Executive Summary

Digital Twin: Manufacturing Excellence through Virtual Factory Replication A Whitepaper by Dr. Michael Grieves This paper introduces the concept of a “Digital Twin” as a virtual representation of what has been produced. Compare a Digital Twin to its engineering design to better understand what was produced versus what was designed, tightening the loop between design and execution. Page 1 of 7 Digital Twin White Paper Copyright © Michael W. Grieves, LLC 2014 Introduction The concept of a virtual, digital equivalent to a physical product or the Digital Twin was introduced in 2003 at my University of Michigan Executive C ourse on Product Lifecycle Management (PLM). At the time this concept was introduced, digital representations of actual physical products were relatively new and immature . In addition, the information being collected about the physical product as it was being produced was limited , manually collected, and mostly paper -based. In the decade that has followed, the information technology supporting both the development and maintenance of the virtual product and the design and manufacture of the physical product has exploded. Virtual products are rich representations of products that are virtually indistinguishable from their physical counterparts. The rise of Manufacturing Execution Systems on the factory floor has resulted in a wealth of data collected and maintained on the production and form of physical products. In addition, this collection has progressed from being manually collected and paper based to being digital and being collected by a wide variety of physical non-destructive sensing technologies, including sensors and gauges, Coordinate Measuring Machines, lasers, vision systems, and white light scanning. I introduced the term “Digital Twin” in Virtually Perfect: Driving Innovative and Lean Products through Product Lifecycle Management (pg. 133 ). I attributed it to John Vickers of NASA whom I work with. We have subsequently used this term in current projects. In light of these advances, it is timely to explore how the Digital Twin can move from a n interesting and potentially useful concept that aids in understanding the relationship between a physical product and its underlying information to a critical component of an enterprise-wide closed- loop product lifecycle. These tasks will both reduce costs and foster innovation in the manufacture of quality products. Digital Twin Concept Model The Digital Twin concept model is shown in Figure 1. It contains three main parts: a) physical products in Real Space, b) virtual products in Virtual Space, and c) the connections of data and information that ties the virtual and real products together. In the decade since this model was introduced, there have been tremendous increases in the amount, richness, and fidelity of information of both the physical and virtual products. On the virtual side, we have improved the amount of information we have available. We have added numerous behavioral characteristics so that we can not only visualize the product, but we can test it for performance capabilities. We have the ability to create lightweight versions of the virtual model. This means While the focus of this paper is on the manufacturing phase, the use of the Digital Twin extends throughout the product’s life to provide value to its user and information on how it actually performed to its manufacturer. This larger use is described in Virtually Perfect. Page 2 of 7 Digital Twin White Paper Copyright © Michael W. Grieves, LLC 2014 that we can select the geometry, characteristics, and attributes that we require without carrying around unnecessary details. This dramatically reduces the size of the models and allows for faster processing. These light-weight models allow today's simulation products to visualize and simulate complex systems and systems of systems, including their physical behaviors, in real-time and with acceptable compute costs. These lightweight models also mean that the time and cost of communicating them electronically is substantially less. They now can be shared not only with the organization but also throughout the supplier network. This enhances collaboration in both reducing time t o understand and enhancing both quality and depth of understanding of product information and changes. As importantly, we can simulate the manufacturing environment that creates the product, including most operations, both automated and manual, that constitute the manufacturing process. These operations include assembly, robotic welding, forming, milling, and other manufacturing floor operations. On the physical side, we now collect more and more information about the characteristics of the physical product. We can collect all types of physical measurements from automated quality control stations, such as Coordinate Measuring Machines (CMMs). We can collect the data from the machines that perform operations on the physical part to understand exactly what operations, at what speeds and forces, were applied. For example, we can collect the torque readings of every bolt that attaches a fuel pump to an engine in order to insure that each engine/fuel pump attachment is successfully performed. Extending Model Lifespans – A Matter of Unifying the Virtual and Real Worlds The amount and quality of information about the virtual and physical product have progressed rapidly in the last decade. The issue is that the two-way connection between real and virtual space has been lagging behind. Global manufacturers today either work with the physical product or with the virtual product. We have not developed the connection between the two products so that we can work with both of them simultaneously. The typical way we do this is to develop a fully annotated 3- D model. We then develop a manufacturing process that will realize this model with a Bill of Process (BOP) and Manufacturing Bill of Materials (MBOM). The more sophisticated and advanced manufacturers then simulate the production process digitally. Page 3 of 7 Digital Twin White Paper Copyright © Michael W. Grieves, LLC 2014 However, at that stage, we then simply turn over the BOP and MBOM to manufacturing and leave the virtual models behind. In many cases currently, we even dramatically water down the usefulness of the model by producing 2- D blueprints for the factory floor. There are manufacturers who are bringing

terminals stationed in the work cells. However, even here there is not real integration and connection between the virtual model and the physical product taking shape on the factory floor. The terminal model merely serves as a reference, and a human has to perform the connection between the virtual and the physical product on an ad hoc basis. As shown in Figure 2, linking the physical product with the virtual product could take the form of the 3- D model not only appearing on the screen but also incorporating actual dimensions from the physical product. The information of the physical product would overlay the virtual product and highlight differences that would need to be addressed. This simultaneous view and comparison of the physical and virtual product will reap major benefits, especially in the manufacturing phase of the product. Digital Twin Fulfillment Requirements In order to deliver the substantial benefits to be gained from this linkage between virtual and physical products, one solution is to have a Unified Repository (UR) that will link the two products together. Both virtual development tools and physical collection tools would populate the Unified R epository. This would enable two-way connection between the virtual and physical product. On the virtual tool side, design and engineering would identify characteristi cs, such as dimensions, tolerances, torque requirements, hardness measurements, etc., and place a unique tag in the virtual model that would serve as a data placeholder for the actual physical product. Included in the tag would be the as-designed characteristic parameter. When the design was released for production, these tags would be collected from the virtual product model and used to create the UR. A lightweight model with the tags and their characteristics and geometrical location would also be created. On the physical side, these tags would be incorporated into the MES in the Bill of Process creation at the process step where they will be captured. As the processes were completed on the factory floor, the MES would output the captured characteristic to the UR. The final step would be to incorporate this back into the factory simulation. This would turn the factory simulation into a factory replication application. Instead of simulating what should be happening in the factory, the application would be