Man/Machine Interactions in Intelligent Robotic Systems

This paper deals with an experimental study of optimal interaction between human operators and intelligent robots in systems for decision making and control. The study uses an adaptive computer which is able to observe a control task, learn the required responses, and assume control responsibility. The initial findings show that the factors of greatest importance include how predictable the intelligent machine is, how many errors it makes initially, how much information the operator receives about machine decisions, and what his feelings are toward the machine.