How to Create a Digital Twin Model
When we think of future technology, we often imagine AI-powered services, wacky new modes of transportation, and immersive digital simulations. While all of these technologies have great potential for growth and are attention-worthy in their own right, some technologies that are less prominent but just as important are often overlooked. Digital twins fall into this mold.
Digital Twins are already around today and are used in many countries, but have not achieved such a level of prominence and application yet that would make them familiar to most companies and consumers. Today, we are tackling this issue and giving you the rundown on digital twins, from what they are to how to create a digital twin model.
What is a Digital Twin and How Does It Work?
The terminology here is simple, and you can probably guess the meaning just by looking at the two words. A digital twin (DT) is a digital version or representation of something that typically exists in the physical world. The “something” in question can greatly vary in size, be it a microchip or a football field. You can think of it as a fully accurate 3D model with the same behavior and data as its real-life counterpart.
At the moment, DTs are mostly built with industrial or architectural applications, so they normally represent non-living objects, but it is possible to build them for living things like plants and animals, though success stories of this kind are likely far off in the future. Sometimes, digital twins are built for something that does not exist yet (but could) as a form of prototype or proving ground.
How digital twins work
After being constructed, DTs maintain the same appearance, features, and performance as their physical counterparts, which makes it convenient to do things with them that you would not rush to do with the physical version. For example, you can run a program with the DT inside and analyze how well the target object is working or test how making some small changes to a process will impact performance. Essentially, you are getting new and accurate data from the twin that would be much harder to collect with the physical version.
Benefits
One of the greatest benefits of DTs is their testing capabilities. With an accurate model, you can apply changes digitally to determine what you can expect if you do the same on a real model. With the right software, this can be done substantially faster than with a physical model, and without any of the risk.
Another key advantage is prototyping. If you build something digitally before you proceed with accurate construction, you have the time and opportunity to find the optimal design and become confident in its reliability. In fact, using 3D data to create digital twins is relatively easy in comparison to taking measurements of existing objects and models.
Many businesses choose to integrate DTs because they are helpful in improving their business processes, in ways such as remote troubleshooting of technical issues and combining with PLM, ERP, and EAM systems to boost production.
Types of Digital Twins
Though there is no official classification of DTs accepted by scholars, we can point out a useful grouping often used in manufacturing:
Component Twin
The CT revolves around a single part or component, which is usually part of a larger object or mechanism. For example, if a company is designing a new sprocket for their motor or checking an existing one, they may create a representation of the sprocket to get an accurate reading of how fast it can spin, how much structural pressure it can withstand, whether its form is optimized, and whether the material it is made of is suitable for its functional tasks.
Asset Twin
An AT is typically made for a single object with many internal components. For example, it could be something like a motor, a TV, a traffic light, or a fishing rod. It is not unusual for this model to be built with CTs in it, but the focus is no longer aimed at the small individual parts, but rather the whole object and how it works with the many components. This is likely the most common type of DT, considering the thousands of companies that build digital models of their products without ever realizing what they are classified as and how they can be expanded.
System Twin
As the name suggests, the system type encompasses a complete functioning system, with several assets that each have their own component. Thus, you have several objects or units functioning together in a single system, and the model analyzes how well the system as a whole functions, with the option to examine each asset and its components within the scope of the system. For example, an ST can represent a manufacturing conveyer belt, or a set of equipment in an automatic car wash, or even a whole building with several floors and utilities.
Process Twin
Any PT seeks to simulate and visualize an entire process. This process might include a multitude of systems working together or a single object carrying out a process — the scale varies dynamically. For example, when this type of solution is used in manufacturing, it can be based on data from every step of the process, from the delivery of raw materials to their transformation, molding, coloration, packaging, distribution, etc. They are mostly used to improve business processes on a large scale, and can be a major boost to PLM (product lifecycle management) systems.
How Digital Twins Are Made
The process of building DTs is complex and very difficult to master in a limited amount of time, but it can be broken down into 4 basic stages.
Step 1 — Planning
Before you build such a solution or get someone to do it for you, you should get a good understanding of what it should achieve. Will it just be a hyper-realistic showcase for your product, or an essential troubleshooting tool? Do you want to have the option to create theoretic scenarios with the tool, or should its main purpose be to provide up-to-date data readings from its physical counterpart? All of these questions should be considered during the planning stage, so that you come to a clear decision regarding the purpose of the DT, which data will go into it, and which level of accuracy you aim for.
Step 2 — Measurements
If you are building the DT as a prototype for a currently non-existent object, you will have full freedom to assign the parameters and performance you like to the virtual product. Otherwise, you will need accurate measurements and data that will go into the working computer model.
These measurements can include both simple procedures (like recording all of the proportions of an object) to more complex ones that require sensors. For example, movement sensors placed on different elements in an assembly line can record how a certain piece of equipment moves during manufacturing — its speed, force, position, and accuracy. Visual measurements like photographs and scans are also vital for the next stage of the process.
Step 3 — Generation
To build a dynamic and adaptive DT, it will need to be built in a software environment — some kind of application where it will later be examined and modified. The visual models of your target can be created through specialized modeling software, while the performance and functional aspects will require a vast sandbox.
For example, you can build an app from scratch that will house your solution, or use some of the ready-made software that already exists for such apps (e.g. Predix and Seebo). The data collected in Step 2 is inputted into the modeling/visualization/coding/twinning software and transformed into the virtual solution you have been anticipating and striving for.
Step 4 — Operation
Once you get your DT up and running, you will likely opt to make some immediate adjustments. Depending on how data-driven your solution is, you may be able to identify issues in a matter of hours and make changes to fix the situation. Of course, you may also find that the issues identified stem from the software and not a piece of equipment or object. In this case, tweaks, bug fixes, and code rewrites may be in order. After you have dealt with the urgent issues uncovered, you will be free to use the newly-built tool to your full advantage, testing, analyzing, and improving as you go along.
Examples of Innovative Digital Twins
While DTs may not have exactly caught on and gained the support of thousands of companies, there is already a solid collection of use cases that are truly inspiring. Below, you can read about several of them:
The City Twin
We have mentioned DTs the size of football fields, but some people think even bigger. The residents of Herrenberg in Germany have built a one-of-a-kind virtual model of their town using this innovative software. The virtual city includes features to monitor traffic flows, wind and emissions, as well as make informed choices about construction projects in this region.
In addition, there is a feedback feature allowing residents to share information relevant to other dwellers of the town. This approach has the potential to help city planners make meaningful changes to the locality and for residents to enjoy them.
Finding new cures
2 major companies (Atos and Siemens) are partnering to build a solution that could transform the pharmaceutical industry as we know it. They are working on a DT of the “Process” type which can be applied to develop new medicines and pharmaceutical products.
As you probably know, the development of drugs, vaccines, and medicinal compounds is a long process fraught with numerous challenges, but this new initiative seeks to speed up parts of the process. Their solution is driven by predictive modeling and data crunching, and designed to reduce waste in the industry, whether it refers to physical materials or fruitless experimentation.
Voyages made warmer
Marine vessels and ships have complex heating systems that use up a lot of their energy and fuel, and unfortunately, much of the heat goes to waste due to inefficiency. To tackle this issue, researchers from the Lappeenranta University of Technology have embarked on a project to improve vessel heat efficiency through DTs. They have built a real waste heat recovery unit and proceeded to create a DT of it. If all goes according to plan, this tool can be utilized by ship operators to make their vessels run more efficiently and stay warmer in cold sailing conditions.
Giving new life to old military hardware
The U.S. Army is working closely with Wichita State University to build a twin of their famed Black Hawk helicopter. This helicopter entered service more than 40 years ago and was eventually replaced by newer models, but the army still has a fleet of these helicopters that it wants to operate, and this requires proper maintenance. Thus, DTs of the helicopter and its hundreds of components will make it much easier for the Army to replace faulty parts and conduct repairs, thus prolonging the lifespan of the legendary aircraft.
Digital Twin Services
Given the complexity of digital twin creation, you will likely need several development partners if you opt to build anything of this kind. There are multiple businesses that specialize in building solutions for digital twins, and 3D-Ace is among them. Please allow us to tell a bit about ourselves.
Our studio has been around for over 2 decades, and this time has allowed us to gain bountiful experience and branch off in many directions. Our calling is creating 2D & 3D content, including game models, animations, visual effects, and industry solutions like walkthroughs and visualizations. In the context of digital twins, we are ready to create assets and visualizations that will make up the graphic part of your model, as well as integrate it into active systems in your business such as ERP, PLM, EAM, and others.
We find DT technology fascinating, and our specialists are always eager for new projects that will let them do what they do best — visualize. If you want to know more about our services and how we could cooperate, reach out to us at any convenient time.
Originally published at https://3d-ace.com on June 19, 2020.
