Breaking Boundaries
Dave Jesse on May 3, 2016

As both avid readers of my blog will know, I am very keen on data sharing. This allows us to learn from each other, both to find out what we are doing badly and to help to identify best practice. Indeed, data driven analysis of safety is now recognised as the mainstay of current aviation safety initiatives.


Size Matters

One of the problems of data sharing is that all too quickly the largest of databases appears to thin out and we are left with too little information and struggle to obtain meaningful statistics.

Let me explain by way of an example. In the last six months, Flight Data Services analyzed about 600,000 flights. From this selection, only 128 flights landed at Kyaukpyu airport (illustrated above) and of these, only 22 of these landed on the southerly runway. 22 samples is barely enough to start making analyses of the characteristics of the runway and certainly not enough to start trending over time.

Even with common airports, if we are interested in operations of an unusual aircraft type, or at particular times of day, it is easy to reach low numbers of operations. Hence, the larger the data pool the more chance our conclusions will be meaningful.



(I originally wrote “Coagulation”, but thought that blood clotting was perhaps inappropriate for an aviation safety blog!)

To gain size in the database, it is necessary to group together data from multiple sources. Only by consolidating data into larger sets can we start to overcome some of the problems of sparse data. This is a well-recognised problem for corporate operators, for example, who may not fly to some airports more than once or twice.

The largest program of this type is the Aviation Safety Information Analysis and Sharing (ASIAS) program in the USA where the FAA have encouraged US operators to pool their data, and some 11,000,000 flights have now been collected into a single huge database.

There is a drawback in this approach. While the FAA are reaping the rewards of work done for the US operations, this does not cover operators from outside the USA flying into American airports or flights by American airlines flying overseas. What is needed is a level of consolidation that has no state boundaries and that will allow us to start looking at aviation on a truly global scale.

Flight Data eXchange (FDX)

This is the goal of IATA’s FDX program. The International Air Transport Association took the view that only by pooling data on a global scale will the aviation industry be able to view safety issues that transcend national boundaries. An ambitious target, but one which no other organisation (except possibly ICAO) could take on.

Initial Steps

With such a mountain to climb, IATA took the pragmatic decision to start in one location so that they obtained good coverage of the region and could have some impact, rather than spreading their efforts thinly across the globe. They started their work in Latin America, where they have already made an impact, and they are now set to expand rapidly elsewhere. Part of their challenge will be to manage the speed of growth and expectations across the world.

Telling the World

One of the problems of all large scale data sharing exercises is that the participants worry that lessons learned might leak into the public domain in a way that could be misinterpreted, and so reflect badly upon them. For this reason it is difficult to advertise successes and hence difficult to identify and quantify the impact that these programs make. There is room for improvement here, especially as the lessons learned may have wider applicability.

In Summary

My view is that this is the only opportunity the aviation industry has to achieve a flight data sharing platform that breaks free from state boundaries, and for this reason it is important that the industry gets behind this program to bring it to its full potential. There can be only one global program, and this surely has to be it.