This paper outlines an analytical framework to perform a data-driven, risk-based assessment of Air Traffic Control (ATC) facilities. Safety associated with an ATC facility is modeled as an influence network using a set of risk factors. A novel hybrid approach employing Adaptive-Network based Fuzzy Inference Systems is introduced to propagate the model. Statistical analysis of system-wide data for each risk factor is performed to identify outliers and understand underlying distributions. They are then used to define Fuzzy Membership Functions for model variables. Analytical Hierarch Process (AHP) is used to determine rules required by the model's inference engine. Finally, the methodology is applied to a set of ATC facilities using real data.