Abstract
Supervisors of highly automated systems monitor the performance of numerous sub-systems and intervene only when a sub-system fails. Such intervention, or fault management, requires the collection of information from multiple, dispersed, data sources. Fault management is supported by automated data monitoring, data location, and response recommendations following sub-system failure. The present study compared fault management across various levels of automated information management in a multi-task environment. In Experiment 1, participants presented simple problems and relatively ineffective assistance, continued to consult the primary data displays with little performance improvement and no automation bias. In Experiment 2, when required to manage more complex data with an apparently reliable decision aid, performance was not improved by the decision aid but responses to false alarms increased. Workload assessed by performance on concurrent tasks did not decrease in either experiment.
Get full access to this article
View all access options for this article.