A HOLDING FUNCTION FOR CONFLICT PROBE APPLICATIONS

by Dave McNally and Joe Walton

Under current National Airspace System operations conflict
alerts for aircraft in holding patterns are often missed or
in error due to the fact that trajectories for holding
aircraft are not modeled in existing Conflict Alert or
Conflict Probe automation. In addition, a controller in one
sector may not know when aircraft are holding in a neighboring
sector. These factors can lead to an increased potential for
loss of separation while aircraft are flying in holding
patterns. The objective of this work is to develop a holding
function that automatically determines when an aircraft enters
a holding pattern, computes a holding region around the
pattern, and probes the holding region for conflict with
other traffic. Since controller workload is generally high
during periods when aircraft are in holding the operational
concept of use assumes the holding region is automatically
computed and the controller is alerted only if another
aircraft is predicted to fly through the holding region. The
holding algorithm is applied only to aircraft that would likely
go into holding during rush periods, e.g., arrivals to
capacity constrained airports. A holding region is computed for
aircraft that settle out on a steady outbound course. Initial
estimates of the holding fix position, turn radius, and
holding pattern leg length are computed and automatically
updated as the aircraft flies the pattern. The holding
algorithm was implemented in the Center/TRACON Automation
System software suite and tested using Host radar track and
flight plan data from the Fort Worth Air Route Traffic Control
Center. Of 37 aircraft that went into holding during a 1 hour
severe weather period, the holding function correctly computed
and updated 34 holding regions that accurately reflected the
holding pattern airspace. One of these aircraft was involved
in an operational error that would certainly have been
prevented had the 3 min holding alert been displayed to the
controller. In 3 cases the holding region was activated
incorrectly, then deactivated following subsequent track updates,
and later computed correctly on the second outbound leg. The
results show that Host track and flight plan data alone may
be used to automatically model and conflict probe holding
airspace in real-time.