U. S. Department of Transportation
Download 2.66 Mb.
|
Executive SummaryRunway incursions are used to identify pre-collision behavior. Understanding those factors that increase the severity of a runway incursion may help identify situations that are more dangerous and potentially mitigate that danger. A runway incursion is defined as the unauthorized presence of a vehicle, pedestrian or aircraft on a runway. Runway incursions are rated according to severity: category D represents the least severe incidents (generally one aircraft) while category A represents the most severe (up to and including a collision). Incidents are also identified by who is “responsible” for the incursion: a controller, a pilot, or a vehicle. The purpose of this research is to examine the underlying factors that contribute to the severity of runway incursions. The research detailed in this report does not seek to explain the causes of particular events, but rather focuses on broader trends in incursion severity. Understanding those broader patterns can provide insight into policy-making and identify areas for future research. Prior to examining any data, a literature review was undertaken to identify hypotheses potentially relevant explanatory variables. However, little quantitative research has been done on runway incursions. Much of the research that has been done has been qualitative in nature. Some identified trends, but generally focus on individual events rather than broad factors that may influence severity. Thus, to the best knowledge of the authors, the research in this report is the first systematic statistical analysis of runway incursions. The analysis focused on the set of all runway incursions that occurred from 2001 to 2010. The FAA curated this dataset, which contains basic information about the incursion and related aircraft. One of the Volpe Center’s innovations was to combine multiple FAA and non-FAA data sources to incorporate information not available in the base dataset. These additional sources included the FAA’s Air Traffic Quality Assurance (ATQA) database and Operational Network (OPSNET) database, while weather and information on airport layout were gathered from other parties. A variety of statistical techniques were also used to examine the dataset. Due to the lack of previous research, much of the effort focused on cross tabulations of the data. This technique revealed interesting relationships among the variables both in terms of incident severity and incident type. A preliminary modeling effort was also undertaken. Some of the major conclusions drawn from the research are:
In addition to identifying factors that contribute to severity, this research effort identified areas for future research. Some of the research that could contribute most to an understanding of the risks related to runway incursions are:
Table of Acronyms
TABLE OF CONTENTS Section Page Appendix A:The most obvious is that there is an effect of experience. As pilots spend more hours in a make and model they are less likely to commit serious incursions. 66 Appendix B:An alternative explanation is that bad pilots do not ever get many hours in a make and model. Under this hypothesis, error rates are fairly constant across experience levels but pilots that commit many serious errors stop being pilots (e.g., they do not enjoy it, cannot get licensed). This would lead to lower hour pilots being concentrated in categories A and B rather than in C or D. 66 Appendix A:Runway Incursion Definition 209 Appendix B:Data Issues 214 Appendix C:Statistical Concepts 262 C.1.Two-way Chi-Squared Tests 262 C.2.Box and Whisker Plots 263 C.3.Kruskal-Wallis Tests 264 C.4.Interpreting Regression Output 265 C.5.A Question of Interpretation: Bayesian versus Frequentist Models 267 C.5.1.Frequentist Econometrics 267 C.5.2.Bayesian Econometrics 267 C.5.3.Making the Decision: Comparing and Contrasting 268 C.5.4.Conclusion 272 C.6.Extensions to the Multinomial Logit Model 273 Appendix D:Future Research 274 Appendix C:Summary of Modeling Results 276
TABLE OF Tables Section Page Section Page
|