Using AI-powered delay prediction to improve daily airline operations

Operator using FADE

By successfully implementing FADE, our Forecast of ATFM Delay Evolution tool into the EUROCONTROL Network Manager's operational systems, EUROCONTROL is now successfully using AI-powered delay prediction to improve daily airline operations across the European aviation network.

FADE is an excellent example of swiftly deploying innovative tools into operational use. It was developed by the EUROCONTROL Innovation Hub (EIH) to answer an airline ‘pain point’ which had been pitched by airlines to the EIH in the first ideation cycle of the EUROCONTROL Air Transport Innovation Network, our user-driven innovation portfolio.

Airlines asked us whether we could help them understand better how delays evolve when a flight is affected by an air traffic flow management (ATFM) regulation, as currently they only obtain limited information on the resulting slot evolution. That makes operational planning not as predictable as they would like, and can have major negative impacts on operational and economic performance, and especially passenger experience (everyone flying hates the uncertainty that delays trigger!).

The result developed by the EIH after six months of developing, and a summer of testing, is FADE, a tool that uses AI and machine learning to provide airlines with two delay-prediction indicators – delay trend (i.e. the probability for ATFM delays to decrease) and predicted delay.

13 airlines in total tested out the R&D prototype last summer (Air France, Transavia, SAS, Vueling, Swiss, Turkish Airlines, Wizz Air, KLM, Air Europa and easyJet in an initial phase, with TUI fly, Austrian Airlines and SunExpress joining them after positive feedback from the original participants).

After positive airline feedback – airlines felt FADE’s predicted delay data helped them plan better, providing flight crew and ground handlers with more accurate information on when to board, and prioritising necessary reroutings – the tool was deployed in the NMP Flight application within the EUROCONTROL Network Manager’s operational systems. The NMP Flight application receives delay predictions from Cloudera, the Network Manager’s Advanced Analytics and AI platform.

For an idea of what the tool does, and how airlines have been using it, check out our video with Camille Anoraud and Ramon Dalmau-Codina from our Innovation Hub, alongside Jens Albek, an operative route analyst at SAS, Mathias Stavroulis, a Transavia flight dispatcher, and Juan Torres Martin, a flight dispatcher at Vueling.

The next step will be the creation of an API for direct integration of FADE into airlines' operational systems.

Latest highlights


ISOBAR – Research is building controller confidence in AI systems

Group picture IRINA kick-off meeting

Driving research & innovation forward with the integration of RPAS into European airspace


ECHO project led by EUROCONTROL successfully delivers a ConOps for higher airspace operations

Aircraft taxiing and being directed

New EUROCONTROL contrail observatory at our Innovation Hub will support contrail avoidance research

Unmanned aircraft over a city

Bringing Urban Air Mobility to the cities of the future: EUREKA set to pave the way for vertiports

  ISOBAR - Winner in the SESAR Exploratory Research category  5th from the left: Ramón Dalmau Codina (EUROCONTROL - Innovation Hub)

ISOBAR and 4DSkyways projects win at the Digital European Sky Awards