1. Siemens Global university challenge 2016:
enabling the digital twin

2. Clustered Digital Twin (CDT) Infrastructure
Idea Category: Business Idea
Idea Title:
• PROPOSING CLUSTERD DIGITAL TWIN INFRASTRUCTURE for DESIGNING & USING DIGITAL
TWINS IN MANUFACTURING ENVIRONMENT
• IDENTIFY ADDITIONAL BUSINESS DRIVERS FOR USING DIGITAL TWINS USING CLUSTERED
DIGITAL TWIN INFRASTRUCTURE
Team:-
Avinash Neema
Avinash Chandra Misra
Business School

3. CDT: Idea Preview
Proposal:
• Identify additional business drivers required for using the digital twin by designing and
analysing CLUSTERED DIGITAL TWIN (CDT) platform software.
Idea Preview:-
• Analysis of Data generated by Digital twins involves lots of computation and computing
resources. At the same time, the target manufacturing companies that might be willing to
use digital twins also want to get the maximum return out of their investment (ROI) in
digital twin technology.
• Hence, it’s an area to analyse means of efficient usage of the information provided by
actual product and optimize the computing resources used by digital twins.
• A scale out software infrastructure provides a robust platform to design, monitor, and
integrate Digital Twins in a plant as well as across the plants.
• On-the-fly Product maintenance is one of the biggest area to investigate in energy industry.
A CDT like model can provide needed support to address plug-and-play feature in
manufacturing plants.

4. CDT:Additional business drivers to Analyse for
application of “Digital Twin” using Clustered
Digital Twin Infrastructure
1. Bidirectional Communication using Digital Twins:
Digital Twins built using CDT infrastructure can work more intelligently not only
getting data from product using sensors but also monitoring and controlling actual
products.
2. Computing Resources Optimization for Digital Twins
Reduce Digital Twin operational cost
3. Enhance deterministic behaviour of Digital Twins by “Task Prioritization” to ensure
resource allocation to most important digital twin, decision making, and runtime event
handling, providing better fault management.
4. Better software algorithms to enhance the life cycle of local computing resources (such
as FLASH memory in local computers installed in a plants, as FLASH memory is costly
and has limited write operations).
5. On-the-fly addition/maintenance of components in actual products as well as addition
of Digital Twins in the Cluster seamlessly.
6. Online health monitoring of products through Digital Twins:
Periodic health monitoring (scheduled by user).
On-demand diagnostic (basic routine if device is not in active mode).
Offline diagnostic routines (component taken offline and tests are performed, and
component is fixed back after maintenance).

5. CDT: Centralized Clustered digital twin platform
software. Advantages..
• Centralized Clustered Digital Twin platform software provides:-
• Resource optimization for Designing Digital Twins.
• Software flexibility to optimize digital twins operational efficiency.
• Load balancing among various end products such as wind turbine through
bidirectional communication through Digital Twins of those products.
• Centralized product health monitoring and diagnostic system. A
malfunctioned component can be taken offline, and offline tests are
performed by repair team. And once component is ready for use, it can be
plugged into the digital twin software.
• Categorization of Digital Twins in product development phase:
• Usually when a product line is released, not all of the customers will be using the
devices (products) in the same way. We would like to categorize those digital twins and
act on the priority based on the application of products.
• Provisioning of digital twins based on applications (devices at user end,
devices for testing, devices for maintenance and repair).

6. CDT: Future Scope & Possibilities
Scope & possibilities:
• Cluster Digital Twin (CDT) infrastructure design can be used for
creating a scale out platform for manufacturing firms that have
countable products such as wind turbines in wind farms, engines in
automobile industries etc.
• If a connected infrastructure is in place, scope is unlimited as lot of
data can be prioritized, resources can be shared, load balancing could
be performed and if devised properly plug and play of physical
components on the fly is feasible.

7. CDT: SWOT and Positioning analysis Vs
Competitors(GE’s Predix)
A
B C
What CDT
offers
What
Competitors
offer
What
Customers
Want
A Strength:
CDT offers Innovative digital twin design feature
On-the-fly Digital twin maintenance support
Runtime event handling. Plug and play of digital
twin as well as device support.
Bidirectional communication (Digital Twin <->
product)
Software infrastructure support for online,
offline and scheduled diagnostic monitoring.
Software design for trying Plug-and-play of
components (after maintenance) because CDT
enabled Twins are programs that work on
network programming concepts of Server-client.
Cost effective software design, better ROI
Local cluster designing support for fast event
handling fro digital twins.
Digital Twin computing resource optimization,
D

8. CDT: SWOT and Positioning analysis Vs
Competitors.. continued
A
B C
What CDT
offers
What
Competitors
offer
What
Customers
Want
B Core common services
Cloud services and Integration
Software as a service,
Digital twin design and monitoring
C Industrial IoT support to create embedded
devices.
Huge cost to use GE’s Predix platform
D Opportunities
Augmented reality support,
Additional employee safety features
Device plug and play support on the fly
D

9. CDT: Big Picture

10. CDT: Big picture WITHOUT Clustered Digital Twin
Design
Twin name Twin1 Twin2 Twin3 Twin4
CPU resource CPU1 CPU2 CPU3 CPU3
Memory Allocated 10GB 10GB 10GB 10GB
Twin Work Load 125% 20% 25% 75%
Digital Twins
with dedicated
resources
Wind farm # 1 Wind farm # 2
Real
Devices

11. CDT: Big Picture WITH Clustered Digital Twin
(CDT)Design
Core 1
M1
Digital Twins
Connected
over CDT
platform
Wind farm # 1 Wind farm # 2
IP1 IP2 IP3 IP4 DIGITAL TWINS ARE CONNETED OVER NETWORK LAYER
CLUSTERED DIGITAL TWIN SOFTWARE INFRASTRUCTERE
CLI- command line
interface
HARDWARE RESOURCES
Resource Scheduler Processor Memory Provisioning
Software Kernel Core1 Core2 Core3 core4 M1 M2 M3 M4
Core 3
M3
Core 2
M2
Core 4
M4
Real
Devices
Dynamic
Core and
memory
Allocation
based on
Hotness of
devices

12. CDT: Load Balancing in Customer Environment
Core 1
M1
Digital Twins
Connected
over CDT
platform
Wind farm # 1 Wind farm # 2
IP1 IP2 IP3 IP4 DIGITAL TWINS ARE CONNETED OVER NETWORK LAYER
CLUSTERED DIGITAL TWIN SOFTWARE INFRASTRUCTERE
CLI- command line
interface
HARDWARE RESOURCES
Resource Scheduler Processor Memory Provisioning
Core 3
M3
Core 2
M2
Core 4
M4
Real
Devices
Dynamic
Core and
memory
Allocation
based on
Hotness of
devices
Load Balancing
HOT COLD
Best Practices Sharing

13. MASTER
CONTROL UNIT
DIGITAL TWIN ACTUAL
PRODUCT
With sensors
installed.
CLUSTER DIGITAL TWIN PLATFORM
Alarms,
Commands Actions
CDT: Signaling event handling on actual products
using Twins developed on CDT platform
Talking to
sensor
network
Because Digital Twin are the computer programs
and models built using CDT infrastructure, they
can be used to trigger actions of sensors install on
actual products.

14. CDT: Resource Optimization and Cost saving
Software Features
• CDT platform enabled digital twins can track live Issues in Customer
Environment such as threshold levels of various component
operating efficiencies.
• Dynamic Resource Prioritization through CDT resource scheduler
according to the of the actual product (wind turbine in this case):-
• activity (Hot: maximum utilization, Cold: least utilization).
• Load Balancing triggering between actual devices using CDT resource
scheduler and the observed data through digital twin.

15. CDT: Seamless Maintenance of Digital Twin
CLUSTERED DIGITAL TWIN PLATFORM
Need maintenance
3 Create clone
4 Synchronize
TWIN1 data into
clone
5 Shut down Twin 1
for maintenance
6 Initiate
maintenance of
product
1
2
TWIN1
Clone of
TWIN1
Twin Fault
detection
Actual
product ma
cloud
4

16. CDT: Seamless Maintenance of Product
Component (Turbine blades) and load shifting
CLUSTERED DIGITAL TWIN
PLATFORM
3 Need maintenance
4 Alarm to operators for
fault detection
6 Load shifting to
another product
5 Initiate maintenance of
product
1
2
TWIN 1
Alarm
Actual
product
repair,
replacement
cloud
Blade
fault1
TWIN 2
Plug and
play
Substitute
Product

17. CDT: Sample health monitoring (diagnostic
operations) using CDT Infrastructure

18. CDT: Customer Value Addition
Value addition to customer:
Flexibility of operations, maintenance using digital twins clustered
infrastructure.
Prioritization of products maintenance post release, resulting in
better use of time and in cost savings of operations.
Resource optimization for digital twins.
More Dollar savings for installing Digital Twins for their products!
Better ROI in digital twins technology which results in saving more
dollars ($) per digital twin investment for a customer.
Scalable digital twin infrastructure.

19. Clustered Digital Twin :
Additional References

20. Drawing 1: relationship between Product,
Twin and CDT

21. Drawing 2: CPU and Memory Provisioning at Run
time using CDT

22. Drawing 3: Load Balancing and Resource
sharing among twins

23. Drawing 4: A Problem Scenario#2

24. Drawing 5: A Solution to Scenario#2

25. Drawing 6: Scenario#3 Seamless maintenance
of digital twin

26. Drawing 7: Scenario # 4 Seamless on the fly
product maintenance -PLUG and PLAY

27. Thank You !