The increased digitization and tendency to automate complex manufacturing processes have dredged up technologies that date back to the 70s. Digital Twin is one of those disruptive technologies that went through multiple advancements and made colossal headway to be recently named as a strategic technology by Gartner. Manufacturers aspire to digital twinning and virtual commissioning
As of today, technology providers assert that a digital twin is an enabler of data-driven production, automation, and a critical component for the transition to Industry 4.0. Digital twin components affect all the key product lifecycle phases.
What is a Digital Twin within the Industrial IoT Paradigm?
A digital twin, in the majority of sources, is mainly explained as a digital representation or sometimes an avatar of any physical object or several objects, systems, processes, and even people. It is more than a digital model because it also replicates the dynamics of how the IoT-enabled device or a system works and operates. And it is even more than a field-proven 3D computer-aided design model (aka CAD).
In the heyday of Industrial IoT, Digital Twin bodes for data-driven manufacturing and decision-making. The latest online survey claims that 51% of companies apply IoT devices to collect data on their equipment health status. At the same time, 37% of all the respondents said that IoT devices help them enable predictive maintenance.
IoT sensors are in the DT core; the technology allows building predictive models to track the physical object’s behavior in real time. The DT core tech also understands how real-time behavior will perform under specific conditions.
This data allows improving the efficiency of manufacturing and decreasing downtime during an outage and continuously enhancing a product even after its release. For example, foreseeing some technical failures of a product that has already reached its customer is easier if the tech is coupled with the Digital Twins technology.
Tesla keeps a digital twin of each manufactured vehicle and based on the received data from multiple IoT sensors — sends necessary updates to a vehicle improving its performance.
Digital Twin vs. 3D CAD Model
3D rendering of a physical object is known as a CAD model helping designers to come with an accurate and comprehensive virtual design of a digital prototype. A digital prototype, unlike a physical one, can be easily altered.
For manufacturing, CAD modeling has been one of those cost-effective methods to construct a virtual model. 3D CAD brings about necessary geometry data and finds its way into systems configuration, simulations, and digital mock-up investigations.
For example, automakers use CAD models to optimize the time and materials needed for creating a single clay model of a vehicle. The automotive industry was one of those pioneering sectors (also military, aerospace, and engineering). The military, aerospace, and engineering started using CAD systems in the late 1960s-80s.
For automotive architects, design engineers, and stylists, CAD modeling has brought the desired flexibility in delivering a vehicle’s prototype. The CAD modeling allowed vehicles to come with plenty of iterations for modifying the vehicle’s ergonomics, anthropometrics, and speed up the design phase.
Digital Twin (DT) is an advanced form of the 3D CAD model; mostly serving Industry 4.0 and Smart Manufacturing purposes.
What makes DT different from CAD is the two-way communication with its physical counterpart. By utilizing real-time data on the behavior of a physical twin in the real world, there’s an opportunity to correct it through testing and modifying parameters in the virtual environment.
DT can substitute 3D CAD model reshaping virtual prototyping and providing manufacturers with a smart mockup of a new product they can test and enhance before investing in the construction of a costly physical prototype that will be most likely redone.
Another distinction from CAD is that Digital Twin can be applied to each stage of Product Lifecycle Management (PLM). With a pool of data visible to all the departments, including sales, marketing, logistics, supply chain, the digital twin can also be viewed as a data-driven tool for integrating cross-functional and geographically dispersed teams.
Digital Twin Types and Use Cases
Thriving smart manufacturing should hinge on three basic types of a digital twin embracing critical stages of a product lifecycle:
- The digital twin of a product simulates any physical object performance in various scenarios doing away with multiple prototypes and minimizing the total development time. Hence, manufacturers can add the required adjustments in the virtual environment to test and validate the product’s functionality, safety, and quality before it goes live.
- For example, Rolls Royce uses digital twins of fan blades to build their UltraFan jet engines to achieve 25% fuel consumption efficiency.
- Kuka robots are produced and maintained banking on their digital twins.
- As one of the biggest aerospace manufacturers, Boeing obtains digital twins of its airplanes resulting in a 40% increase in first-time quality and performance over the whole aircraft lifecycle.
- The UK’s Network Rail impresses with its digital twin of 16,270 km railway network created to automate the design process, significantly save costs and time on on-site manual measurements.
- Another even more unusual use case is a digital twin of a smart city covering architecture, engineering, and construction needs. A city is a dynamic object changing over time.
- Municipalities take a step forward by incorporating a digital twin technology into their urban development and management strategies.
- The trend has been gaining traction in the past few years. And as of today, there’s a city upgrading program Virtual Singapore with over $70 million invested in the development of a dynamic 3D city modeling platform.The digital twin of Singapore will allow visualizing how new smart buildings will fit the current capital city infrastructure as well as facilitating roads construction planning, installations, etc.
In terms of solar energy production, DT of a sunny Singapore can be used to analyze what buildings are the best options for installing solar panels and accumulating the most substantial amount of energy. The solar projections can also help to project the energy consumption peaks depending on the season.
- The second large city of Netherlands, Rotterdam also plans to utilize the city’s digital twin to improve infrastructure maintenance, energy efficiency, road, and water traffic, as well as to optimize work of firefighters during the emergencies, etc.
- However, the UK plans to upstage the city mentioned above projects by creating a national DT called “Brit-twin” that will be a comprehensive digital replica of entire Great Britain’s infrastructure.
The digital twin of production is virtual commissioning. Virtual commissioning promises to reshape manufacturing across different industries. A type of digital twin focuses on the digitization and full automation of a shop floor.
- Tronrud and Siemens primarily utilize digital twins in packaging equipment engineering. Since digital twin can fit any industry, virtual commissioning has been already used in the production of sawing machines (ABB and HewSaw).
- Motor assembly for robots (Siemens factory), oil and gas facility (Kongsberg).
- BP, a substantial British oil and gas company, has a digital twin called APEX, a simulation and surveillance system that helps engineers to decrease the optimization time of a production system to 20 minutes instead of hours.
The digital twin of performance gathers operational data of products, machines, and the entire production line to emulate and predict performance failures, energy consumption peaks, as well as downtime risks.
- Digital twin technology allows General Electric monitoring their wind turbines performance and overseeing their wind farms operations and productivity.
- Kaeser Compressors, a manufacturer of compressed air and vacuum products, uses digital twins of their compressors to predict when they may fail and thus, helps minimize downtime.
Although the number of real-world DT use cases is only growing, many manufacturers still play it safe as there are many challenges in overcoming conventional approaches to design and testing.
Perhaps, for this reason, A-STAR and ARTC (research and technology centers) created a testbed or a digital twin platform providing organizations with an ability to develop and test their DT PoCs for manufacturing equipment.
Reorganization of factories, shop floors, and training people to use digital twins at their workplace will be another significant challenge requiring sizeable investments. The good news is that this tech trend adoption is only ramping up. Especially after Gartner has recognized DT as one of the top technologies that will drive disruption in different industries over the next few years.
What is Virtual Commissioning and its Benefits
Digital twin technology has given a lift to virtual commissioning – a practice of building a 3D virtual replica of a real manufacturing environment. The appearance of virtual commissioning was influenced by the fact that the actual commissioning could take up to 20% of the entire project delivery.
Also, the increased level of automation and its complexity in manufacturing made it impossible to stick to the old models of pre-commissioning.
Simply put, virtual commissioning is a simulation model (a digital twin) of a factory/plant, a shop floor, or a small-size manufacturing cell or a machine. The fundamental purpose is to emulate and run the entire or partial production process virtually. It also tests vital functions and performance by applying technology before the launch of a production line.
This testing allows the detection and eliminating design flaws, bugs in the PLCs code, for example, and solving a range of technical, functional or performance issues in advance.
One survey showed that the time for real commissioning (real system and a controller) could be decreased by 75% due to a higher quality of the production system at the very start.
Virtual commissioning is highly recommended for automation control systems such as robots, PLCs (Programmable Logic Controllers), NC (Numerical Control) machines, VFDs (Variable-frequency Drives), and motors to decrease the risks of process downtime and changeover.
Consider other benefits of virtual commissioning:
- Accelerated on-site commissioning and production line set-up;
- Minimized risks of equipment failures, collisions, and downtime;
- Streamlining a conventional process of software testing – Factory Acceptance Testing and Site Acceptance Testing;
- Optimization of a production cycle of highly complex manufacturing lines;
- Reduction of prototype waste and higher savings on expensive materials;
- Early detected bugs in software incur fewer expenses.
These and other benefits are the main reasons why manufacturers aspire to implement virtual commissioning software. In the future, the technology will constitute full automation of all production processes to fit Industry 4.0 requirements.
The most significant players at the market who already offer virtual commissioning solutions and digital twin solutions are Siemens (Tecnomatix), Dassault Systemes (DELMIA), Visual Components Essentials, Machining (industrialPhysics), MapleSoft (MapleSim), and others.
Digital Thread vs. Digital Twin
Another common concept of PLM, as well as PDM (Product Data Management), is a Digital Thread which helps ensure an end-to-end approach to Industry 4.0-driven manufacturing. When manufacturers started racing towards the end-to-end automation, both technologies (Twin and Thread) came forward as its irreplaceable ingredients.
Digital Thread is a framework that allows integrating siloed elements, including all digital data of complex manufacturing processes. Plus, it suggests an interconnected communication of multiple participants involved at each stage.
Contrasting the terms, the Digital Thread is a digital representation of the whole manufacturing value chain – from the requirements and design stage to assembly.
Then there is aftermarket support. Digital Twin – is the digital representation of an asset. At the same time, Digital Thread allows improving Digital Twin traceability. The traceability is across all the lifecycle phases and as a result, brings the following benefits to manufacturers:
- Shortened development lifecycles and reduced costs.
- Continuous enhancement of a product through the whole value chain, its quality, and performance by minimizing defects and reducing downtime.
- Better transparency and efficiency of overall manufacturing processes and operations. This efficiency is due to the improved cross-discipline collaboration between multiple teams sharing the same digital data for decision-making.
- Accelerated manufacturing velocity, higher productivity, and flexibility.
- Most importantly, digital thread combined with PLM allows transforming the whole manufacturing delivery model focused on continuous improvement.
At the moment, Digital Thread, as well as the Digital Twin technology, are key innovative strategies. These processes are used by such companies as Siemens, General Electric, IBM, Microsoft, and Oracle.
In 2017, Aberdeen Group study reported that best-in-class firms already deployed PLM solutions. Relying on the digital thread technology in the product design (53% of firms), manufacturing processes (55% of firms), and the production stage (57% of firms).
Although Digital Thread was first used in Aerospace and Defence. In recent years, automotive, transportation, energy and utilities, and machine manufacturing have become the most significant industries (85% of global total).
These successes have influenced Digital Thread Market growth. The latest report says that the global market size will only increase by $100 million in 2019, reaching 49.6% CAGR in respect of revenue by 2024.
Key Technologies that Drive Digital Twin Evolvement
How did a digital twin and later a digital thread become such aspired technologies over the last few years? Digital twin definition was explained by Dr. Grieves in 2002 and referred to as “Conceptual Ideal of PLM” or “Information Mirroring Model.”
NASA and U.S. Air Force first utilized the technology. The Apollo 13 team rescue operation in 1970 also used this tech. During the process, NASA used a mirroring system (a preceding technology of a Digital Twin) to explore the open space and emulate the best way to return astronauts home.
At that time, the Digital Twin technology didn’t gain much support from manufacturers until industry leaders encountered the term in a Gartner’s report. The report described DT as “an underpinning technology of IoT.”
When the interest in the technology among tech trailblazers started growing, Digital Twin joined the list of Top 10 strategic technology trends in 2017 and 2019. Digital Twin solutions are working well, along with AI and Blockchain tech. As of today, 75% of surveyed organizations plan or already use DT technology.
What exactly was Digital Twin tech in 1970 and 2002? What did this tech lack then — but in 2019 can already be offered to manufacturers?
Internet of Things and Cloud Computing
The rise of another huge tech trend – Internet of Things — and its successful integration in our daily lives gave birth to using the Digital Twinning ideas. That idea had to gather dust for years (e.g., connected car and generally, the whole concept. Think: “connected to everything,” autonomous driving, intelligent assistants such as Alexa, etc.).
What prompted IoT going mainstream?
- Relatively reduced costs on sensors connected to the cloud;
- Increased cloud storage volume and data throughput capacity;
- Impressive growth in computing power;
- The improved mobile network bandwidth, speed, capacity, and decreased latency.
Internet of Things, in concert with Cloud Computing, functions as a “linchpin.” This is the connectivity between a real object and its digital avatar that enable real-time data collection and transmission. The data collection is handled by a DT providing assets (systems or devices) with the ability to “communicate.”
The communication then relays information about wellbeing and performance. Physical and digital twins can continuously interact due to the IoT sensors, mounted on a physical object and connected to the cloud. Real-time data collected from them is an input fed to a virtual counterpart which can be used for the PLM optimization.
Together with AI and data science, Digital Twin can empower Industrial IoT. But more importantly — it enables, automation of complex operations as the main priority of smart manufacturing.
In the Forrester report, there are already 15 recognized IIoT software providers (such as IBM, SAP, Amazon, Bosch, Siemens, Microsoft, Hitachi, and others). Each of these companies is focused on the integration of digital twin software and enterprise solutions with insightful data analytics.
Revealed by IDC, is that in 2019, manufacturers across disperse industries are expected to invest $197 billion in their IoT-driven operations, as well as Production Asset Management systems. What’s more staggering is that this year, the total IoT spending will be most likely around $745 billion globally as against $646 billion in 2018.
Big Data Analytics
To process massive volumes of raw data continuously received from sensors, Digital Twin should be integrated with Big Data analytics used for organizing, analyzing, and turning structured, semi-structured, or unstructured data into actionable insights.
Combined they can enhance the whole product lifecycle (from the product design to its maintenance, repair, and operations).
According to this study, the convergence of Big Data and Digital Twin promotes the establishment of smart manufacturing across different industries removing the “barriers between different PLM phases.” I this way it is decreasing the overall product development and especially, the verification time.
Empowered with Predictive Analytics models, digital twins also embrace essential metrics to enable predictive maintenance. The maintenance is based on the historical data on risk factors, failures, required maintenance, operations scenarios.
The predictive analytics also takes into account the machine configuration, and even energy consumption (remaining useful life, mean time to failure, and mean time between failures, end of life metrics).
What’s more, the digital avatar can use predictive analytics algorithms and received data to run behavior-focused simulations. These algorithms can be run in the virtual world to forecast how the device will “behave” under specific circumstances.
There will always be conditions that have to be watched, such as extreme operating conditions. Also, manufacturers can decide on the most pertinent maintenance scenario by previously running simulations with the digital twin and train employees how to apply them to the equipment in the virtual environment first.
Advancements in Augmented Reality integrated into the Digital Twin creation improved data visualization techniques in overlaying the image of a 3D twin model on its physical object. Moreover, augmented reality applications, as well as Digital Twin, can be applied to the entire value chain (e.g., AR in automotive R&D, manufacturing, supply, sales, and support).
These technologies are coupled to provide 360° view of a physical counterpart, understanding its dynamics in real-time, which allows boosting decision making. These decisions also include accelerating product development — its testing, and refinement and assisting in configuration guidance and predictive maintenance.
Another use case of an integrated DT with AR is a Virtual Twin, which allows bridging up a digital twin with its remote physical counterpart. Siemens introduced a PoC of a Virtual Twin, “a 3D model of a physical smart factory.”
By using Microsoft HoloLens technicians and engineers can have faster access to a Digital Twin of a smart factory and digital data to control and manage the shop floor remotely.
To summarize, in this article, we tried to embrace the basics of the digital twin technology looking closely at virtual commissioning and its benefits to manufacturers. We also described the essential difference between a digital thread and a digital twin, as well as some key technologies that drive digital twins.