THE IMPACT OF AUTOMATED PLANNING AIDS ON SITUATIONAL AWARENESS, WORKLOAD, AND SITUATION ASSESSMENT IN THE MONITORING OF UNCREWED VEHICLES

Abstract

Uncrewed vehicles (UVs) are pervasive in civilian and military operations. While technological advances allow most UV functions to be automated, human operators are frequently tasked to monitor UVs and intervene when system capabilities can not meet situational demands – often occurring as unanticipated situations. When monitoring UVs, good situational awareness (SA) allows human operators to recognize the need for their intervention. Although increasingly automated systems limit the need for human intervention, operator SA may be compromised as task functions are allocated to a system resulting in poor monitoring performance. In UV operations, the functions of route generation and selection may be critical opportunities for operators to build situational awareness. If a human operator, rather than a system, completes the processes encompassed in these functions (e.g., the evaluation the environment and comparison of route alternatives), the task considerations evaluated by the operator might bolster their SA and allow them to anticipate when interventions will be required to support the UVs. Improved SA is expected to result in better monitoring behaviours, fewer accidents, and improved overall performance. To evaluate the role that route generation and selection functions play in the monitoring of UVs, participants completed a virtual UV monitoring task with assistance from an automated route planning system. The system operated at three levels of automation (LOA) where the generation and selection functions were completed by either the operator human or an automated system – at the low LOA the system was an interface for the participant to generate the route, at the moderate LOA it generated routes which were ranked by the participant, and at the high LOA the system generated and selected one route. Once a route was determined, participants monitored vii the UVs as they traveled their outlined routes and intervened if the UVs were approaching an unsafe area. SA, situation assessment (visual attention), perceived workload, and performance were evaluated between the three levels of the automated system. The level of the automated system did not impact any variable that was assessed, indicating that the automation of UV route planning functions may not be detrimental to operator SA or overall performance. Findings and limitations can inform future examinations into the relationship between the planning process and SA, as well as the allocation of functions to a human operator or automated system to optimize performance in UV operations.