Curfew infringement prevention

We use AI to minimise flight delays, which impact airlines’ punctuality and the passenger experience (the combination of which is covered in Regulation (EC) No 261/2004 on passengers’ rights), and may generate night curfews imposed at many European airports.

The airport regulations are then integrated into the Network Manager Operations Center (NMOC)’s morning briefings on the day of the operations. Regulation Outlook AI uses information from airport events in NM’s “Airport Corner” and from NOTAMs, as well as weather and traffic forecasts to estimate the likelihood of a regulation being required at any of the network’s busiest airports. It already predicts the hours when regulations are more likely - with 84 % accuracy. Moreover, it addresses concerns about “AI explainability” by providing the rationale behind the prediction, be it atmospheric phenomena, industrial action, works in progress at the airport, over demand, etc.

AI models currently used, essentially deep neural networks, are of such complexity that it is practically impossible for their creator to understand their inner workings. Explainable artificial intelligence is a set of processes and methods that allows human users to comprehend and trust the results and output created by machine learning algorithms.

The MIRROR operational decision support tool was used as a sandbox during the development of the Curfew AI module. It currently provides NM operational staff with a real-time view of each aircraft’s flight legs, thus allowing identification of potential knock-on delays. The Curfew AI module will be made available in the coming months to MIRROR’s several hundred industry users following an update of the MIRROR interface.

In addition, an application programming interface running on EUROCONTROL’s cloud infrastructure will provide a broader access to the Curfew AI module, in particular to airlines for integration into their flight operational centres as a support tool.