Human and Computer Task Allocation in Air Defense Systems

Abstract
The advent of computer systems that offer direct support for high- level cognitive tasks has called basic assumptions about the proper respective roles of computers and humans into question. Reallocation of cognitive tasks from human to computer has sometimes resulted in user rejection of resulting systems. In cases where automation has been complete or nearly so, system performance may suffer from the lack of contributions best made by human operators. In either case, systems become overly complex and costly, and perform less than optimally. Within a general air defense background, a small experimental study has been conducted to examine basic issues related to the allocation of cognitive tasks in human-computer systems. Research questions include examination of (1) variables which determine the relative superiority of humans or computers, (2) the impact of information load on the optimality of human decision rules, (3) flexible versus fixed allocation schemes, and (4) issues related to the optimal locus of control of the task allocation process. A prototype research system was developed to simulate relevant aspects of the air defense environment. Experimental subjects acted as operators of the prototype system and made decisions which required them to identify each approaching simulated aircraft as friend or hostile. Four conditions were compared, ranging from a manual condition in which subjects must make all ID decisions to a completely automated condition in which the computer made all ID decisions.

This publication has 0 references indexed in Scilit: