Queuing Network Modeling of Human Multitask Performance and its Application to Usability Testing of In-Vehicle Infotainment Systems.

Human Factors;Cognitive Modeling;Industrial and Operations Engineering;Engineering;Industrial and Operations Engineering

Feng, RuijiaSarter, Nadine B. ;

University of Michigan

关键词: Human Factors;    Cognitive Modeling;    Industrial and Operations Engineering;    Engineering;    Industrial and Operations Engineering;   

Others  :  https://deepblue.lib.umich.edu/bitstream/handle/2027.42/113590/fredfeng_1.pdf?sequence=1&isAllowed=y

瑞士|英语

来源: The Illinois Digital Environment for Access to Learning and Scholarship

【 摘 要 】

Human performance of a primary continuous task (e.g., steering a vehicle) and a secondary discrete task (e.g., tuning radio stations) simultaneously is a common scenario in many domains. It is of great importance to have a good understanding of the mechanisms of human multitasking behavior in order to design the task environments and user interfaces (UIs) that facilitate human performance and minimize potential safety hazards. In this dissertation I investigated and modeled human multitask performance with a vehicle-steering task and several typical in-vehicle secondary tasks. Two experiments were conducted to investigate how various display designs and control modules affect the driver;;s eye glance behavior and performance. A computational model based on the cognitive architecture of Queuing Network-Model Human Processor (QN-MHP) was built to account for the experiment findings. In contrast to most existing studies that focus on visual search in single task situations, this dissertation employed experimental work that investigates visual search in multitask situations. A modeling mechanism for flexible task activation (rather than strict serial activations) was developed to allow the activation of a task component to be based on the completion status of other task components. A task switching scheme was built to model the time-sharing nature of multitasking. These extensions offer new theoretical insights into visual search in multitask situations and enable the model to simulate parallel processing both within one task and among multiple tasks. The validation results show that the model could account for the observed performance differences from the empirical data. Based on this model, a computer-aided engineering toolkit was developed that allows the UI designers to make quantitative prediction of the usability of design concepts and prototypes. Scientifically, the results of this dissertation research offer additional insights into the mechanisms of human multitask performance. From the engineering application and practical value perspective, the new modeling mechanism and the new toolkit have advantages over the traditional usability testing methods with human subjects by enabling the UI designers to explore a larger design space and address usability issues at the early design stages with lower cost both in time and manpower.