Looking 10-20 years into the future, in the rapidly advancing world of autonomous, special ops, cargo, and Medevac aircraft, all of the basic aviation functions will likely be handled completely and competently by AI agents embedded within air vehicles. Under nominal conditions (and many basic off-nominal situations) the AI-controlled vehicle will operate autonomously and independently without input from on-board personnel. However, no mission is ever completely nominal, and many open questions remain about how onboard personnel and AI controlling the vehicle should collaborate effectively. Currently, Medevac is accomplished aboard expensive, scarce, specialty-built aircraft with large medical and pilot crews. With the nature of military conflicts evolving towards the potential of large-scale battles with large numbers of casualties and wounded soldiers, there is an emerging opportunity for the use of autonomous aerial vehicles to minimize crew needs. We developed a scenario in which a medic without any piloting or AI programming training, is teamed with an AI pilot to transport a patient aboard an autonomous aircraft.
We recruited 9 certified Emergency Medical Technicians and had them complete four simulated medical evacuation missions under various nominal and emergency conditions. All were generally successful at accomplishing the mission after basic training on the AI interface, however, workload and error rates varied with the complexity of the scenario. Users requested better transparency into the AI’s status, decision-making, and planning as well as a better interface. The most requested interface modality was voice and natural language to preserve the medics’ visual attentional resources for patient care.