DIGITAL TWINS IN www.reallygreatsite.

com

MANUFACTURING
Non Traditional Manufacturing Processes

Name ID Number Contribution

Arnav Trehan 2022ABPS1155P 33.33%

Kaarthik Nair 2022ABPS1331P 33.33%

Advait Nambiar 2022ABPS1093P 33.33%

Introduction
A digital twin is a dynamic, real-time virtual replica of a
physical asset, process, or system that continuously
updates with live data. By integrating AI, IoT sensors,
and simulations, it enables businesses to analyze
performance, predict failures, optimize operations, and
make smarter, data-driven decisions.

Origins
The concept of digital twins, virtual replicas of physical
assets, originated from NASA in 2010 to improve
spacecraft simulation, though the underlying principles
of computer simulation which existed earlier.

Purpose
Virtual Representation
Simulation and prediction
Optimization
Lifecycle Management
 Background State of the Art
Companies like Tesla analyze real-time data from
sensors to predict equipment failures, improve
workflow efficiency, and reduce downtime.
Singapore’s Virtual Singapore project helps
analyze traffic patterns, optimize energy
consumption, and improve emergency response
planning.
Philips’ heart models help surgeons simulate
complex procedures before performing actual
surgery, reducing risks and improving success
rates.
Other
15%

Manufacturing
40%
Smart Cities
20%

The image traces the evolution of Digital Twin technology from 1970 to 2018, showing
its growth through computer-driven simulations (1970s), applications like AutoCAD
(1980s-2000s), system design with Dr. Grieves’ concept (2000s-2015), and widespread
Healthcare
adoption by 2018, marked by GE’s wind farm and use by major companies. 25%
 Issues and Challenges
Data Integration: Legacy systems in manufacturing (e.g., pre-
2000 factory setups) clash with modern IoT protocols, causing
data silos and delays in twin deployment.

Interoperability: Disparate platforms (e.g., 01 Real-Time Processing: High data rates (e.g.,
Siemens MindSphere vs. AWS IoT TwinMaker) 0.5 GB/flight from Rolls-Royce Trent XWB
lack seamless integration, complicating multi- engines) lead to processing lag, delaying
vendor deployments. This often hinders 05 02 critical insights like failure predictions. A
single factory twin generates 1 TB/day,
scalability as well. Due to lack of collaborative Technical overwhelming storage and increasing
efforts the needs often exceed current
infrastructure capabilities Issues operational costs

04 03
Latency in Connectivity: 5G networks, while fast, Domain Modelling: Complex systems like jet engines
face latency issues in remote areas, hindering real- (30,000 parts) demand precise digital replicas;
time updates for rural smart city twins incomplete models can lead to false predictions,
risking safety.
 Issues and Challenges
Ethical Concerns: Human digital twins, such as organ models, raise ethical concerns regarding data
consent and privacy. There is a risk of misuse in insurance or employment decisions, potentially
leading to discrimination. Clear regulatory guidelines and transparent data governance are
essential to address these issues.

Lack of skilled labour: A shortage of skilled

01 High Costs: Setting up a factory digital twin
professionals is slowing the deployment of can exceed $5M, making it inaccessible to
digital twins, with 45% of companies citing
skill gaps as a major challenge
05 Non 02 SMEs. While large companies benefit from
efficiency gains, affordable, cloud-based
Technical solutions are needed to expand adoption.

Issues
04 03
Privacy Risk: Industrial digital twins are highly
Lack of Standardization: The absence of universal
vulnerable to cyberattacks, as demonstrated by a
standards leads to interoperability challenges
2023 ransomware incident that caused $2M in
between platforms like Siemens MindSphere and GE
operational losses
Predix.
 Architecture
1. Physical Layer (Bottom-most layer)
Contains physical devices and sensors (temperature, pressure, vibration
sensors, IoT-enabled devices, actuators, robots, and machines).
These sensors collect real-time data about the manufacturing process,
machine conditions, and environmental parameters.

2. Data Transfer & Collection Layer

Data collected from physical sensors and devices is transmitted to data

sources and controllers.
It uses communication protocols such as OPC UA (Open Platform
Communications Unified Architecture) to transfer industrial data securely.

3.Data Storage & Processing Layer

This layer includes Local Data Repositories for storing and processing raw
sensor data.
Data is structured and indexed for further analysis.
Example: The temperature data is stored in a local database for quick access.
 Architecture
4. Communication Layer
Acts as an intermediary between local storage and cloud platforms.
Uses IoT Gateway to transmit data to cloud systems.
SQL (Structured Query Language) is used for querying and retrieving data

5. Cloud Computing & Storage Layer

This layer consists of Cloud-Based Information Repositories that provide scalability
for data storage.
Cloud systems allow remote monitoring, AI-driven analytics, and large-scale
simulations.

6. Virtual Layer (Top-most layer)

This is the Digital Twin environment, where real-time data is used for simulation,
emulation, and predictive analytics.
It enables:
1. Predictive maintenance (anticipating machine failures).
2. Process optimization (identifying efficiency bottlenecks).
3. What-if scenario analysis (testing different operational strategies).
Example: The system predicts that the temperature may rise above safe levels in the
next 2 hours and suggests preventive action.
 Analytical Models in Digital Twins in Manufacturing
Analytical models in digital twins refer to the use of data-driven techniques to monitor, analyze, and optimize
manufacturing processes. These models help manufacturers interpret real-time and historical data to identify patterns,
detect inefficiencies, and make informed decisions.

The process begins with IoT-enabled sensors embedded in machines and tools collecting massive amounts of real-time
data, including machine speed, environmental conditions, and material usage.
Raw data is often incomplete or noisy, so the system first cleans, organizes, and structures it for analysis.
The analytical model then analyzes the cleaned data to identify trends and deviations from normal behavior.
The model applies predictive analytics to forecast future issues based on past trends.
Prescriptive analytics then suggests actions to fix or improve the system.

Example: Predictive Maintenance in an Aerospace Component Factory

An aerospace manufacturer uses analytical models in digital twins to monitor jet engine parts during production. Sensors
track temperature, stress, and vibration in metal components.
The analytical model detects small cracks or overheating before they cause failures.
Engineers receive early warnings, allowing them to replace parts before a breakdown occurs.
This reduces unplanned downtime by 30% and improves safety and reliability.

In short, analytical models enhance efficiency, reduce costs, and improve product quality without manual intervention. They help
manufacturers minimize downtime by detecting early machine failures, ensure quality by identifying defects in production,
optimize supply chains through demand forecasting, and lower costs by reducing waste and improving energy efficiency.
Case Study: Rolls Royce Jet Engines
Challenge: Managing a jet engine’s 30,000 parts to prevent unexpected failures.

Problem
Unplanned engine failures led to costly downtime,
grounding flights and disrupting airline schedules,
resulting in significant revenue losses.
High maintenance costs arose from reactive repairs
Inefficient engine performance increased fuel
consumption, raising operational costs and carbon
emissions

Overview:
Rolls-Royce developed a digital twin to monitor its jet
engines in real-time, focusing on performance tracking
and predictive maintenance.
The initiative aimed to optimize operations across its
global fleet, reducing downtime and improving overall
efficiency.
It leveraged advanced technologies to provide
actionable insights, enabling proactive decision-making
for airlines and maintenance teams.
Case Study: Rolls Royce Jet Engines
Implementation
Equipped engines with thousands of IoT sensors to collect real-time
data on temperature, pressure, vibration, and fuel consumption
during flights.
Developed a virtual model using 3D simulations and physics-based
modeling to replicate engine behavior under various operating
conditions.
Integrated sensor data into a cloud-based platform (e.g., Microsoft
Azure), enabling seamless data flow and storage.
Applied AI and machine learning algorithms to analyze data, predict
potential failures, and optimize engine performance.

Results
Enabled predictive maintenance, identifying issues before failures
occurred, significantly reducing unplanned downtime for airlines.
Optimized fuel efficiency by fine-tuning engine performance,
lowering operational costs for airlines. The image depicts a Rolls-Royce Trent 1000 jet engine
Reduced carbon emissions through improved fuel consumption, alongside its digital twin, a holographic 3D model
supporting Rolls-Royce’s sustainability goals. showcasing real-time data integration for predictive
Enhanced overall reliability, allowing airlines to operate more maintenance and performance optimization.
efficiently and maintain schedules.
Solutions to the Data Integration: Middleware platforms like MuleSoft help
Issues and 1 bridge legacy systems with modern IoT protocols, eliminating
data silos and ensuring seamless information flow.
Challenges -
Real-Time Processing: Edge computing reduces reliance on the
Technical Issues cloud, enabling faster decision-making—Rolls-Royce, for
2 example, cuts cloud load by 30% with this approach. Data
compression and prioritization algorithms help factory twins
handle 1 TB of data per day, reducing storage costs by 20%

Domain Modeling: AI-driven simulation tools like NVIDIA

3 Omniverse refine complex models automatically, enhancing the
accuracy of digital twins for applications like jet engine design.

Low-Latency Connectivity: Expanding satellite networks such

4 as Starlink ensures high-speed, low-latency connections in
remote areas, supporting real-time applications in smart city
twins.
Interoperability: Open-source frameworks like Eclipse enable
5 seamless integration across platforms, ensuring compatibility
between systems like Siemens MindSphere and AWS.
Solutions to the
Enhancing Data Privacy: Implementing blockchain technology
Issues and 1 can secure and encrypt sensitive health twin data (e.g., Philips’
Challenges - Non heart models), ensuring transparency and GDPR compliance.

Technical Issues Reducing Implementation Costs: Government subsidies and

cloud-based SaaS models can make digital twins more
2
accessible to SMEs, addressing the 60% currently priced out of
adoption.

Strengthening Cybersecurity: AI-driven cybersecurity tools like

Darktrace detect and neutralize threats in real time, preventing
3
costly attacks such as the $2M ransomware breach on a factory
twin.

Standardizing Interoperability: Establishing universal

4 standards through IEEE/ISO (e.g., IEEE P2806) can ensure
seamless integration between platforms like Siemens
MindSphere and GE Predix.

Launch industry-backed training programs and certifications

5 to upskill professionals, bridging skill gap and accelerating
deployment.
 Future Prospects of Digital Twins

6G Integration AI Ethics and Human Twins Adoption by SMEs

By 2030, 6G networks will reduce Ethical AI frameworks will govern Affordable SaaS models and
latency to microseconds, human twins (e.g., organ models), subsidies will drive 70% SME
enabling real-time twins for ensuring privacy and fairness in adoption by 2030, democratizing
global smart cities healthcare by 2035 Digital Twins

140

120

100
Market Size ($B)

80
Sustainability Focus Risk of Over-Reliance
60
Digital Twins will optimize energy
40 use, cutting emissions by 30% Over-dependence on twins may
in industries like aviation by lead to systemic failures—e.g., a
20
2030 2024 jet engine defect grounded
0
2026 2027 2028 2029 2030 flights due to prediction errors
Year
Market Projection: $125 billion by 2030, with 80% of
industries using twins
Capabilities of Digital Twins
Digital Twins are most effective where, ......
Systems where it is too Large or complex systems for Systems where predicting
1 expensive, risky or impossible to 2 which change is being 3 process variability is
do live tests considered important

Systems where decision

Systems where you have Systems where you need to
4 5 6 makers must be fast and
incomplete data communicate ideas
accurate

Conclusion
Digital Twins offer a powerful tool to mirror and manage complex
systems, as we’ve seen through their applications in industry,
healthcare, and urban planning. While technical and ethical
challenges persist, the solutions we’ve explored—such as advanced
data processing, secure frameworks, and standardized protocols—
provide a clear path forward. Looking ahead, their role in emerging
fields like education and global supply chains signals a broader
impact, though careful implementation will be key. In closing, Digital
Twins represent a significant step in our ability to understand and
shape the world
THANK
YOU!