Adaptive Automation: Using secondary task measurement to determine function allocation

Moh, Heng Seng (2012) Adaptive Automation: Using secondary task measurement to determine function allocation. [Dissertation (University of Nottingham only)] (Unpublished)

[img] PDF - Registered users only - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader
Download (1MB)


The present study investigated the behaviour of secondary task performance in Adaptive Automation to determine whether it could be used as a sensitive index of operator’s mental workload to trigger changes in support system. The majority of this study is to verify the method that is been used in performance-based automation is appropriate for which it used secondary task measurement as an indicator to initiate changes in LOA to assist human operator in the right time with right amount. This investigation did not start a new experiment but used experiment data that was done by previous researchers. Previous experiment used simulation software called Automated Cabin Air Management System (AutoCAMS) which simulates a life support system in the operational context of spaceflight. During the simulation, participants were required to ensure system stabilization and repair all 10 faults when occurred. Meanwhile, participants also need to respond quickly to 2 secondary tasks.

With the experiment data, two simple analyses were made which are identifying the pattern of secondary task performance around the occurrence of fault and determine the effectiveness of using detector in secondary task measurement to detect fault events. These analyses were based on secondary task performance should worsen whenever fault event occurred due to additional workload. However, the result from the analysis did not appear to be inclined with expectation that the patterns were random and the detector could not accurately detect fault event but some false alarm. It was concluded that secondary task performance could not correctly indicating operator’s mental workload that could be used to trigger initiation in support system.

Based on the result, secondary task is very sensitive that could be easily disturbed by any variation within the system. One of the factors that was discussed is the interference issue caused by additional tasks that share the same input, output, and time which resulted in performance impairment. Meanwhile, human operator self-defined his own method to manage all the tasks concurrently such as anticipation that resulted in no change can be observed in secondary task measurement.

To decrease the effect of disturbance on secondary task measurement, some refinement on the simulation and analysis were suggested. Those refinements include redesign secondary task, use other method of analysis, and other approaches of measuring operator’s mental workload. Unfortunately, limited time assigned to this study caused further investigation could not be made.

Item Type: Dissertation (University of Nottingham only)
Depositing User: EP, Services
Date Deposited: 08 Apr 2013 13:30
Last Modified: 19 Oct 2017 13:13

Actions (Archive Staff Only)

Edit View Edit View