|M.Sc Student||Ahuvia Shir|
|Subject||Navigation Learning in Automated Systems: A Comparison of|
Methods for Keeping the Operator "in the Loop"
|Department||Department of Industrial Engineering and Management||Supervisors||Professor Avi Parush|
|Professor Anat Rafaeli|
The main issue of the current research is the potential degradation of spatial knowledge acquisition as a result of automation implementation and the “out of the loop” problem. A suggested model of navigating with automation is presented in this work. According to this model, there is a continuum of reliance on automation, from under reliance up to over-reliance. We assume that a medium level of reliance on the automation will keep the operator in the loop, and will encourage attention to environmental information as a basis of high level spatial knowledge.
The present study employed two methods. The first method lowered the availability of the automation by presenting self-location indication by user request and not continuously. The second method presented "quizzes", where the participant is asked to indicate his present location.
103 participants performed navigation tasks in an on-screen virtual complex building. First, participants practiced navigation in 16 tasks, in one of four conditions. Then, participants were tested for their spatial knowledge.
The main finding is that a combination of the two methods yielded a higher level of spatial knowledge. The presentation of quizzes during the practice phase produced a higher level of spatial knowledge. However, displaying self position by user request, by itself, did not produce an improved learning. While the manipulation of quiz presentation directed the subjects explicitly toward developing survey knowledge in an active way, the manipulation of self position display still allowed the subjects to remain in a passive role.