An exploratory research on European ATM Network resilience through supervised learning

Kurzfassung

Air Traffic Management (ATM) is a complex system challenged by different aspects such as layered management subsystems, various business models for airlines, and dynamic environment. These aspects expose planned operations to daily disruptions leading to delayed passengers. Disruptions on one hand and system saturation on the other hand, suggest that ATM needs to become more resilient. Through a data-driven approach, this thesis works on conceptualizing the European ATM network resilience. Since system resiliency is correlated with situational awareness, this study proposes a network state definition and then explores learning methods to extract better predictions (delay and delayed flights).