Tarja Kettunen, ISA Software Ltd.

Extended Abstract

One of the primary constraints to the efficiency of air traffic today and in the foreseeable future is insufficient airport capacity. Previous airport capacity studies have concentrated purely on runway capacity, without looking at the impact of actual operations on the surface, although an increased interest in surface operations can be seen in the Advanced Surface Movement Guidance and Control Systems (A-SMGCS) domain. To truly analyze the gate-to-gate efficiency of any future concept, it is clear that surface operations and the effect of surface-related problems, ultimately measured as delay, need to be incorporated in the macroscopic system-wide view. The predictability of surface operations is also of great interest in the context of trajectory-based management, widely agreed as the ATM strategy of the future. This study, carried out for the United States Federal Aviation Administration (FAA), provides the data from which suitable parameters can be identified to model the influence of surface operations in future operational concept validation experiments.

The Airport Operations Analysis complements the FAA’s Airport Capacity Benchmark reports for 31 of the busiest US airports by analyzing the actual surface operations for departing and arriving flights at each of the 31 airports. The resulting metrics are based on the scheduled carrier operations data that the airlines themselves are required to file monthly and which is available from the US Department of Transportation’s Bureau of Transportation Statistics (DOT-BTS) Inter-modal Transportation Database (TransStats). The data analyzed is one month of flight operations taken from the database for September 2000 and is considered to be well representative as it constitutes a major part of the selected airports’ traffic. Excluded are general aviation and military operations, non-scheduled flights, international traffic and cargo operations.

Airport operation metrics are produced with the objective of ascertaining generic operating time distributions and airline-specific operating time distributions for each airport. The analysis is carried out both for the whole analysis period and for hourly data periods to allow for the influence of heavy demand periods that are typical of many of the airports being considered. Airline operations at their own hub airports are of significance since the predictability and accuracy of surface operations at hubs have great impact on the airlines operations on a NAS-wide level. High-level comparisons are made across the different airports to show the numbers of movements of each type (departure / arrival) and the associated ground operations times (taxi-in, taxi-out). Comparisons across airlines are not considered relevant, since their operating conditions at any given airport are not comparable (agreed gate locations affecting taxiing times, specific terminal buildings owned by a given airline etc.).

The results show that taxi-out times at a given airport are not necessarily proportional to the amount of operations. This would tend to suggest that the impact of the airport infrastructure is significant. The three New York Metropolitan airports show the largest taxi-out time in the analysis period, with La Guardia (LGA) showing an average taxi-out time as high as 35 minutes.

Taxi-in time rankings by airport follow the tendency of the taxi-out times, except that the taxi-in times are significantly shorter and the range is not considerable even between the lowest and the highest (from 3 minutes at San Diego airport to only 11 minutes at Dallas Forth Worth airport).

The analysis tries to find explanatory factors for the operating times considering the airport infrastructure and operational procedures, however, no data is available to identify the operating configuration for these airports for each of the days that were considered, so we assume typical operations as described in the FAA airport benchmark study.

In the future steps of the study, standard statistical analysis techniques will be used to produce predictability estimates for ground operations at each of the airports. It is expected that the predictability will differ by the time of day (by the traffic load).

The results obtained from this analysis are currently being used in our on-going concept validation experiments to develop the concept of use for management by trajectory and the allocation of 4-D trajectory-based contracts for strategic management of the US airspace (CONUS).