C4IR Conference Series 2019

1. Digital Twins
What are they and why should we care?
5th HVM New Materials Conference Summit 2019
Cambridge, UK 6-7 November 2019
www.cir-strategy.com/events

2. Our (working) defini7on
A digital twin is a dynamic virtual representation of a physical object or system, usually
across multiple stages of its lifecycle. It uses real-world data, simulation and / or
machine learning models, combined with data analysis to enable understanding,
learning and reasoning. Digital twins can be used to answer what-if questions and
should be able to present the insights in an intuitive way.

3. It is likely that you will end up with mul7ple, federated, digital twins –
that will need to share data and possibly be integrated in real 7me.
Product design
OperaNon & Maintenance
Reliability &
resilience
IT Systems
Manufacturing
process
Training
Business
Processes OrganisaNon
Classes of problems that DTs can address
• Design
• EvaluaNng alternaNves
• TesNng designs
• Visualising
• Planning
• Comparing different plans / heurisNcs used
for planning
• ConNngency, resiliency planning
• Reliability engineering
• OperaNng policy evaluaNon
• Training
• Real Nme decision making
• Decommissioning

4. Simulations are, by necessity, bounded and an approximation
Digital Twins need to make a number of trade-offs…
Depth
How detailed and accurate
are the results?
Breadth
How generic / specific is it?
Complexity
How expensive will It be?

5. c
Digital Twins model data from the real world to enable us to take better decisions that
impact the real world
Data Security and Privacy, Governance and Management etc.
Digital Twin Real World
Data
Decisions
Business NOT IT
driven

6. c
c
Digital
Model
(now-cast)
Situational
Awareness
Predictive
Engine
(forecast)
Visualisation
Real-world
sensor data Physical
Model data
Visualisation
and Interaction
High level component view of a Digital Twin
Modelling output
Estimated states
Data Security and Privacy, Governance and Management etc.
Actionable Insights / Alerts
Data Analysis
Digital Twin
System
Component
Sensors
Real World
Places
People

7. Digital Twin Enterprise
IntegraNon
Data
Systems of Record
Simulation Modelling
Visualisation
Process management
IoT
Analytics and AI
Governance
Security
Integration
Enterprise business processes
Enterprise
IT

8. IoT Data SimulaNon
Modelling
Visuali-
saNon
AnalyNcs
AI
Process
mangmnt
SoR
Governance Security IntegraNon
Sensors
Schemas
Real World data
AggregaNon
IoT Pla[orm Edge / Fog Comms
Standards
Cleansing
Archiving
Eng drawings
BOM
Asset Structures
Asset state
current & historic
SimulaNon
(Test Data)
Manufacturing
Weather
External Sources
e.g. supply chain
Environmental
Types – see
comments
Depend on
DT domain
ERP CAD /Eng EAM PLM GIS CMDB Requirements
MulN-dimension
predicNve models
StochasNc SimulaNon
Deep ML
DeterminisNc simulaNon PredicNon of
unseen states
PredicNve
maintenance
Math. representaNon
FuncNon
simulaNon
Data lake
SimulaNon
couplers
Human
readable /
interpretable
StaNsNcal
representaNon
DescripNve
DiagnosNc
PredicNve
OpNmisaNon
Pabern
recogniNon
MIS
Feature
ExtracNon
Event pabern
idenNficaNon
Abstract
representaNon
Geometric
Physical
re-imagining
3D in 2D AR
VR
Real Nme
dashboard
SimulaNon
In-situ
Data eye-
balling
Use cases: Repair,
afer market …
AlerNng
Business
process models
Workflow
SimulaNon
“tool-chain”
Resource
allocaNon
Rules of engagement,
CBA , prioriNsaNon
Data
lifecycle
Who can
see what
Master data
management
SimulaNon
in- and
output
version
control
Data
catalogue
Model
governance
Interface
Perimeter
(IoT sensor
data)
IT / OT
Data
Security
and Privacy
(GDPR)
Ethics
Intrusion
detecNon
Data
ApplicaNon
Process
OT / IT
Protocol
ESB
API
Management
ECM
IdenNty and
Access
Mangement
End point
mngmnt
People
Real
World
Places Devices
EncrypNon
Network
Gateways
Device
IdenNty
Assets Context Processes OrganisaNons
4D
Discrete event Process Monte Carlo
Digital Twin Reference Architecture

9. hbps://www.nic.org.uk/wp-content/uploads/Data-for-the-Public-Good-NIC-Report.pdf hbps://www.cdbb.cam.ac.uk/Resources/ResoucePublicaNons/TheGeminiPrinciples.pdf
National Infrastructure Commission: National Digital Twin or Brit-Twin

10. Examples
Aston MarNn
Red Bull Racing
N-1 Resilience
modelling

11. Data
Systems of Record
SimulaNon Modelling
VisualisaNon
Process management
IoT
Digital Twin Reference Architecture
AnalyNcs and AI
Governance
Security
IntegraNon
Real World