The military operates in dynamic and complex task environments with inherent uncertainty that can make it difficult for human operators to form appropriate strategies when using autonomy. Uncertainty about system capability often results in disuse or misuse of autonomy, and as a result, performance and situation awareness are compromised. A key component in successful teaming is trust. Methods of transparency are emerging to support human-autonomy teams (HATs) by communicating information about autonomy’s actions, decisions, behaviors, and intentions to develop shared awareness and shared intent. Most efforts investigate transparency during task execution, but there are limits to how efficiently transparency content is communicated in these settings. DARPA recently conducted exercises with swarms of drones demonstrating the inability of an operator to respond appropriately despite system transparency. Given that and the ability to process at high speed, HAT operation might prohibit information delivery and operator perception and comprehension of the information about system behaviors, states, plans, and rationales before action must be taken. Consideration of transparency information during task performance can siphon valuable mental resources during high task load, and lack of spare mental capacity during these periods limits the ability of operators to learn and retain insights about how the autonomy functions. Addressing operator limits, implementation of transparency beyond mission execution must be sought. This paper presents research integrating multiple methods of transparency and applying them to different phases of a HAT lifecycle (e.g., pre, during, and post task) within the domain of counter small, unmanned air systems. Two important questions are explored: 1) What is the impact of this method across the HAT lifecycle to common ground, trust, communication, and reliance (e.g., pre, during, and post task), and 2) How do HAT interactions over time shape and change transparency communication needs?
Keywords
HUMAN FACTORS
Additional Keywords
Transparency, Trust, Human Agent Team, Counter UAS