.webp)
On May 24, NAVER Team and Korea Water Resources Corporation began a full-scale collaboration to build a digital twin service based on a water management platform. Korea Water Resources Corporation announced that it plans to collaborate with NAVER Team based on its experience in developing Digital Garam+, a digital twin water management platform introduced at CES 2023.
Recently, there has been a lot of interest in digital twin projects, such as building cloud-based smart cities. In this article, we'll share our know-how and experience with the ultimate goal of building a successful digital twin platform.
Understanding the Digital Twin Business
Digital twin overview
A digital twin is a virtual representation of a machine, piece of equipment, or object in the real world on a computer. It is based on using real-time data to update it throughout its lifecycle. It refers to a virtual model of an object or system that aids in decision making through techniques such as simulation, machine learning, and reasoning.
Why you should care about the digital twin business
The global market for digital twins is expected to grow 38% annually, reaching $16 billion by 2023. One of the reasons the digital twin business is starting to gain traction is the growing number of ways it can be utilized. For example, here's why.
- Data is coming from a variety of sources and can be utilized in a variety of ways, even within the same industry.
- We can now look for value across the entire product lifecycle, not just product maintenance.
While there are different opinions as to why this technology has suddenly gained traction, there is no disagreement that AI has acted as a catalyst. With the advancement of the technology, the information gathered can be used in a digital twin environment to simulation, and then Because we're able to provide optimal results while teaching the AI to learn the results.
Big Data vs. Digital Twins
Digital twins encompass the concept of leveraging vast amounts of big data to create a visual representation of a simulation and connect it to the real world in real time. As a result, the user interface of a digital twin tends to be more immediate and user-friendly.
반면 Big data can be considered simpler than the digital twin business because it presents analytical results in two-dimensional tools such as tables, charts, and graphs. Simulating a digital twin doesn't end with simply making predictions from data.
Digital Twin Business Structure and Examples
Technologies that make up a digital twin
A digital twin is an aggregation of the technologies mentioned below. It is a combination of technologies that were originally used and are now available under the name of digital twin.
- Product Design and Product Data Management : CAD・PLM (Product Lifecycle Management)
- Product Simulation and Engineering : Computer Aided Engineering (CAE)
- Factory and manufacturing line simulation: 3D factory/plant simulation software
- Feed digital data back to the physical world as 3D information: Augmented Reality (AR) and Virtual Reality (VR)
- Feeding data from physical space into digital space: IoT and 3D scanning
Industry use cases
At each stage of design, construction, and maintenance, the construction industry relies on digital twin platforms for efficient process design, site safety, and productivity.
- Kagoshima Construction
Developed '3D K-Feild', a digital twin of a construction site, to remotely monitor construction sites
Displaying people, object, and car data acquired by IoT sensors installed in the field in a virtual space.
- Komatsu
'Smart contracts' build digital twins by acquiring terrain data with drones
Efficiency in the surveying process and automatic creation of processes
Dramatically improved surveying efficiency by completing a task that took over 4 days in 20 minutes
Visualize process progress to help leaders and decision makers get things done faster
City Planning - Smart Cities
In Singapore, Building Information Modeling (BIM) was used to digitally twin the entire land mass of the country. This is known as "Virtual Singapore," where you can see 3D models that incorporate terrain information and real-time data about buildings, transportation, water levels, people's locations, and more. This has allowed us to maintain each piece of infrastructure in the city and create efficient city planning.
NAVER has unveiled a digital twin solution 'ArcEye', which is capable of mapping large spaces with great precision. It made headlines when it was applied to NAVER's new building '1784', which was completed this year. In response to the increasing demand from companies and organizations that want to build digital twins of everyday spaces such as shopping malls and skyscrapers, the service boasts "core technology, specialized equipment, and cloud infrastructure". It has attracted the attention of many operators because it breaks down all the steps required to build a digital twin by function.
Recently, the team at Naver, through ArcEye, announced that Saudi Arabia is investing more than $500 billion to build the 'Neom City' to win the order.
How to implement a digital twin and practical tips
Implementation starts with data design
The vast amounts of data generated by manufacturing sites that operate around the clock are sometimes left unused or discarded. Therefore, based on big data-based process simulation, production planning optimization, digital twin platforms began to gain traction.
"Creating a digital twin for a typical sized organization requires millions of data points and relationships," said Harry Powell, Director of Industry Solutions at TigerGraph. To query this data, you need to understand the relationships between thousands of objects across dozens of links." In other words, in the digital twin business, where you're simulating the behavior of thousands of objects, you need a data model that can query objects and object relationships.
Simulation techniques are inherently time-consuming, depending on the complexity of the environment. However, when they are driven by learning modeling to reduce simulation time, the time is improved but the accuracy suffers. In fact, a study of simulation techniques using deep learning at Y University found that mixed-precision training was more accurate than traditional physics-based simulation methods faster, but the accuracy wasn't much different from what it was before. Eventually, to improve accuracy, we had to The conclusion was that it was important to make the training data diverse.
Data processing for real-world accuracy
Developing a simulation model that accurately mimics the real world is essential to utilizing digital twins. Poor simulation accuracy can lead to unreliable decision-making and have a critical impact on industrial environments.
In order to develop an accurate simulation model, we need to use Actual production data is essential. In other words, unlike big data, digital twins are a business where accuracy is critical during the implementation process. After implementation, the focus should be on achieving the desired end with the simulated data. In this process, Accuracy in data processing becomes more important.
Conclusion: Think about purpose and data design and implementation before embarking on a digital twin project.
A large part of the reason so many digital twin projects fail is because they try to do too much before they get started. You don't need to focus on fancy graphics, just getting the data right is all you need to get the most out of your digital twin.
While the purposes of digital twin projects vary, it is important to recognize common goals. It is important to clarify the purpose of the analytics data to be applied to the digital twin project and to process and implement it.
The goal of a digital twin project is to do more than just use big data, so it's important to plan carefully to make sure you're not just doing analysis and visualization with big data.
In addition, before proceeding with the digital twin project, it is important to collect high-quality data and process it to establish a virtual world that resembles reality. As explained earlier, real production data is required to develop an accurate simulation model, and the process of processing it into training data determines the accuracy of the model.
To summarize
- A digital twin is a model that is a virtual representation of the physical world.
- Digital twins have been used in a variety of industries, most notably in architecture and production, and more recently in urban planning and smart cities, where they are being actively researched and developed.
- In order to succeed in the digital twin business, it is important to pay attention to the accuracy of the simulation model. It is important to check the error between the real world and the virtual world, and to identify and process the purpose of the simulation data.
References
- https://blog.naver.com/chomdan_/222999019465
- Digital twins are all about data...don't go for the flashy images | The Korea Economic Daily
- Digital Twins - 4 Latest Examples & 6 Benefits
- https://www.samsungsds.com/kr/insights/digital_twin_trend_2.html
.svg)
