The Effects of Interruptions and Retention Interval on Prospective Memory in Simulated Air Traffic Control

Abstract

In air traffic control (ATC), operational procedures or workload can prevent controllers from immediately completing task actions, requiring them to delay the task action until an appropriate future time. For example, a controller may identify an aircraft conflict (a violation of minimum separation between two aircraft), but be unable to immediately issue a resolution, thereby requiring them to remember to execute the resolution in the future. Evidence from ATC incident reports and controller interviews suggest that controllers sometimes forget to complete delayed intentions, and this has serious safety implications. Laboratory studies of prospective memory (PM) suggest that interruptions can impair PM performance and that retention interval can affect the probability of PM retrieval. Such effects would be highly relevant to ATC because controllers must remember delayed intentions over variable retention intervals and do so in face of interruptions from other safety-critical task demands. The aim of the present study was to examine the extent to which task interruptions and length of retention interval influence PM performance in an ATC simulation. The ATC simulation required participants to accept/handoff aircraft entering/exiting their sector, prevent aircraft pairs from conflicting, and perform a delayed-execute PM handoff task. The delayed-handoff task required participants to press an alternative response key when handing-off certain target PM aircraft instead of the routine hand-off response key. The interrupting task required participants to manage an additional ATC sector for 27s which occluded the primary sector display. The two within-subjects factors manipulated were interruption (interruption task or no-interruption) and retention interval of the delayed-handoff task (37s, 77s, or 117s). We found a high overall PM error rate (33%), which replicates previous studies that have found participants frequently make PM errors in simulated ATC. Strong Bayesian evidence in support of the null hypothesis indicated that the interrupting task did not affect PM performance. Interestingly, longer retention intervals only reduced PM performance when participants were not interrupted and thus performed the primary ATC task continuously. These findings suggest that the switch in task context caused by the interruption triggered the reinstatement of deferred task goals, thereby nullifying the impact of retention interval.