Investigating the causality of potential collisions on the airport surface

Maintaining safe and efficient operations at airports while the National Airspace System grows and evolves requires understanding of potential hazards and the conditions that lead from these potential hazards to actual incidents and accidents. Such understanding is necessary to model the risks expected under future operational conditions, to develop mitigation strategies for anticipated risks, and to predict the effectiveness of the mitigation under a range of conditions. An essential step is to develop and validate predictive, quantitative models to characterize the relationships between causal factors and collision risk. Predictive models relate risk to factors such as traffic levels and airport properties, and support risk mitigation planning for the Federal Aviation Administration's (FAA) Next Generation Air Transportation System (NextGen) initiative. Faced with an absence of credible models for collision risk in airport surface movement, we developed methods for counting aircraft interactions and estimating parameters relevant to collision potential in previous work sponsored by the FAA's Office of Aviation Safety (AVS) [1]. In the current work, we describe our progress in developing quantitative, predictive models of risk as a function of airport conditions, based on the previous work. A key element is defining criteria for identifying a subset of aircraft interactions in the airport movement area as Potentially Hazardous Interactions (PHI), as described in other recent work [2]. Our research focuses on three main topics: how to predict rates of more serious but rarer incidents from the more readily-observed precursor events; finding the factors that are most strongly linked to changes in these rates; and systematically describing how the rates of PHIs vary between airports. These results are the first step toward predictive models that could help to guide strategies for technology development or procedural changes to mitigate these risks. We present results of applying these analysis methods to sample data based on surveillance data measured at multiple U.S. airports. These results constitute a basis for inferring the effectiveness of different operational improvements ranging from airport surface management practices to enhanced aircraft equipage benefits in limiting the rate of potentially hazardous operations under anticipated future airport traffic levels.