Adaptive and Scalable Information Modeling to Enable Autonomous Decision Making for Real-Time Interoperable Factories

  • Adaptive und skalierbare Informationsmodellierung zur Ermöglichung autonomer Entscheidungsfindungsprozesse für interoperable, reaktive Fertigungen

Hoffmann, Max; Jeschke, Sabina (Thesis advisor); Vogel-Heuser, Birgit (Thesis advisor)

Aachen (2017)
Dissertation / PhD Thesis

Dissertation, RWTH Aachen University, 2017

Abstract

This work presents a framework that enables the transition from grown manufacturing and automation systems that have evolved over long-term production life cycles to an interoperable factory of the future. This paradigm shift from tightly-coupled automation systems to loosely-coupled flexible information and communication infrastructures in manufacturing is performed by making use of object-oriented interoperability standards. The proposed framework facilitates the development of information models for manufacturing environments with the goal to enable autonomous decision-making based on real-time information from the field.In terms of general demands for higher degrees of digitalization and the implementation of flexible manufacturing solutions within modern factories, major challenges to production planning and automation have to be faced in the near future. This work specifically focuses on the requirements of legacy and \textit{running} systems with regard to technologies that are introduced by umbrella terms such as Industry 4.0 or Cyber-Physical Systems. As proposed in this work, the shift from current factories to digitized manufacturing environments is performed in two steps -- the implementation of scalable information modeling and communication solutions into the factory environment and the establishment of autonomous systems.The implementation of communication and interoperability solutions is performed by an introduction of interface standards that support both tightly-coupled architectures of legacy systems as well as service-oriented approaches for flexible information exchange mechanisms for loosely-coupled system architectures. With regards to these requirements the OPC Unified Architecture meta-modeling standard is utilized as it fulfills the demanded compatibility with deterministic bus systems and enables the integration into flexible networking environments at the same time. By making use of object-oriented information modeling capabilities of OPC UA it is possible to carry out digital representations of arbitrary factory environments.This digital representation of a production site constitutes the starting point for the implementation of so-called smart agents into the factory environment. Using information modeling techniques, virtual representations of physical machines and devices are carried out by means of Cyber-Physical Systems. Smart software agents that coexist in both - the real and the digital - world are able to autonomously communicate, organize and optimize the production based on algorithms and dedicated software modules.The realization of these smart manufacturing solutions is evaluated by means of industrial use-case that reflect the requirements and automation demands of today's production sites. A software demonstrator as well as validation scenarios are carried out to illustrate the applicability of the proposed solutions to real-world industrial processes.