Báo cáo

1. Introduction (giới thiệu chung)

1.1. Definition (định nghĩa)

A digital twin is a digital model of an intended or actual real-world physical product, system,
or process (a physical twin) that serves as the effectively indistinguishable digital counterpart
of it for practical purposes, such as simulation, integration, testing, monitoring, and
maintenance

A digital twin is set of adaptive models that emulate the behaviour of a physical system in a
virtual system getting real time data to update itself along its life cycle. The digital twin
replicates the physical system to predict failures and opportunities for changing, to prescribe
real time actions for optimizing and/or mitigating unexpected events observing and
evaluating the operating profile system.

DT is a technology that combines multiple fields. DT currently lack a unified definition,

which is still in constant development and evolution. The definitions of data twin in different
units and stages are listed as follows:
 2012 NASA defined DT:

DT is a comprehensive multi-physical, multi-scale, probabilistic simulation system

for vehicles or systems. It uses the best physical model to describe the historical use
of equipment to reflect the life of its corresponding physical equipment.

 2017 DT defined by the Defense Procurement University:

The integrated multiphysics, multiscale, probability simulation, using the best

available models, sensor information, and input data to mirror and predict the
life/activity/performance of the ircorresponding physical Twin, enabled by Digital
Thread

 2019 Stark Damerau defined DT:

A DT is a digital representation that contains the feature description of its selected

object or its product and service system, and obtains the attributes, conditions and
behaviors of the object through models, information and data in a single or even
multiple life cycle stages.

 2020 DT defined in a white paper published by the China Institute of Electronic

Technology Standardization:

DT refers to making full use of data such as physical model, sensor update, operation
history, and integrating multi-disciplinary, multi-physical, multi-scale, and multi-
probability simulation process to complete the mapping in the virtual space, thereby
reflecting the full life of the corresponding physical equipment Cycle process
 Figure 1: An early Digital Twin concept by Grieves and Vickers.

1.2. History (Lịch sử )

The digital twin concept gained recognition in 2002 after Challenge Advisory has hosted a
presentation for Michael Grieves in the University of Michigan on technology. The
presentation involved the development of a product lifecycle management center. It contained
all the elements familiar with the digital twin including; real space, virtual space and the
spreading of data and information flow between real and virtual space. While the terminology
may have changed over the years the concept of creating a digital and physical twin as one
entity has remained the same since its emergence. While its commonly thought to be
developed in 2002, digital twin technology itself has actually been a concept practiced since
the 1960s. NASA would use basic twinning ideas during this period for space programming.
They did this by creating physically duplicated systems at ground level to match the systems
in space. An example is when NASA developed a digital twin to assess and simulate
conditions on board Apollo 13.

After the launch of Apollo 13 on April 1970, no one could have predicted it would become a
fight for survival as the oxygen tanks exploded early into the mission. It became a famous
rescue mission as the world held its breath, with technical issues needing to be resolved from
up to 200,000 miles away. A key to the rescue mission, however, was that NASA had a digital
twin model of Apollo 13 on earth which allowed engineers to test possible solutions from
ground level. Of course, systems have now become predominantly virtual rather than
physical simulations. With the concept already being practiced for a few decades, the phrase
‘digital twin’ was first mentioned in 1998 and was being referred to a digital copy of actor
Alan Alda’s voice in Alan Alda meets Alan Alda 2.0. Although the digital twins have been
highly familiar since 2002, only as recently as 2017 has it become one of the top strategic
technology trends. The Internet of Things enabled digital twins to become cost-effective so
they could become as imperative to business as they are today.

1.3. Components of DT

The digital twin system comprises three main components: hardware, data management
middleware, and software. These Components forms the structure of the Digital Twin
ecosystem. Each component plays a distinct yet interconnected role, contributing to the
creation, operation, and enhancement of virtual replicas that mirror real-world systems. This
collaboration enables industries to unlock insights, efficiencies, and innovation in the pursuit
of operational excellence.
 Figure 2: Digital Twin Components

1.3.1. Hardware Components

Hardware components form the physical foundation of Digital Twins. These include
sensors, actuators, devices, and computing infrastructure that capture and process
real-world data, enabling the creation of a digital replica.

Imagine an industrial robot arm on a manufacturing assembly line. The robot arm is
equipped with various sensors, such as position sensors, temperature sensors, and
force sensors. These sensors gather data about the robot's movements, the forces it
applies, and the temperatures it operates in. These hardware components serve as the
eyes and ears of the Digital Twin, providing crucial data that reflects the real-world
behavior of the robot arm.

1.3.2. Middleware Component

Middleware acts as the bridge between the hardware and software components,
facilitating data communication, integration, and management. Middleware ensures
that data from hardware sensors is collected, processed, and made accessible to
software applications for analysis and simulation.

In an automotive manufacturing setup, there are multiple robots, conveyors, and

machines. Middleware solutions gather data from sensors attached to these devices
and standardize the format for software consumption. It aggregates data like speed,
position, and status, making it accessible to the Digital Twin's software component.
This middleware layer plays a vital role in maintaining data consistency and ensuring
that software applications receive accurate and relevant information.

1.3.3. Software Components

Software components are the brains of the Digital Twin. They use data from
hardware sensors and middleware to create virtual representations, run simulations,
and enable real-time analysis and decision-making.
 Consider an aircraft engine design process. Engineers create a Digital Twin of the
engine, incorporating design specifications, material properties, and real-world data
from hardware sensors placed on physical prototypes. Specialized software models
simulate how different parts of the engine interact under various conditions, such as
different altitudes and temperatures. These software components allow engineers to
visualize stress points, identify potential failures, and optimize the engine's design
before it's physically built.

1.4. Types of DT

Figure 3: Types of DT

1.4.1. Process Digital Twin

Process twins, the superior type of digital twins, combine system twins into a single
entity to investigate system synchronization and cooperation. This system provides
the most comprehensive picture of how things happen inside the factory or plant.
Therefore, it enables a far more in-depth and flexible output examination.

Process twinning provides data on results and allows for the adjustment of inputs,
such as raw material feeding rate and manufacturing climate, without interfering with
the manufacturing procedure or lowering the standard of production.

Executives can now quickly and risk-free test various business predictions instead of
depending on feelings or theory. Additionally, it helps to monitor key company KPIs
and facilitate data-based decision-making effectively.

For instance, a process twin could simulate every aspect of the plant, even down to
the workers running the machinery on the production line. In contrast, a system twin
might simulate a manufacturing process.

1.4.2. System/Unit Digital Twin

System twins are copies of assets that show how multiple components come together
to form functioning components at the system level. They provide an extensive
perspective of the plant or manufacturing facility.
 This enables the testing of various system settings for maximum efficacy or the
identification of new possibilities for the development of additional income streams.
The scope of the system twin includes a group of assets used in a specific function.
These are the manufacturing of a primary product inside the plant or the delivery of
energy.

At this level, continual tracking and digital twin modeling go beyond the primary
digital twins' ability to detect errors and breakdowns. It is a route to obtaining
information necessary for making tactical choices and achieving complete process
transparency.

1.4.3. Asset/Product Digital Twin

A product or asset digital twin is a digital replica of tangible assets such as

machinery, buildings, and automobiles. It offers updated information on an asset's
atmosphere, performance metrics, and functional health.

Typically, they are made up of many component twins or use the information
produced by component twins to model a complicated asset, such as an engine,
pump, or building. An asset twin examines how well various components work
together and function as an integrated unit.

By using asset/product digital twins, engineers can find areas for possible
development and gain insights into the operation of their equipment. This lowers
downtime for businesses and boosts operational effectiveness.

1.4.4. Component/Part Digital Twin

The most basic type of digital twin technology is known as the component twin or
parts twin. It is equivalent to the simplest components of the system, such as a
particular piece of machinery or a product, like a sensor, switch, valve, etc.
Such sophisticated parts may have their functioning monitored and current
circumstances simulated. This helps to assess their reliability, durability, and
effectiveness due to their digital model.

Meanwhile, a part type of digital twin just slightly resembles the factory architecture.
However, it allows for better monitoring of the equipment components and prompt
maintenance. This guarantees the reliability of that manufacturing operation and the
standard of the end product.

2. How to build a digital twin?

2.1. What is Azure Digital Twins?

Azure Digital Twins is a platform as a service (PaaS) offering that enables the creation of twin
graphs based on digital models of entire environments, which could be buildings, factories,
farms, energy networks, railways, stadiums, and even entire cities. These digital models can
 be used to gain insights that drive better products, optimized operations, reduced costs, and
breakthrough customer experiences.
Azure Digital Twins can be used to design a digital twin architecture that represents actual
IoT devices in a wider cloud solution, and which connects to IoT Hub device twins to send
and receive live data.

2.2. Key Features of Azure Digital Twins

2.2.1. Digital Modeling:

 Uses Digital Twins Definition Language (DTDL) to model physical environments,

such as buildings, factories, or cities.
 Defines entities, relationships, telemetry, and metadata, enabling complex system
representations.

2.2.2. Real-Time Data Integration:

 Ingests data from IoT devices, sensors, and other systems via Azure IoT Hub, Event
Hub, or Event Grid.
 Maintains live updates reflecting real-world changes.

2.2.3. Insights and Analytics:

 Query models using Azure Digital Twins Query Language (ADQL) to analyze
relationships and data patterns.
 Integrates with services like Azure Machine Learning, Power BI, and custom
applications for insights.

2.2.4. Simulation and Prediction:

 Run simulations on the digital twin to predict outcomes, optimize processes, or train
AI models without disrupting physical systems.

2.2.5. Scalability and Security:

 Supports large-scale digital ecosystems with robust security, compliance, and

integration options.

2.3. How It Works?

 Model Creation: Define the system components and their relationships using DTDL.
 Graph Construction: Instantiate digital twins and their relationships in Azure
Digital Twins.
 Data Integration: Connect IoT devices and systems to feed real-time telemetry.
 Analysis and Visualization: Use tools like Azure Maps, dashboards, or custom apps
for insights.
 Azure Digital Twins is a powerful tool for organizations looking to digitalize
complex systems, allowing them to gain insights, optimize processes, and innovate
with data-driven strategies. For more details, check out the
official Microsoft Azure Digital Twins documentation.

3. Technologies and Benefits of DT

3.1. Technologies associated to DT
3.2. Benefits of DT
4. Industries where DT can be majorly beneficial
4.1. Transportation
4.2. Healthcare
4.3. Retail
4.4. Smart cities
4.5. Manufacturing
5. Current Challenges and Limitations in DT
5.1. Challenges
5.2. Limitations
6. References