Plug & play control: Control technology towards new challenges

Control engineering is in many ways a mature technology that has found its way into almost every industrial sector with an almost countless number of successful applications. Nevertheless, there still remains significant challenges to overcome that prevent the technology from further spread. One of these challenges is related to the fact that large industrial processes are always live systems in the sense that they are subject to constant change in terms of instrumentation (sensors and actuators) and in terms of subsystems that are added or removed. This means that an advanced control system might be based on a system model that is valid only for a rather short time span and might be turned off by the operator, when it seizes to operate satisfactorily for this reason. Moreover, most of the advanced design methodologies offered by our community are monolithic in nature, in the sense that the only way to modify the control system is to perform a completely new design of the whole system in response to what might be even tiny changes in instrumentation. Such a constant redesign of the whole control system is usually not feasible due to cost and commissioning issues. This paper is a position paper in the sense that it probably provides more questions than answers. The objective of the paper is to highlight some of the challenges that the control community is facing, if the advanced methods should be made applicable to a wider range of applications, in particular those that can be described as live systems for which sensors and actuators are added or removed, and likewise, for which subsystems are added or remove dfrom time to time. The challenge is defined as creating control systems that can automatically accommodate changes of this radical nature. A number of industrial case studies to exemplify the challenges are described.