Human Factors Analysis of Safety Alerts in Air Traffic Control.

摘要

Controllers receive several types of alerts from Air Traffic Control (ATC) automation systems that warn of potentially hazardous situations, including Conflict Alerts, Mode-C Intruder alerts, and Minimum Safe Altitude Warnings. This report provides a human factors analysis of ATC alerts and recommends changes to the alert algorithms and presentation that should increase controller effectiveness and overall system safety. We collected automation data from en route, approach control, and tower facilities that show how often alerts occur, how controllers respond to alerts, and when controller actions occur relative to the alerts. Of all the alerts examined, the majority received no response from controllers; many were so brief that controllers must have resolved the situation prior to the activation or the alert situation resolved itself without action by the controller. Of the alert situations where actions were taken, controllers most often took action before the alert activated. The results suggest that (1) many alerts are valid according to the alert algorithms but do not provide useful information to controllers, (2) these nuisance alerts are extremely common in the field, and (3) high nuisance alert rates may desensitize controllers and lead to poor performance. We recommend that the Federal Aviation Administration address the problem of nuisance alerts by improving safety alert algorithms, improving alert presentations, and providing better alert suppression functions. To improve safety alert algorithms, we recommend using data from this study to obtain better measures of critical reaction time parameters for alert algorithms. To reduce the impact of nuisance alerts, we recommend using alert presentations with multiple levels of urgency.