Experimental investigation of submersible buoyancy regulation and support task performance under different levels of automation and varying workloads

A submersible mechanical support operator is mainly responsible for monitoring and handling the submersible Buoyancy Regulation and Support (BRS) system. Operators are frequently faced uneven task frequencies due to unpredictable failures, leading to different Workload Level (WL) and potential safety risks. To enhance the task performance of submersible, this study explores adjusting the Level Of Automation (LOA) under different WLs to achieve effective allocation of human-machine functions. This study designed 7 LOAs (LOA1-LOA7) and recruited 36 subjects for experiments to analyze the impact of different LOAs and varying WLs on submersible BRS system task performance. This study primarily focused on determining the optimal LOA under three WLs by examining five key aspects, including subjective WL, work performance, Situation Awareness (SA), eye movement indicators, and Electrocardiogram (ECG). The results showed that under Low Workload Level (LWL), LOA2, LOA6, and LOA7 significantly improved work performance, but LOA6 and LOA7 resulted in a state of underload. Therefore, LOA2 was identified as the optimal LOA for LWL. Under Medium Workload Level (MWL), LOA4, LOA6, and LOA7 significantly improved both subjective WL and work performance. However, LOA6 and LOA7 also significantly reduced the SA levels of the subjects, making LOA4 the optimal LOA for MWL. Under High Workload Level (HWL), although LOA2 to LOA7 significantly improved the accuracy and response time, LOA6 and LOA7 notably reduced the subjective WL level and work time percentage. At LOA7, the subjects exhibited SA, the largest saccade amplitude, and the lowest saccade velocity. Therefore, LOA7 could be identified as the optimal LOA for HWL. The research findings underscore the importance of rationally considering dynamic function allocation strategy to enhance submersible operator performance under varying WLs.