Abstract:
The recent advancements in Information and Communication Technologies (ICT) and production systems have made manufacturing shop floors a data-rich environment. Massive volumes of data are produced daily in manufacturing setups, presenting new opportunities towards improving flexibility and quality of production processes. Such improvements can be achieved through the development of the so-called Smart manufacturing systems and, ultimately, Smart Factories.
Smart manufacturing is considered the future state of manufacturing, in which the produced data by sensors, embedded devices, and production systems are acquired, transmitted, and analyzed in real-time as required for a better understanding of the production processes and facilitating the knowledge-based decision-making in manufacturing companies. These systems aim at increasing efficiency, quality, and flexibility of production by exploiting and integrating advanced technologies and paradigms such as the Internet of Things, Cyber-Physical Production Systems, big data analytics, and machine learning.
Given the importance of big data and real-time analytics in smart manufacturing, this research work focuses on the development of a generic framework to guide the implementation of industrial big data management and analytics system (BDMAS) as part of smart manufacturing. To this end, a comprehensive list of goals and a set of functional and non-functional requirements are identified for the development of an industrial BDMAS based on empirical knowledge and a thorough literature review.
Subsequently, a conceptual framework for the Cyber-Physical Data Management and Analytics Systems (CP-DMAS) is proposed, along with a detailed description of its main components as well as the most prominent open-source technologies and tools available to develop the CP-DMAS. A microservice-based system architecture is also introduced for the development and implementation of the CP-DMAS prototype. The proposed framework, system architecture, and implementation strategies are validated through a set of practical experiments conducted by mimicking an industrial setting consisting of heterogeneous sensors, equipment, and manufacturing components.