Citation

Dunlay WJ, Horonjeff R. Transp. Res. 1974; 8(3): 205-217.

Copyright

(Copyright © 1974, Elsevier Publishing)

DOI

10.1016/0041-1647(74)90007-0

PMID

unavailable

Abstract

This paper seeks to define and estimate the frequency of aircraft interactions (called conflicts) which entail controller intervention. The task of conflict detection is viewed as a stimulus-response process in which the strength of stimulation is a particular closest-approach separation between aircraft, and the corresponding probability of response is the fraction of times controllers judge that separation to be a potential violation of the 5 nautical mile minimum separation standard. Data from human factors studies of air traffic control are used to estimate response probabilities for a wide range of closest-approach separations. Two types of response probability are defined. The first type (normative) predicts when controllers should intervene based on an analysis of human and equipment errors in the air traffic control process. The second type (descriptive) predicts when controllers actually intervene based on data from real-time simulations of conflict detection. These response probabilities are incorporated into an empirical model for estimating the expected number of conflicts in a specified time using data from ATC flight-progress strips.