Modeling human attention and performance in automated environments with low task loading

Automation has the benefit of reducing human operators' workload. By leveraging the power of computers and information technology, the work of human operators is becoming easier. However, when the workload is too low but the human is required to be present either by regulation or due to limitations of automation, human performance can be negatively affected. Negative consequences such as distraction, mind wandering, and inattention have been reported across many high risk settings including unmanned aerial vehicle operation, process control plant supervision, train engineers, and anesthesiologists. Because of the move towards more automated systems in the future, a better understanding is needed to enable intervention and mitigation of possible negative impacts. The objectives of this research are to systematically investigate the attention and performance of human operators when they interact with automated systems under low task load, build a dynamic model and use it to facilitate system design. A systems-based framework, called the Boredom Influence Diagram, was proposed to better understand the relationships between the various influences and outcomes of low task loading. A System Dynamics model, named the Performance and Attention with Low-task-loading (PAL) Model, was built based on this framework. The PAL model captures the dynamic changes of task load, attention, and performance over time in long duration low task loading automated environments. In order to evaluate the replication and prediction capability of the model, three dynamic hypotheses were proposed and tested using data from three experiments. The first hypothesis stated that attention decreases under low task load. This was supported by comparing model outputs with data from an experiment of target searching using unmanned vehicles. Building on Hypothesis 1, the second and third hypotheses examined the impact of decreased attention on performance in responding to an emergency event. Hypothesis 2 was examined by comparing model outputs with data from an experiment of accident response in nuclear power plant monitoring. Results showed that performance is worse with lower attention levels. Hypothesis 3 was tested by comparing model outputs with data from an experiment of defensive target tracking. The results showed that the impact of decreased attention on performance was larger when the task was difficult. The process of testing these three hypotheses shows that the PAL model is a generalized theory that could explain behaviors under low task load in different supervisory control settings. Finally, benefits, limitations, generalizability and applications of the PAL model were evaluated. Further research is needed to improve and extend the PAL model, investigate individual differences to facilitate personnel selection, and develop system and task designs to mitigate negative consequences.

Department dc:contributor.department

Massachusetts Institute of Technology. Institute for Data, Systems, and Society.

Grantor dc:publisher

Massachusetts Institute of Technology

Year dc:date.issued

2016

Author dc:creator

• Gao, Fei, Ph. D. Massachusetts Institute of Technology

Advisor dc:contributor.advisor

• John Carroll.

• Institute for Data, Systems, and Society.
• Engineering Systems Division.

Statement dc:rights

• M.I.T. theses are protected by copyright. They may be viewed from this source for any purpose, but reproduction or distribution in any format is prohibited without written permission. See provided URL for inquiries about permission.

Licence dc:rights.uri

• http://dspace.mit.edu/handle/1721.1/7582

Language dc:language.iso

eng

Handle dc:identifier.uri

http://hdl.handle.net/1721.1/106592

OAI identifier oai:identifier

oai:dspace.mit.edu:1721.1/106592