An approach towards decentralization of systems communication in Industry 4.0

Abstract

Decentralization of system communication of Industry 4.0 is the main thesis, where different types of system type topologies (Centralized, Distributed, and Decentralized) are explored. With the different topologies, the importance of systems decentralization in the context of logistics is explored for a socially networked industry. A socially networked industry is a hierarchical network of systems that collaborate and execute tasks with optimizations performed at every network level. To develop such a networking paradigm, decentralization is necessary, unfortunately it is not possible to provide the same strategies for all communicating systems. Therefore, the heterogeneous systems in logistics are classified broadly into two systems types based on the available communication data bandwidth and energy constraints. Upon classifying the systems into low power, low data-rate systems and high power, high data-rate systems, networking architecture for decentralized communication are developed. Two decentralized networking concepts are developed as part of this research work called the Decentralized Brains and the DezCom Context Broker, which is deployed in the industrial research facility of FLW, TU Dortmund. A sensor floor with 345 low power wireless sensor nodes is used to develop and evaluate the performance of Decentralized Brains, and a multi-robot system along with servers deployed in a cloud environment was used for analyzing dezCom context broker. Networking flooding primitive and decentralized networking software for low power wireless sensor networks were developed, and the codebase for Decentralized Brains is published. A guide for developing decentralized industrial applications is discussed with the advantages and disadvantages of developing decentralized applications for specific industrial use cases.