Abstract
Function allocation, mode errors, and misuse of automation have become increasingly important research topics as technology has become more sophisticated. These and related problems will emerge as critical research issues as self-organizing, agent-based automation becomes more prominent. While current research has studied how people control a small number (2 to 10) of agents, the future will likely introduce the challenge of supervising hundreds of agents. Complex automation that is built from hundreds of loosely connected intelligent agents may exhibit powerful new adaptive behaviors. This emergent behavior may not be an intuitive extension of individual agent behavior. Addressing the challenge of this new technology will require a theoretical understanding of human behavior that goes beyond a task-based description of well-defined scenarios. Several important research questions include: What is the best way to support effective management of the agents? How can potential problems among the agents be identified and predicted? What display concepts are most effective for controlling large numbers of agents? How can the emergent agent behavior constrained so that it is comprehensible to human operators?
Get full access to this article
View all access options for this article.