Quantitative assessment of workload and workability in simulated submersible control tasks

A submariner is responsible for controlling heading, attitude and depth in a manned submersible. The control tasks are typical multi-window tasks and the workloads (WLs) are often imbalanced. These easily cause potential human-machine control risks. Therefore, the real-time monitoring and evaluating of submariner status should be emphasized in a manned submersible. In this paper, two indices, Workload Index (WLI) and Workability Index (WAI), are built to evaluate simulated submariner status. With 30 subjects, experiments involving six simulated submersible control tasks were conducted at six different WL levels in order to evaluate the behavior of the proposed indices under varying workload conditions. The experimental results showed a significant strong correlation between the WLI and NASA-TLX scores ( r r m = 0.847 , p < 0.001 ), and a moderate correlation with accuracy ( r r m = − 0.367 , p < 0.001 ). A significant strong correlation was found between the WAI and accuracy ( r r m = 0.930 , p < 0.001 ), along with a moderate negative correlation with NASA-TLX scores ( r r m = − 0.620 , p < 0.001 ) and a weak negative correlation with average working time ( r r m = − 0.178 , p = 0.029 ). Overall, WLI and WAI showed systematic variations across workload conditions and meaningful relationships with subjective workload and task-performance measures under the present simulated task conditions. Based on these findings, thresholds for WLI and WAI were established respectively, which could be applied for the further dynamic function allocation.