In this study, a task analysis method that can evaluate human- machine interfaces for given task procedures is proposed. The method supports a framework to represent how the operator performs tasks with the interface. The description syntax sketches the structure of given tasks indicating the relationships that the system variables have among themselves within each elementary task. Several important cognitive aspects of the task and the interface, that are related to the mental workload and interaction complexity, can be revealed by the features of this syntax. These aspects are possible interference between elementary tasks, the existence of prerequisite tasks, and the need to refer the result of remotely previous tasks. Based on these aspects, we evaluate interfaces and then suggest a better interface.
A computer aiding tool has been developed to support the description of tasks and identification of the cognitive aspects with the framework. The new framework and the computer aiding tool were used to analyze examples drawn from DURESS, a thermal-hydraulic process simulation developed by Vicente as an exemplary complex human-machine system, to show the advantages of the proposed analysis method.