ABSTRACT
Introducing automation into ground vehicles can benefit from lessons learned in many other domains, such as aviation and military systems. In this chapter, I first describe human–automation interaction failures that have led to accidents in aviation. I then present and describe seven features of human–automation interaction: (1) the degree of automation and its dependence on stages of human information processing; (2) automation reliability; (3) the kinds of errors that imperfectly reliable automation can lead to; (4) the concept of out-of-the-loop-unfamiliarity (OOTLUF) and its relation to situation awareness; (5) mode complexity; (6) automation transparency; and (7) flexible automation. A third section describes the research contributions, first outlining the strengths and weaknesses of accident analysis, and then focusing on experimental research in two domains: alerting systems and OOTLUF in decision aiding. In a final section, I suggest solutions to human–automation interaction challenges that have been identified in other domains related to flexible automation, transparency, and training.
