An adaptive framework for configuration of parameters in an Industry 4.0 manufacturing SCADA system by merging machine learning techniques

摘要

Industry 4.0 (I40) is characterized by a shift from traditional production systems, where the human supervisor or the operator is the main force behind all operations, to smart factories driven by artificial intelligence (AI) where machines and intelligent production techniques are taking over the control of operations. Machine learning (ML) is one of the artificial intelligence techniques introduced in the industrial sector for boosting flexibility, customizing production, transforming traditional production equipment to smart and intelligent systems but currently having relatively fewer applications than in the information technology (IT) environment. Factory operational tools such as Supervisory Control and Data Acquisition (SCADA) systems intensively used in production processes are also affected by this transformation. This paper uses a combination of some machine learning concepts and algorithms to develop a framework for the configuration of parameters in an industrial environment via its SCADA system, changing a simple traditional human-machine interface (HMI) to a self-configurable I40 compatible device and offering an innovative solution that contributes to the improvement of conventional manufacturing operational processes. The adaptive framework scope is to introduce small traditional manufacturing plants with limited resources to the use of basic smart manufacturing principles and AI concepts. The experimental results of a small industrial rubber manufacturer used as a case study in this paper clearly show the merit of the new framework built based on machine learning techniques.