ABSTRACT

Systems which aim to benefit from the technologies developed in the emerging Augmenting Cognition (AugCog) field employ a closed-loop architecture to optimize interactions between the operator and the system. In the terms of Augmented Cognition the closed-loop nature of the system reflects the real-time monitoring of the status of the operator, and the estimation of the ability to cope with current and possibly future task demands. Knowledge of the status of the operator enables strategies to be employed which seek to focus the skills of the operator on time-critical task performance. Implicit in the operation of such as system is knowledge of the current tasks, goals and consequently intents of the operator. These enable strategies designed to mitigate excessive workload to be targeted in a context sensitive fashion to support the overall efficiency of the man-machine platform. A number of techniques may be employed to mitigate the excessive workload of operators. Broadly these fall into the categories of: attentional management, adaptive interfaces, adaptive automation, decision support, task scheduling and context sensitive support of responses. Critical components of any system which is designed to dynamically change the manner in which control interactions are employed are those issues of operator expectancy and trust. The primary mechanism for maintaining levels of trust in system operation is the development of a shared mental model between the operator and the system. A common knowledge base, or understanding of the principles of operation of the system by the operator, supports the development and consolidation of the mental model, and leads to increased trust in the system which will be reflected in high levels of predictability. The aim of implementing such a system is that the operator will be able to rely upon the systems as they may rely upon other team members. Indeed in common with many systems with multiple operators, well designed AugCog applications enable the distribution of tasks and consequently workload. This is effective team working, and enables a single operator to perform the tasks of many.