De-Fixating the Operator: Current Behavior Models Reviewed

Abstract Human controllers take a prominent place in the supervision and control of complex dynamic systems. Despite the achieved high safety standards there is a widespread acceptance that human error and thus erroneous behavior is inevitable in the end. Current efforts to cope with erroneous behavior are based on behavior models using a control perspective. Since these models cannot process erroneous behavior, an alternative perspective to model human behavior is proposed, namely autonomy. For autonomous living systems the critical variable is sense of influence.