The U.S. Navy’s P-8A aircraft performs missions in a team-based manner, requiring clear and concise communication among specialized individuals. However, students often train individually. In this case, instructors role-play the other crewmembers, while also managing scenarios and assessing students. Integrated autonomous agents offer a technical solution to allow students to perform within the simulated training environment, while mitigating instructor workload. Ideally, automated agents track mission phases and work in concert with verbal commands from the student. However, to maximize the benefit of autonomous agents, it will be critical to ensure that there is an appropriate level of automation transparency to support instructor situational awareness (SA) and trainee performance without over-burdening the instructor. This likely has impacts on SA, trust, workload, and performance.
A study was conducted in order to examine the effects of automation transparency on situational awareness, workload, performance, and other associated variables. Forty-five individuals participated in a study where they were required to manage drone teammates in a simulated fishing task. Participants monitored three drones that provided information about the area and verbally directed two drones to move and collect fish. Participants were limited to only verbal communication to represent aircrew coordination. Performance, SA, workload, and automation trust were assessed across three levels of automation transparency (low, medium, and high). Overall results indicated no significant impacts to performance or situational awareness dependent on transparency level. However, higher levels of automation transparency negatively impacted workload. Although more research is needed into automation transparency, this study provides a base to begin understanding how autonomous agents may affect students and instructors, and how these interfaces should be designed.
Keywords
Additional Keywords
Trust in Automation, Automation Transparency, Situational Awareness, Workload, Synthetic Agents, Autonomous Agents