Simultaneous Monitoring of a Number of Auditory Sonar Outputs

Abstract

A series of experiments is described in which performance under ``vigilance'' conditions was observed in a task which required simultaneous monitoring of a number of simulated active sonar outputs presented binaurally over headsets in the form of frequency-separated pure tones. The amplitude of each tone was independently modulated by a low bandwidth noise source. The signal to be detected was a 200-msec fixed increase in amplitude of one of the tones; this signal could occur on any of these tones at random. Performance levels were compared with those observed when only a single output was presented. Short-term studies showed that detection rate was markedly degraded by the requirement for multiple monitoring, and that the extent of this degradation was systematically related to the number of outputs presented. It was also found that degradation occurred even when only one specified output of a number simultaneously presented actually carried signals, and the remaining outputs thus constituted ``noise.'' In subsequent long-term studies conducted in a ``watchkeeping'' situation it was evident that the extent of the degradation when all outputs required attention was not reduced by extensive practice. When the sessions of multiple-output monitoring were prolonged in these watchkeeping studies, such within-session decrements in performance that occurred persisted over the 11-day observation period, but did not appear to be related to the time of day or night in any simple manner.