Supervisory control in a dynamic and uncertain environment: laboratory task and crew performance

Automation, Task, and Context Features

“…Specifically, a Markov decision process was employed to model automation use as it related to three autopilot design featuresthe ability of the autopilot relative to the crew's ability to control the helicopter, and the time it took to either engage or disengage the autopilot-and three task context features-the duration, time criticality, and frequency of secondary editing tasks. This modeling effort was prompted by an earlier lab study finding that crews did not use the autopilot for its intended purpose-as a task-offload aid (Kirlik, Plamondon, Lytton, Jagacinski, & Miller, 1993); that is, pilots did not consistently delegate the flying task to the autopilot when engaged in a secondary task. The model suggested that the reason for pilots' behavior was a mismatch between the intended strategy for autopilot use and characteristics of the autopilot (long engagement time and constraints on its ability to fly as fast as with manual control) and the task context (frequent and time critical editing tasks).…”

Dynamic Decision Making

“…Psychological research on dynamic decision making began with Toda's (1962) pioneering study of human performance on a game called the "fungus eater," in which human subjects controlled a robot's search for uranium and fuel on a hypothetical planet. Subsequently, human performance has been examined across a wide variety of dynamic decision tasks including computer games designed to simulate stock purchases (Ebert, 1972;Rapoport, 1966), welfare management (Dorner, 1980;Mackinnon & Wearing, 1980), vehicle navigation (Jagacinski & Miller, 1978;Anzai, 1984), health management (Kleinmuntz &Thomas, 1987;Kerstholt, 1994), production and inventory control (Sterman, 1989;Berry & Broadbent, 1988), supervisory control (Kirlik, Plamondon, Lytton, & Jagacinski, 1993), and fire-fighting (Brehmer & Allard, 1991).…”

Intelligent Multimedia Multi-Agent Systems

“…Thus when an intelligent agent has to accomplish a task which involves interaction with the environment, it creates an intemal model of the environment through an encoding process, performs mental computations on the contents (symbolic or subsymbolic) in this intemal model, and extemalizes the output through a decoding process. As noted by , Kirlik et al (1993) and Suchman (1987) most studies in traditional cognitive science do not separate extemal representations from intemal representations or equate representations having both intemal and extemal components to intemal representations. This confusion often Ieads one to postulate unnecessary complex intemal mechanisms to explain the complex structure of wrongly identified intemal representation, much of which is merely a reflection of the structure of the extemal representation.…”