The Robustness of Human Advantage in Swarm Leader Identification
Event Type
Virtual Program Session
Human AI Robot Teaming (HART)
TimeWednesday, October 12th2:15pm - 2:30pm EDT
DescriptionControl of robotic swarms through control over a leader(s) is a dominant approach. Resilience to attrition is a primary advantage attributed to swarms yet the presence of leader(s) makes them vulnerable to decapitation. Algorithms which allow a swarm to hide its leader are a solution. In prior work we found that an Adversary NN trained to identify a leader performed substantially better than human observers. When the swarm was trained to hide its leader, however, the advantage reversed. The present study investigates the robustness of human leader identification by testing identifications made in the presence of medium and high levels of visual clutter. Clutter degraded human performance to some extent but human accuracy remained well above that of the Adversary in deception conditions. This study confirms the robustness of the human superiority effect and argues for the inclusion of humans in AI systems which may confront learned deception.