Towards Meaningfully Integrating Human-Autonomy Teaming in Applied Settings

This paper describes a theoretical model of apply and integrating human-AI teams in applied contexts. The model was developed from a thorough review of empirical findings of human-AI teams revolving around five central pillars including individual differences, training, autonomy transparency, autonomy reliability, and levels of autonomy. Additionally, the theoretical model introduces bi-directional transparency between the human and AI team members to help bridge the gap between the human and AI understanding one another. The interaction between each pillar and the outcomes associated with the use of the theoretical model are discussed along with its extensibility to future developments and ability to scale to a variety of different contexts.