuel savings, better slot adherence, less late gate changes, less network delays, and higher arrival and departure punctuality. The departure punctuality might be the most important element in this list for an airport and its stakeholders. Why? Because airports, airlines and service providers need to allocate reserve capacities for strategic (already planned buffers) and tactical (operational day) delays.

In my recent post about airport punctuality myths I mentioned that I conducted some research about the effect of ECTRL A-CDM on departure punctuality. There were two problems with this. First, as already pointed out in the other post, departure punctuality measured in percent is an imperfect metric. And I am still going to use that impudently. Secondly, there is not much data publicly available.

For the following analysis I used data from the Association of European Airlines (AEA) who themselves receive data from airlines and Central Office of Delay Analysis (CODA). Unfortunately, the data only covers AEA airlines and it does not include passenger figures. But for the sake of this analysis I assume that AEA airlines traffic is the most critical (in terms of turnaround process for instance).

What did I analyze? I investigated all A-CDM Airports according to the DPI implementation status and split them in two categories: (1) Fully compliant (DPI operational) and  (2) Locally implemented. I was expecting the following effect from an A-CDM implementation: (a) sustainable improvement of punctuality after implementation, (b) solid punctuality during adverse conditions (i.e. last winter), (c) continuous improvement, i.e. actual year at least as good as last year with more traffic or better than last year with less traffic and (d) a positive punctuality trend in the last 12 months.

Executive summary:

The following table lists the investigated airports and the findings. Example charts of individual airports can be found below.

In a nutshell:

• Being an 'A-CDM' airport does not mean that your departure punctuality improves sustainably.
• Even though the number of flights decreased at most European airports, punctuality has not improved.
• Most A-CDM airports spiral downwards in adverse conditions, i.e. winter operations.

And one more thing which I noted during the analysis:

• It is a crying shame that we work in an industry which publishes departure times and at most European airports only about every second flight actually departs on time.

Here are some example charts of the investigated airports. If you would like more of them, just let me know - jweder (at) neuropie.com.

 

Three Myths About Airport Punctuality

While Ryanair is boasting to be the most punctual European airline, Oman Air celebrates one day of 100% on-time performance on April 26 this year. Unfortunately there is no public data that proves any of these claims.

What about airports? In April 2011 London Heathrow published a record 94% departure punctuality.  Now they are Britain's worst airport according to the Civil Aviation Authority (CAA).

I think it's time to wade through some of the myths and confusion surrounding airport punctuality.

1. Punctuality is an internationally standardized metric.

Punctuality targets are usually defined in terms of 15-minute punctuality, i.e. a flight is counted as departing on-time if the plane goes off-blocks within 15 minutes of the scheduled time of  departure. But what is "15 minutes"? Is it more than 14'59" or more than 15' or even more than 15'59"? Now some airports and airlines may disagree on this but even ACI, IATA and AEA don't. If they do, then it will probably be 15'59", stealing another minute from the passenger, airline and airport capacity just to look better.

And, have you compared off-blocks times of airport and airline? Some airports check and validate differences within +/- 1 minute, most don't. Result: airport and airline report different punctuality figures.

2. Punctuality is a great metric to benchmark.

In previous posts I was highlighting some flaws of punctuality as a metric.

• Cancellations of delayed flights improve punctuality
• Indicator ignores size of aircraft and number of passengers affected
• No assessment about the financial loss possible
• Mix of performance which can be influenced with exogenous factors
• Difficult to determine improvement measures

Why is it still used, even as internal target? Because it is easy to understand and when it is bad, it can be easily attributed to exogenous factors.

3. Eurocontrol A-CDM improves airport punctuality.

There is no evidence that ECTRL A-CDM improves punctuality. Yes, it might reduce taxi times but with a negative effect on departure punctuality. With the lack of evidence I did some research myself and present the findings in my next post.

I know it is easy to lament. What do I have to offer instead? An approach and metrics to actually address the issue together with other stakeholders. Because the airport can have that 360° view and not just the silo view by every other stakeholder.

10 characteristics of Airport CDM culture

In their implementation manual Eurocontrol defines the requirements for an 'Airport CDM culture':

• Agreed relevant data should be shared between all partners involved
  at the right time.
• Data shared should be of sufficient quality to facilitate improved traffic
  predictability and planning capabilities for all partners involved.
• Decisions should be made by the partner best placed to make them.
• Decisions made should be shared with all other partners.

This sounds good, in theory. But how do we know? Here are 10 characteristics on how you can evaluate whether you have a CDM culture or not:

1. Complete openness (well, I understand if you want to keep your financial data privately, but airlines: what is the problem with sharing your pax figures?)
2. Admit your mistakes and errors (airport, please lead by example)
3. No finger pointing
4. Common and shared database (e.g. AODB as a central hub for flight information)
5. Mixed working groups and task forces, so everybody participates in finding a solution
6. Goodwill of all partners to find solutions
7. Joint design of measures for improvement
8. Experimental approach / test runs (do not underestimate the power of trials!)
9. Free discussion of findings (the goal is always the same: we want to improve)
10. Unbureaucratic implementation of measures

Who is afraid of EC Regulation 691/2010?

The EUROCONTROL Performance Review Commission (PRC) has been analyzing the performance of the European Air Traffic Management System since 1998. Initially, the focus was on en-route ATM performance. Some indicators from this framework were selected in 2010 for regular monitoring of operational airport ANS operational performance. Three of the flight efficiency indicators have been adopted as airport capacity indicators for Reference Period 1 (2012-2014) in the SES II performance scheme:

• Air Traffic Flow Management (ATFM) Arrival Delay
• Taxi-Out Additional Time
• Arrival Sequencing and Metering Area (ASMA) Additional Time
  

So this is all about ANS and EUROCONTROL.Why, despite delivering data to EUROCONTROL, should an Airport Operator bother about it? Well, it may fall back on them.

Let us take the Taxi-Out Additional Time as an example. The indicator measures the delta between the Actual Off-Block Time (AOBT) and the Actual Take-Off Time (ATOT). Now if you us airline data, they may report very long taxi times. Longer times than the Airport Operator. Why? Firstly, they may measure the AOBT differently because they use a different source than the airport, for instance ACARS. Secondly, since stations are evaluated against outbound delays, they will try to 'optimize' the AOBT to remain within the 15 minutes IATA threshold and report accordingly to their headquarters. You know what I mean.

Now here comes the problem. Whoever will get the blame (and potentially penalties) about long taxi times will have to demonstrate that it was not them causing it. As an Airport Operator, would you like to engage with the parties about this sort of questions when the performance regime has been implemented and the results have been published? Maybe the Airport Operator will already want to engage in the process to make sure that  questions about data quality, measuring standards, including reasons and impact of delays, is being addressed now.

Cost of delays in Europe 2010

Eurocontrol is publishing delay statistics of airports participating in the ECAC (European Civil Aviation Conference). In their recent report they provide an overview of the delay situation in 2010. This report was used by a number of news agencies (in German only, apologies) to list the worst airports in 2010.

I used the report to assess the total delay cost of ECAC in 2010. According to the information I retrieved from the report, adding a few assumptions on the average size of aircraft and seat load factor, I came up with impressive figures about departure delay cost:

In 2010 there were more than 7 billion Euros opportunity cost to passengers, more than 2 billion Euros operating cost to airlines and more than 400 million Euros operating cost to airports.

And I did not even include the cost of cancellations which were estimated at more than 100,000 flights at the April peak of the volcanic ashes crisis.

The real potential of Airport Collaborative Decision Making

The current scope of Airport Collaborative Decision Making (A-CDM) is focused on Air Traffic Management (ATM) problems, i.e. the turnaround process and departure sequencing process. In the words of a CDM project manager: "I do not care if a flight is delayed for one hour, as long as I have a predictable and accurate departure sequence."

This is limiting the full exploitation of the CDM philosophy.

To fully unleash the potential we should include ALL processes at the airport which include aircraft, passenger, baggage and cargo. Furthermore, CDM is not about information sharing only but also about communication and objective critique. Before making a decision in a silo, for instance a short term change of baggage belt, that functional area should consult immigration and the ground handler. And, if things go wrong in a process consistently (based on facts), we should work together on an action plan to improve.

Because, in the end, isn't CDM about excellence in the way we operate and perform, together and as individual stakeholders?

How to fix airport delays

Airport operators sometimes tell me that because of the interconnected structure of air travel, any real solution to resolve delays needs to be system-wide. What they mean is that government and airlines need to come up with long-term plans and actions to increase capacity and remove congestion. The industry legitimately applauds NextGen, A-CDM, and other national/international airspace programs.

What can and will airport operators do now to improve the situation? They will certainly act as soon as they are being fined over delays. For instance, Dublin airport  has been fined for delays to passengers passing through security and consumer dissatisfaction. Yet, this a fine for delays processing passengers through the terminals but not for flight delays. In fact, flights can still be delayed without queues.

Just like tackling any other problem, let's first start to believe we can resolve it. Maybe the problem requires innovation to surmount. Maybe we need first to understand the impact of the problem which is, for instance, the cost of delay rather than on-time performance in percent. And, maybe we can use that data to convince our partners in joining improvement programs and recognize that the airport itself is a system which can resolve issues.

Resolve reactionary delays

Nearly half of all flight delays in Europe are actually "reactionary" or "knock-on" delays caused by previous late departures or arrivals, according to a new study of Eurocontrol. Unfortunately the study is an analysis and does not, despite some suggestions on adjusting the framework for reporting delays and improving the quality of the delay data captures, provide any means to tackle this issue.

We have come up with an approach to resolve reactionary delays. We found that we can attribute the delay cost to the original causer of the delay. That causer may still be within the airline (e.g. aircraft maintenance) but it is a much better decision tool than just 'reactionary delays'. It is still about responsibility here ... nobody feels responsible for reactionary delays.