EUROCONTROL Network Manager invests in AI to turn data into business value

Aircraft blueprint.

After a successful first phase of digital transformation 10 years ago, the EUROCONTROL Network Manager (NM) is now engaged in a second digital transformation phase by developing a new data centric system architecture and leveraging machine learning and AI techniques. This will allow for the introduction of intelligence and automation in real-time business processes, augmenting human capabilities and supporting informed decision-making.

As with most digital technologies, artificial intelligence (AI) is now widespread. Learning, reasoning, heuristic search and problem solving algorithms are found in a very wide range of applications. Most industrial and economic sectors are deploying these techniques in their engineering methods and products. Artists are experimenting with AI in their creative tools.

The EUROCONTROL Network Manager is planning a technology breakthrough and AI will be one of the most important enablers for the next generation system. AI already contributes to – and benefits from – an accelerated momentum of technology development, which is opening a wealth of opportunities in many different fields. AI technologies help medical professionals and benefit environmental protection and monitoring programmes, in agricultural projects and in the modelling and management of cities, infrastructures and industries. They contribute to safer and more efficient mobility and transport systems, offering effective tools for multi-modal and multi-lingual interaction and information querying.

The new NM’s system architecture involves much more than just a technological upgrade of the system that performs the demand-and-capacity balancing within Europe’s air traffic management (ATM) system – it will be a key enabler for the entire and profound digital transformation of the continent’s ATM network, changing the way the European network is managed and the relationships between all aviation stakeholders. AI techniques will become the mediator between the users and this new digital world. The challenge will be to be able to accurately analyse and qualify the safety properties of components and systems to go above the current safety standards.

The first digital transformation of NM was already a catalyser for the ATM digital transformation. It allowed NM to make ATM data widely available via its business-to-business services (NM B2B), promoting interoperability based on open standards, which is a core goal of the Interoperability Strategy of NM, a goal aligned with the European Aviation Strategy and the Digital Single Market. The NM B2B services have stimulated creativity and innovation and have contributed to safer and more operational-efficient and cost-efficient services. They have also enabled new entrants to the market, creating jobs. They form the backbone required for the inter-regional data exchange that supports the global ATFM vision.

At the March 2019 World ATM Congress, the European Commission’s Single European Sky (SES) recognised the contribution of NM’s interoperability initiatives to the digital transformation of the ATM and granted the SES Innovation Award to the ''Interoperability with the EUROCONTROL Network Manager” initiative.

“This initiative recognised the achievements of NM and its ten partners – five air navigation service providers, three airports, one airline and a computerised flight plan service provider – in collaborating on a variety of digital data exchanges with NM. The interoperability initiative has to date delivered more than 100 services and connected to NM more than 200 systems from organisations worldwide, and demonstrated the tangible operational benefits of making ATM data digitally available.”

Idalina Mendes Videira Network Strategy and Development expert EUROCONTROL

The first phase of digital transformation capitalises on the Service-Oriented Architecture (SOA) technologies to redesign and open systems to achieve more levels of interoperability with customers. The NM’s Technology Division is now moving into the second phase to ensure that the huge amount of data being store can be exploited. NM has several years of archived data, which has been used for post-ops reporting, but not leveraged for supporting real-time operations. The idea is to put some intelligence behind the unexploited data to help humans increase their capability to understand the dynamic traffic situation they face and make the right decisions. AI will be at the core of this project.

Interested? Subscribe!

Never miss out on the latest developments from EUROCONTROL

NM has been adding increasing levels of automation to its system architecture over the last few years. The introduction of data science techniques to analyse the processing of flight plans in 2017, has been allowing to further automate the process and is a fundamental reason for the leap in the number of flight plan transactions which NM is able to process automatically.

The volume of data being processed by NM is rising

  Peak day in 2018 System sizing (WW ready)
Flights 38,000 300,000
Msg In 400,000 3 million
Msg Out 700,000 6 million
CPR 7 million 60 million
B2C/B2B requests 4.5 million 35 million
B2B Pub/Sub 8 million 100 million

With the application of new AI processes, NM will be able to understand the effect of an event before it happens. So if there is a weather phenomenon forecast or a sudden runway closure, NM will know the impact of this on the network and calculate the best course of action to take.

The introduction of new machine learning capabilities within the NM’s systems architecture will not just improve predictability, it will transform the way humans and machines work together to manage the network. Currently humans play the role of a stop-gap to fill the constraints of the system but this paradigm will need to be reversed so that the machine and the human can work in a complementary way. This means NM needs to understand better the issues of demand-andcapacity balancing through the actions of different actors. A complementary aspect, which will increase predictability, is the gathering of data from new sources. As an example, an experiment with ADS-B satellite information on aircraft positions before entering the European airspace demonstrated the increase of predictability of the system by up to 10%, which is a huge improvement.

NM is now receiving advanced information on flight operations in the USA from the Federal Aviation Administration (FAA) and Brazil using System Wide Information Management (SWIM). In the near future, more information could become available from flow control centres in the Middle East and Singapore.

The increase in predictability, resulting both from the use of AI and the use of additional sources of data, specially outside Europe, is expected to raise confidence in the system and to unleash unused capacity, which is currently kept in order to be able to deal with the high levels of uncertainty we have today.

As data is exchanged between stakeholders in increasingly automated ways, the door is open for new artificial intelligence capabilities to start changing the way all stakeholders in the system interact with each other at an operational level.

“We are starting to see that airlines are looking again at the way they are undertaking their flight planning. In the past, much of the work has been outsourced to third-party flight plan service providers. But as aircraft operators have seen the benefits of interacting directly with the NM via the NM B2B services – in terms of ease of use and low cost – they are starting to undertake more automated communications directly. Having the NM data gives the airlines the network situational awareness, which allows them to move from the old paradigm of “file and forget” to being an active actor in the overall traffic optimisation. A key enabler for NM to accommodate theses increasing levels of data exchange is to migrate our system to public-service cloud-based data providers. This will give us the required scalability and elasticity to cope with growth.

“To embrace the digital journey means that we have to manage changes at all levels, operational, technical and cultural. And it is not only the initiative of one party, we all need to work towards the same goals and this will mean a transformation of the way we work together. We will see new types of jobs, new opportunities and innovation labs (or digilabs) spreading out. These are not just to test the technology but more to establish a different human-machine interaction paradigm, in terms of user experience and, as well, to teach the machine and make it learn from our complex environment, which will be key to make AI technologies become a commodity in the European ATM network.”

Pierre Hanoune IT Architecture and Strategy expert EUROCONTROL

Read our full Skyway issue

Learn more about the new era for aviation powered by AI and our role in it.