A Decade in Review: Situation-Awareness-Based Agent Transparency and Human-Machine Teaming

Over the past decade, the Situation awareness-based Agent Transparency (SAT) framework has become a foundational model for the design of effective human-machine systems. As machine autonomy increases, the opaque nature of artificial intelligence can impede collaboration. The SAT framework addresses this challenge through a three-level structure—Perception, Comprehension, and Projection—that clarifies an agent’s reasoning and intent for human partners. This article reviews a decade of research applying the SAT model across domains such as human-robot interaction and autonomous driving. Findings consistently indicate that SAT-based transparency enhances team performance, improves operator situation awareness, and fosters calibrated trust, often without increasing cognitive workload. The model has evolved from a unidirectional tool to a dynamic, bidirectional paradigm that enables genuine human-agent teaming. Future research should focus on developing adaptive transparency, refining human-to-agent communication, and investigating long-term effects to create more effective and responsible systems.