Human-Autonomy Teaming and Explainable AI Capabilities in RTS Games

Real-time strategy games often times mimic the appearance and feel of military-like command and control systems. Artificial intelligence and machine learning research enjoys utilizing the environments produced by these games and are often focused on creating intelligent agents to beat them or accomplish high scores. Instead of creating these agents, or bots, to beat real-time strategy games, this work instead focuses on creating machine learning-driven agents to work with human players. However, due to the advancements in the deep learning field, machine learning models have become harder to understand, and are called ‘black-box models’ due to the inability to see the inner workings of such algorithms. To remedy this, we describe how we take human-autonomy teaming and explainable artificial intelligence techniques to shed light and provide inside for players to understand recommendations by the system.