Data-Driven Approach for Runway Braking Condition Assessment with Forecasting Capability

Traditional runway condition reporting is limited due to its reliance on runway contamination information and pilot reports of braking action. A database of 4.9 million aircraft landings by Aviation Safety Technologies, labeled with runway condition codes computed from aircraft sensor outputs provides a unique opportunity to enhance and modernize condition reporting using data-driven methods. This paper introduces a machine learning model trained on this landing database, which predicts runway condition codes using a cascading Xgboost architecture. The method incorporates a novel multiple-ROC threshold setting procedure for linked classifiers which maintains the shape of the runway condition code distribution. Notably, the model can be used in a forecasting setting as it only requires weather information from METAR reports, a description of the runway, and aircraft type as input. To test its effectiveness, the method is applied to a collection of 30 historical runway excursion incidents, consistently assigning at best ”Medium to Poor” braking action to all cases with reduced friction. The model can serve as a valuable decision aid for aircraft operators, complementing traditional runway condition reporting. Additionally, it can function as a forecasting tool to inform runway maintenance decisions.