The effects of the type of rest breaks on return-to-task performance in semi-automated tasks with varying complexities

• Main Author: Hoyi, Zandile
• Format: Others
• Language: English
• Published: Rhodes University 2017
• Online Access: http://hdl.handle.net/10962/44490

Automation in the aviation industry is acknowledged as a useful tool in reducing pilot workload (Hoh, Smith & Hinton, 1987; Beringer & Harris Jr., 1999). Typically, the role of the pilot (operator) shifts from active participation in a process to a task of monitoring the system with the resumption of control should the automation ‘fail’ (Byrne & Parasuraman, 1996). Unfortunately, the skills necessary to do so would likely degrade from non-use, during this process (Landry, 2012). This project investigates the “attentional demands” for the human operator during interaction with semi-automated operations of the flight. According to Dr Abbott (1996), FAA human factors specialist, one of the problems causing disharmony between crews and their automated systems is the incorrect upset recovery, owing to the human being out-of-the-loop (OOTL) from the system. Recovery, or rather return to task, is the ability of the pilot to loop back into control, once situational awareness has been decreased due to lack of alertness and a decrease in arousal. Different types of rest tasks are commonly prescribed fatigue countermeasures in the industrial setting and have been showed to elicit beneficial effects on prolonged human performance. Understanding the effects of different rest break activity and time out-of-the-loop during semi-automated flying on return to task performance has been adequately studied, thus highlighting its importance in the context of flight safety. The present study requested participants to perform a tracking task in a laboratory where they changed from activity (30 minutes) to a break (2 vs. 30 minutes) and back to the activity (20 minutes). The task varied in the complexity of the activity (pure tracking vs. tracking plus memory plus rule-based decision making), the type of break (passive rest vs. actively supervising) and the duration of the break (2 minutes vs. 30 minutes). Performance was measured as effective response time in the tracking task and number of correct responses to secondary cognitive tasks. Physiological measures included heart rate (HR), heart rate variability (HRV- time and frequency-domain), eye blink frequency and duration. The Karolinska Sleepiness Scale was used as a subjective measure. With regards to the most appropriate rest break tasks, the study concluded that active, administrative tasks, which allowed the operator to maintain some form of situational awareness by monitoring the automated system, achieved favourable effects of being more alert than the passive rest break of being disengaged from the system. In terms of the most appropriate rest break durations, the shorter duration of being out-of-the-loop from controlling the system proved to be more advantageous than the longer out-of-the-loop duration. In looking at the workload levels of arousal, the results suggest that the higher workload level is better at maintaining the alertness of operators. This study functions as a foundational framework for future investigations around the topic of human-automation interaction, looking specifically at return-to-task performance.