Open Cognitive Control System Architecture

In recent years, the automation industry is moving towards Industry 4.0 and cyber-physical systems. There is an increasing demand for smarter plants and plant modules. In this paper, we are proposing an architecture for the open cognitive control system. This will provide the benefits of a control strategy based on the data analysis. With the possibility to connect with the cloud, this will enable easy updates and exchange of task modules that would ensure the system to keep up with the technological advancements. This will also improve the accuracy and lifetime of the automation system. To meet the requirements of flexibility, speed and reliability of the system, the architecture has been based on distributed and concurrent systems design. A detailed description of the architecture along with its challenges have been presented in this paper. A comparison of different implementation techniques for certain blocks has also been discussed.

I. Introduction

Mankind has been developing and finding ways to make their life easier since the early ages. Industrial automation is one such big step that was the outcome of the Industrial Revolution. With the invention of microcontrollers and processors, hardware-software co-design has given rise to an intelligent automation system. For current state-of-the-art systems, the localized software-hardware combination provides deterministic real-time control. This poses a problem for the flexibility and scalability of the plant. Distributed-system based architectures were introduced to the plant to get flexibility in the plant control like the PABADIS: Plant automation based on distributed systems approach [1] . Modularized plant architectures were also introduced to be able to have flexibility, ease of installation and fast time to market [2] . With the increasing number of sensors and actuators and a higher demand for flexibility and scalability, various controller architectures came into the picture. One such architecture is Future Automation System Architecture (FASA), which provides an approach for a flexible, scalable, platform independent and real-time capable execution framework [3] . Another such architecture is using the virtualization technology in the automation industry, that allows running of different applications with diverse operating systems on the same hardware while providing them with the necessary resources [4] .