Coherent Copilot: Human Versus Synthetic Voice for Air Combat Performance

Autonomy, machine learning, and other forms of artificial intelligence have become commonplace in society by civilians and government entities. Human interaction with these machines is now vital for mission success, improved situation awareness, and increased performance in nominal conditions. These interactions can vary in complexity, causing exponential differences in user ergonomics and interfacing. The use case of these intelligent systems often dictates the level of user-friendly ability implemented. If the system is complex, human operators may become task saturated leading to a higher chance of an error occurring. To prevent fatal errors from occurring, human autonomy teaming can be leveraged, wherein shared intent is conveyed through collective outcomes between human and machine teammates. One method of accomplishing such a task is through communication channels with a voice automated command-response protocol. Some modern cockpit command systems use synthesized voices to facilitate communications, though recent research suggests that human-like, synthesized voices lead to overall performance improvements. Furthermore, interactions between the system and the user can be emboldened by utilizing a human voice rather than an automated synthetic communication medium. This paper aims to explore various methods of human communication and begins the plans to create an automated copilot system with advanced interactive skills. The coherent copilot system should have human-like communication skills, and uniquely adjust to user preferences. Future iterations of the coherent copilot project will include human experimentation, and analysis of results.