曝光台 注意防骗
网曝天猫店富美金盛家居专营店坑蒙拐骗欺诈消费者
influence the controller’s choice to use or not use the automation.]
iii. Decision aids
[Advances in computer science and artificial intelligence are
providing powerful new computational decision aids that greatly
expand the potential to support cognitive activities in air traffic
control, e.g., monitoring, problem detection, and planning.
However, the introduction of new concepts of operation, including
“free flight” and reduced separation minima, increases the
likelihood for novel or unexpected situations to occur. Because
automated decision aids are not very good problem solvers when
situations depart from those the systems have been designed for,
the automated aids are inherently “brittle”. Breakdowns in
performance can occur whenever attempts are made to automate
the decision making process and assign the controller the passive
role of following instructions. In short, one has to be very careful
to ensure that decision aiding does not prevent the system from
capitalizing on the insights and expertise of the controller. In
addition, one has to insure that sufficient time budgets are
available to a controller in the event a decision aid fails.. In order
for decision aids to be useful they must be able to communicate
with controllers in a form controllers can understand. Decision
aids, if they are a mystery to controllers, can actually slow
operations.]
APPENDIX A
FACTORS POTENTIALLY AFFECTING SEPARATION SAFETY
A-23
iv. Warnings/advisories
1) Flight path prediction
[ATC automation that seeks to communicate a warning of an
impending conflict directly to the controller requires that it
compute an aircraft’s future position using the aircraft’s flight
plan, performance, track, and wind data. However, the
accuracy of flight path prediction is critically dependent on
the accuracy and timeliness of the data, the variability of the
winds, the certainty of pilots’ intentions, as well as the
effectiveness of detection and warning algorithms .
Unreliability or unexpected variability in flight conditions
limits the look-ahead time for accurate flight path prediction.
If separation minima are reduced, an error in flight path
prediction could result in a conflict with less time available to
achieve successful resolution..]
2) Conflict probe
[Unfortunately, many proposed conflict warning systems for
the controller (e.g., conflict probe) are vulnerable to false
alarms. The false alarm question is not trivial. As noted
earlier, a comparable warning system on the flight deck,
TCAS, suffered considerably in its early development from
problems of excessive false alarm rates that resulted in
mistrust and lack of pilot usage. At the same time, poorly
designed controller warning and advisory systems will also
foster disuse. What makes the false alarm problem more
critical when separation minima are reduced and/or traffic
density increases, is the necessity to insure that the reporting
threshold for a conflict alert/warning is sufficiently lenient that
few, if any, “true” conflicts are missed. However, if a warning
system is designed to minimize “misses” then the problem of
increased false alarms is immediately encountered. This is an
intrinsic problem in developing an effective and sensitive
conflict warning system.]
h. Controller errors
[Controller errors are generally defined in terms of loss of separation.
However, in considering risks associated with reduction in the
separation minima, controller error must include a much wider range
of inappropriate behaviors that result from mismatches between the
cognitive demands placed on the controller and the resources
available to solve those problems. To understand how the ATC
environment can increase cognitive demands in terms of controller
workload, memory, problem solving, decision making and the like is to
understand how that environment contributes to controller error.
Understanding the circumstances under which controller errors
develop, from both a systems and a behavioral viewpoint, should make
SEPARATION SAFETY MODELING
A-24
it possible to evaluate the consequences of ATC changes, like
reductions in separation minima on controller error and safety.]
i. Training
[A critical question in designing effective training concerns how well
training in a simulator or on the job transfers to actual performance
on the job. On-the-job training, as typically practiced in ATC
facilities, trains controllers on task elements that are identical to those
on the job. Although the concept of on-the-job training has been well
中国航空网 www.aero.cn
航空翻译 www.aviation.cn
本文链接地址：
a concept paper for separation safety modeling(68)
