Digital twins have become increasingly popular as Industry 4.0 and IoT continue to develop. When combined with Huawei’s Industrial Intelligent Twins and big data analytics, digital twins provide a full chain of intelligent services, covering design, production, logistics, sales, and services. They can also help enterprises analyze and extract value from data to build leadership in the industry.
When a digital twin is created for an industrial robot, it accurately displays the angle, speed, and acceleration of robotic arms in real time. Based on data about the robot's previous and current state, the digital twin can predict if and when a fault will occur, making s O&M more efficient and reducing unexpected downtime and labor costs for onsite services.
The idea of a digital twin was first proposed by Professor Michael Grieves from the University of Michigan. He described it as "a virtual presentation of a physical product" in the product lifecycle management (PLM) concept. As IT becomes more advanced, the meaning of digital twin is also evolving. According to NASA, a digital twin is about using data, such as physical models, real-time status, and running history, to reproduce, map, and synchronize the shapes, attributes, behaviors, and rules from the physical world in a virtual world through simulation. The simulation is based on established experience and knowledge, so as to reflect the whole lifecycle process of the physical world.
With a digital twin, we can digitalize entire processes from product design and production planning to manufacturing execution, making product innovation and manufacturing more efficient. Imagine building a complete digital twin for an entire city. By monitoring a whole urban environment in real time with the digital twin model, we could better utilize city resources and enable intelligent traffic scheduling.
There are several key technologies behind digital twins:
How do you link the physical world with the virtual world? By modeling the physical world. That means you need to digitalize the physical world with models and map the physical world to a virtual world.
When a product is put into actual production and running, its status changes along with changes to materials, processes, personnel, or environment parameters. These changes happen in the digital space in real time through synchronization through digital twins.
Machine learning and AI analyze the large amounts of data generated by sensors installed in the physical world in real time. This allows the digital twin to detect exceptions and perform predictive maintenance based on established experience and knowledge, so that enterprises can make more accurate predictions and informed decisions.
Unified modeling languages like DTML can support unified digital twin modeling and open ecosystem building, contributing to an interconnected digital world. The combination of digital twins with IoT devices and business systems can be scalable, enabling enterprises to track the past and predict the future, perform real-time predictions, and leverage intelligent data analytics and AI services.
Digital twins can be great, but they can be challenging to implement. First, the physical objects that need to be modeled are usually complex systems. For example, one production line in a steel plant can have more than 6,000 measurement points, and the relationships and interactions between those points are complex.
Second, digital twins require powerful real-time computing capabilities. Physical objects are constantly generating data and real-time mapping between digital twins and physical objects is very computing-intensive. Measurement points on a production line report data every few milliseconds and thousands of service indicators need to be calculated in real time, so that raises the bar for the model's real-time computing resources.
HUAWEI CLOUD IoTA centers on the digital twin model and is deeply integrated with asset models through the standard modeling language DTML. During data analytics, developers can easily reference IoT model data to get their job done more efficiently. Numerous algorithm models, which are pre-integrated in HUAWEI CLOUD Industrial Intelligent Twins and draw on the best practices of Huawei and the industry, are available to help users greatly improve data analytics efficiency.
Devices in the real world are not discrete. They are connected through complex relationships, such as space, organization, person, and context. IoT companies need to connect the physical world to the digital world, better understand their devices, and quickly and efficiently analyze data. These have become basic services that IoT services desperately need.
HUAWEI CLOUD IoTA provides a set of open and scalable advanced modeling languages (DTML), which contains basic words for things like object attributes and relationships between private events. With this, users can create their own vocabulary for their own industries to accurately define the digital forms of their complex physical objects.
For buildings, as an example, we can build complex relationships between things and other things, between things and space, and between things and people. Then we interpret the data generated from these connections in the context of the digital model. Finally, we can use an IoT+ asset model to build a digital twin that is synchronized with the physical world in real time. This way, a unified and consistent data foundation can be provided for data analytics through model abstraction.
HUAWEI CLOUD IoTA provides powerful tools and capabilities to build digital and intelligent infrastructure.
HUAWEI CLOUD IoTA uses "What You See Is What You Get (WYSIWYG)" graphical modeling to simplify the development of complex digital twins. A tree structure is used to describe the internal relationships of complex physical objects, including spatial relationships, combination relationships, and upstream and downstream relationships. Virtual measurement points support multiple calculation operators, including basic arithmetic, scientific notation, triangle functions, sliding windows, and stream computing. Asset model templates can also be defined and quickly copied.
HUAWEI CLOUD IoTA's high-performance model engine supports high concurrent and real-time computing capabilities that can easily complete 100,000 concurrent computing tasks to ensure that digital twins can map optimal statuses in real time.
HUAWEI CLOUD IoTA supports spatio-temporal convergence and analysis, enabling users to analyze data based on a digital twin across multiple time and spatial dimensions. It has also embedded an interface for interconnection with AI. Developers can deploy AI with just a click, imbuing digital twins with intelligent inference and new capabilities.
With digital twins at its core, HUAWEI CLOUD IoTA, can build a digital model for warehouses and act as a one-stop platform for data collection, cleaning, storage, and analysis capabilities.
When an asset enters or leaves a warehouse, an RFID reader automatically scans its RFID tag information and reports it to the cloud IoT platform. The cloud performs further data analytics based on the digital twin model to determine whether the asset is inbound or outbound. The entire process goes as follows:
(1) Check whether the assets have entered and left correctly based on inbound and outbound orders, and provide the onsite service dashboard that gives onsite operators real-time visibility on the check results.
(2) Collect information of asset inbound and outbound, and update the statuses of assets immediately after they are taken into or out of the warehouse. These capabilities support transparent management of inbound and outbound assets.
Making sense of the massive amounts of data generated by both the physical world and the Internet is a crucial step towards building a digital world that supports real-time interaction. With HUAWEI CLOUD IoT, users can tap into their IoT data and digital twins to push the envelope of what's possible and stay ahead of the digital curve.