Without the States agreeing, the fundamental change will not happen

There are four major system transformation steps: moving from analogue to digital air traffic management technology; connectivity through a network; accumulation and aggregation of data and the use of algorithms to put intelligence into this data.

In terms of the analogue to digital, we're moving to an information-centric approach, with a horizontal layered architecture where we decouple the various layers according to the standard information technology (IT) setup. You have a data layer, an integration layer and an application layer.

With the second step of connectivity, we have to look at Europe and ask: how connected are we? We're not connected at all. We still have independent area control centres and connectivity is only done by the controller talking to the pilot, changing frequency and then contacting the next area control centre.

Then the next step is aggregating data. Aircraft, for example, transmit GPS-based location information via automatic dependence surveillance – contract (ADS-C) which is fantastic because we don't need to ping an aircraft anymore with the radar and it means we can collect this information across the network and optimise its flight through the airspace. We can now put intelligence into the entire network instead of reducing it to a sector of airspace; so with two aircraft flying towards each other at 500 miles per hour we can manage the separation much earlier – within the network.

If we have all this information, we can then apply algorithms for separation in the airspace and eventually the entire network. There is a lot of talk now about artificial intelligence (AI) in air traffic management but how do you apply AI if you are not connected and if you don't have access to all the data? To use AI to optimise management of airspace we need to have those first three steps.

At Skyguide, between Zurich and Geneva centres, we have introduced e-coordination measures. We are also sharing radar data across borders, but again this is still just point-to-point; we don’t have access to all surveillance data in the network. Operationally we’re still independent and there’s no information flowing from one area control centre to the next to improve the network. Our operational concepts still optimise the management of a single sector. So even though we try to move towards the pan-European network system (PENS) and the digital backbone of a network that provides that connectivity, we are not there yet.

It's the second step – connectivity. We must get away from the idea that we need locked and closed local ways of managing airspace. Why do we think that way? We have reached the limit of our ability to increase capacity by reducing the size of sectors even further and putting more controllers in place. I see this as the biggest hurdle because as soon as we talk about finding a pan-European solution to this, some will say: “yes, but only if I have control over it”. The lack of centralised or shared service is one of the main roadblocks in the drive towards developing pan-European ATM services.

It would be a completely different system and all the roles will change, especially ANSPs. Instead of us providing information about aircraft location, this will be done by the aircraft itself and that completely challenges the core of what an ANSP is. We can do something along these lines of trajectory-based operation in Switzerland, but we have only a small airspace; it needs to be done at European level.

No, but we are closer than we were five years ago, especially with the latest update to the ATM Master Plan. There is an expressed wish from customers and suppliers to go in that direction. And it is good that we now have bigger players, countries like Germany and France with a critical mass to drive the process forward.

The tipping point is not there yet but there are good signals coming from ANSPs who are unhappy with the systems they have today – the old technology is much too costly and becoming unsustainable. And of the big three suppliers, two are committing to it.

The Virtual Centre programme has been identified by the Performance Review Commission as a “flagship programme”, but now we need to develop the economic regulatory side of things to encourage the adoption of these new technologies. Aviation is still state-driven, state-mandated, state-supported. So in Europe, even if ANSPs push in a certain direction the states will still try to maintain their control or their influence, particularly on the economic regulatory side where we need to set the incentives correctly. We don't see any change in the model of how money is earned in this regulated environment with the tariff system. The old thinking is still prevalent: the longer you are in a country’s airspace and the heavier the aircraft is, the more money you can make. By doing that we ignore the real cost drivers.

Airspace is defined geographically, with each state having a mandate for sovereignty reasons of up to 50km of airspace, to manage, maintain and control it. National interests, national security issues are all part of that so any changes to the status quo will be a challenge.

But it might not be impossible. Skyguide is a pioneer in this area here because we manage a chunk of French airspace and some parts of Southern Germany and Western Austria. This required bilateral agreements in which both sides benefit.

If we call it “nirvana” it will never happen – so I prefer to rephrase this question as: “How do we cope with the capacity crunch in 10 years, and when we have crewed and uncrewed airspace users flying?” Today's model is not sustainable.

I remember a few years ago speaking with one of the suppliers and he said: "Modern IT is not used in air traffic management. We could do so much more if it were. I have software that is old and ageing, and I can't find enough new people who can maintain that software. I have customers who are really unhappy because they want more and I can't deliver it. Or I can only deliver it with extensive delays because of the complexities of the old system. My production costs are exploding. In summary, this is not a sustainable business model."

But I think this is the way forward and we need the suppliers to change their approach. That will then have an impact on regulation, because safety regulation is not yet ready for digitalisation. The concept of a virtualised machine does not exist in today's regulations so we need to address that, too.

And economic regulation needs to incentivise that kind of investment because the Performance Review Commission is still incentivising monolithic equipment rather than modern IT services and the underlying old business model which disconnects cost drivers from revenues.

Now, this transformation process is not easy and at Skyguide we have had many problems. Many are because we are still using legacy systems from 30 or 40 years ago.

If other ANSPs would follow our example and say “OK, I’m not going to put my iTEC or iCAS or my TopSky system into each individual area control centre but move it to a location-independent data centre and then provide the service to the other area control centres.” That would be the first step.

The UK has tried to do this in its Swanwick and Prestwick centres but in the end found it was not possible yet.

It should be happening. The question is when? That is the real issue. Perhaps we might have to wait for some capacity crunch which will force the politicians to take an interest. But I'm an engineer. I say, if you see a problem we should try to solve it and not wait until something happens.

The Airspace Architecture Study was for me a breakthrough document and that was only made possible in 2018 because of the problems we had in Karlsruhe that summer.

It is not something I look forward to but something will have to change in the governance model of this industry.