Internet of Things News
Internet of Things Resources
July 1, 2016
With all the talk in the last few years about the Internet of Things (IoT) and the Industrial Internet of Things (IIoT), some concepts and notions regarding the use of 3D data in manufacturing are being revisited. One term from the previous decade getting a reboot is “digital thread.” What started as a way to describe using model-based definition (MBD) for delivering data to the factory floor has become a metaphor to unite physical and digital manufacturing processes in ways previously considered unrealistic or impossible.
The phrase digital thread was coined at Lockheed Martin to describe using 3D CAD data to directly drive CNC (computer numerically controlled) milling or composite programming systems for carbon fiber placement. In both cases, the physical piece is the result of an unbroken data link that stretches back to the original computer model of the part; the unbroken data path was the digital thread.
A consortium of aerospace and defense manufacturers gathered in 2011 to discuss the digital thread concept for the first time. The goal of their newly formed Computational Manufacturing Alliance (CMA) was to find common ground where both makers and users of the technology who needed to create a digital thread could work out data interchange issues. The goal was rooted in an interoperability agenda of plug-and-play connectivity where data producers and data consumers would be linked in a common data stream originating from a common data source. From that beginning, it has evolved into a vendor’s alliance, where software makers including Right Hemisphere, Delta Sigma, FARO, Kuka Integration and VRSI, work on tools to support a common workflow between engineering, process planning, manufacturing process and measurement.
Since then, the idea of the digital thread has expanded beyond a single model driving a single process, to a more global expression of the goals for a model-based enterprise (MBE) in manufacturing. Private industry, the government and academia are all working on expanding the scope of what full engineering-to-manufacturing connectivity can become. Along the way, other concepts have come to the front as the vision for a digital thread expands. At the top of the list are two concepts: make the digital thread a bi-directional data flow, and utilize a digital twin philosophy of data utilization.
The U.S. National Institute for Standards and Technology (NIST) is working with all stakeholders in advancing the digital thread philosophy. It defines the digital thread as “a way for different machines in a manufacturing process to all follow the same set of digital instructions.” This is slightly different than Lockheed Martin’s original definition, but a more holistic one. There is not just one digital thread from one CAD model to one machine in a factory; elements of the same model must connect to a variety of destinations, and there must be two-way communications.
Existing manufacturing processes use design feedback loops, but they are generally informal and not sufficiently automated (if at all). Initial designs specify fit and form requirements as geometric dimensions and tolerances, included in an MBD as the Product Manufacturing Information (PMI). But there is no formal means of sharing this design intent with part inspection, and certainly no two-way data flow. NIST is advancing the notion of “PMI 2.0” to drive a new standard of high-level information requirements and to close the gaps, connecting geometric, PMI and functional data in a two-way flow with manufacturing. NIST has started a GitHub repository where it is working with interested parties and hosting open software that addresses technology needed to create the envisioned digital thread.
From Digital Thread to Digital Twin
If the digital thread is a two-way line connecting engineering with manufacturing, there has to be a way for the data on the engineering side to update based on input from manufacturing. This is the foundation for the evolving concept of a digital twin. More than a CAD model, the visual and technical database of design intent becomes a digital mirror image of not only the imagined product, but the finished one as well. The digital twin holds analysis data, scan data from prototypes or finished product, as well as inputs from all corners of the enterprise where data is contributed manually.
CAD vendors are starting to describe their models as digital twins, but the exact definitions vary (see “Digital Twins Land a Role In Product Development”). What is common is the notion of two-way communications between the digital model, the machines that make the physical instantiation—and increasingly, the manufactured object itself, as internet-enabled products come to market.
Outside of the CAD vendor community, other stakeholders are pushing for consensus on what constitutes both the digital thread and digital twin. One leader in seeking a common definition and deployment of digital twin technology is the U.S. Air Force (USAF). In a research report approved for public release, USAF says it is advocating day-to-day use of a “tightly integrated digital thread and prototyping process to enable agile development.” Another approved release describes the combination of digital thread and digital twin as the “game-changer that provides the agility and tailorability needed for rapid development and deployment, while also reducing risk.”
In the USAF Global Horizons report on technology visions, digital twin is defined as “the creation and use of a digital surrogate of a material system to allow dynamic, real-time assessment of the system’s current and future capabilities.” The “digital surrogate” is a “physics-based technical description of the weapon system resulting from the generation, management and application of data, models and information from authoritative sources across the system’s lifecycle.”
The USAF technology vision paper also enumerates specific benefits in going to a digital thread/digital twin process:
• streamlined resolution of system performance issues;
• reduce late discovery of system performance deficiencies;
• identification of risk at decision points;
• quantification of risk at decision points;
• informed trade space exploration;
• yield and rate improvements via agility on the shop floor (adaptive machining, virtual assembly, etc.); and
• infrastructure to generate, capture, organize and use relevant data.
The Next Generation Buzzword
Having coined the term digital thread and having used the concept as a technology adoption mission slogan, Lockheed Martin is not resting on past efforts. It is now describing the next generation of digital connectivity. Led by Dennis Little, vice president of Space Systems, Lockheed Martin aims to combine three separate digital domains—virtual reality, 3D printing and digital processes—to “radically streamline its entire approach to creating complex products.” Lockheed Martin is calling this the Digital Tapestry and the company envisions it as the final goal for a fully digital product lifecycle.
“The Digital Tapestry ties everything in our production operations together digitally, from concept to product realization,” says Little. “It’s an end-to-end digital approach where everything is connected—from concept, design, simulation, manufacturing and assembly, to testing and getting the final product to the customer. Our goal is not to break a single thread in the digital process.”