Is data complexity a threat for flight safety?

“Everything simple is false. Everything which is complex is unusable.” This thought, originally expressed by philosopher Paul Valéry in 1937, might occasionally pop up when you work with data models or numerous guidelines in a complex domain such as ATC or AIS/AIM (Aeronautical Information Service/Management). Let’s use the data exchange model ‘AIXM’ as an example.

The success of AIXM

AIXM, introduced by Eurocontrol around 2000, stands for Aeronautical Information Exchange Model. It was designed to exchange aeronautical information (e.g. data on airports, routes and airspaces) between databases/systems. The success of AIXM is embodied by the European AIS Database (EAD), a single database that gives access to up-to-date AIXM data provided by all EAD member states. A growing number of States worldwide are publishing (a fraction of) their aeronautical data in AIXM format.

Before the introduction of AIXM the main source for aeronautical (‘static’) data exchange was the AIP (Aeronautical Information Publication) issued and regularly updated by each State. It probably still is the main source. The format of the AIP is designed for humans; it is too flexible for automated interpretation by systems. Humans are needed to find and interpret the appropriate data in the AIP and manually type it over in other systems (e.g. flight management systems). Mistakes during this task are likely, which is a serious hazard for aviation. Therefore automating data transfer using a data model such as AIXM is generally considered to improve data integrity and therefore flight safety.

AIXM too complex or too simple?

Now let’s get back to Valéry’s statement at the start of this blog. The main problem of AIXM is that the model is both too simple and too complex. Its complexity is mainly due to the design philosophy that AIXM should be useful for a wide range of applications. In itself a logical and laudable intention. However, since AIXM model designers cannot really know all the applications that will use the data, the chance that AIXM data does not really meet the requirements of each individual target application is huge. The required data may not fit in AIXM.

[blockquote text=”“Everything simple is false. Everything which is complex is unusable.”” text_color=”#004361″ show_quote_icon=”yes”][vc_separator type=’transparent’ position=’center’ color=” thickness=’5′ up=” down=”]

For AIXM designers this might lead to the conclusion that AIXM is still too simple. Their natural impulse is to extend or revise the AIXM model whenever AIXM fails to serve the needs of target applications. For system developers and data editors such revisions are sometimes very helpful, but in the same time it means growing complexity. They will increasingly fail to apply the model in the right way (or it will cost too much time).

These weaknesses become visible whenever you try to apply AIXM for new applications, such as eAIP (Electronic Aeronautical Information Publication), eTOD (electronic Terrain and Obstacle Database), or aeronautical charts. Parties modelling and exchanging aeronautical data for a specific application will often find it much easier and more efficient to use other formats than AIXM. This makes it questionable whether it was a good idea to develop a single exchange model for a wide range of applications.

The risks of complex models

Inconsistencies and poor efficiency due to the growing volume and complexity of data models and guidelines seem to be underestimated. These factors may be just as risky as human typing errors. People entering the data have no clue of how target systems deal with their input. The more time is spent on data modelling problems and reading guidelines the less time and focus remains for the collection and checking of the most critical data. Organisations often do not keep up with the growing data complexity. We might already have passed the point where most data model specifications and guidelines that are added to increase flight safety are actually doing the opposite, not because they are bad in itself, but because they add too much to complexity. It seems therefore a good idea if regulators and organisatijonons such as ICAO and Eurocontrol put much more focus on decreasing the volume and complexity of data models and guidelines.

About To70. To70 is one of the world’s leading aviation consultancies, founded in the Netherlands with offices in Europe, Australia, Asia, and Latin America. To70 believes that society’s growing demand for transport and mobility can be met in a safe, efficient, environmentally friendly and economically viable manner. For more information, please refer to

About post author

Leave a Reply

Your email address will not be published.

© Copyright To70 2021