Time and control in joint human-machine systems

Many proposals to enhance human control of dynamic systems focus on interface design and the provision of specific support for the interaction between humans and machines. Examples are process displays, alarm systems, computerised procedures, decision support systems, etc. Although it is essential to get the interface between humans and machines right, two things should be kept in mind. First, a focus on the interaction implies that the activities can be described as a sequence of discrete exchanges of input and output. In reality, the control of a process is a continuous flow of activities that reaches into the past as well as into the future. Second, the ability to remain in control is a quality of the joint human-machine ensemble, rather than of humans (or technological artefacts) alone. System design should therefore support not only people's internal cognitive activities but also the external collaboration among humans, technologies, and organisations. System design should specifically consider how the joint system may fail to achieve its objectives and how the human-machine ensemble can lose control. On the level of the joint system, control can be lost if the available resources - and especially time - are insufficient to evaluate the current situation and select the next actions. In order to understand how such conditions may arise, it is necessary to have a workable model of the dynamics of control in a joint human-machine system. The paper presents an extension of the Contextual Control Model, which previously has been used to describe both individual and team performance in process control. The extension provides a way of modelling how external variability may affect the dynamic equilibrium that is necessary for efficient control, hence provides an enhanced basis for the design of joint systems.