BUILDING URBAN DIGITAL TWINS · 2020-06-04 · 15 INTEROPRABILITY OF INSPIRE DATA • INSPIRE: major achievement in uniform monitoring • Next step: Use semantic Web and linked data

1

BUILDINGURBANDIGITAL TWINSStefan Lefever (imec)Philippe Michiels (imec)Raf Buyle (Flanders Information Agency)

WHAT IS A DIGITAL TWIN?

A Survey on Digital Twin: Definitions, Characteristics, Applications, and Design Implicationshttps://ieeexplore.ieee.org/abstract/document/8901113

How to tell the difference between a model and a digital twinhttps://link.springer.com/article/10.1186/s40323-020-00147-4

An architecture of an Intelligent Digital Twin in a Cyber-Physical Production Systemhttps://www.degruyter.com/view/journals/auto/67/9/article-p762.xml

HISTORY

The “twin” concept originates from NASA’s Apollo program

Initially physical, later digital Later expanded to other

systems and environments

3

WHEN CAN WE CALL IT A DIGITAL TWIN?

A Model of the ObjectA Digital Twin must be a digital representation of a physical asset, including as realistic as

possible models and all available data on the physical asset.

An evolving set of data relating to the objectThe data must contain process data, acquired during operation as well as organizational

and technical information created about the asset.

A means of dynamically updating or adjusting the model in accordance with the dataIt must be possible, to run a simulation on the Digital Twin of the behavior of the physical

asset.A real-time link between the virtual object (the model) and its real-world counterpart can

be made.

4

ONE OF MANY DEFINITIONS

Note: a digital twin without a physical twin is ( just) a model

5

An integrated simulation of a real-life system that uses models, sensor information, and input data to mirror and

predict behavior of its corresponding physical twin.

WHY DO DIGITAL TWINS MATTER?

EXAMPLE

Variables such as tower high, blade length, average wind speed, etc. driver the yield of a windfarm

IoT feeds digital twins – who are hungry for real data

Collecting sufficient qualitative data in a way to use them in digital twins is hard

7

WIND FARM DESIGN & OPERATION

https://www.i-scoop.eu/internet-of-things-guide/industrial-internet-things-iiot-saving-costs-innovation/digital-twins/

MOBILITY & ENVIRONMENT PROTOTYPE

8

USE CASE TO UNDERSTAND DT DYNAMICS

INTRODUCING BUDI AND DUET

9

CITIES / PUBLIC BUILT ENVIRONMENT

PRIVATE BUILT ENVIRONMENT https://www.digitalurbantwins.com/

Build Environment for Digital Twins

DT OPPORTUNITIES

Digital twins allow to better monitor and understand the interactionshappening between different aspects of complex systems and provide insightsin the drivers of the behavioral properties of these systems.

INSPIRE: Monitor the NOW DUET: Understand and predict the tomorrow

10

11

INSPIRE

Ecosystem for Environmental Data for supporting sustainable development Breaking the boundaries between member states

Cross-domain ecosystem for Urban Data and Models also supporting sustainabledevelomentBreaking the boundaries between

member states, regions, municipalitiesand policy domains

https://en.wikipedia.org/wiki/Infrastructure_for_Spatial_Information_in_the_European_Community

DUET

THE CHALLENGES WE FACE

13

DATA DATA DATA

THE CHALLENGES IN BUILDING DIGITAL TWINS

AVAILABILITY QUALITY INTEROP

• Insufficiënt adequate data sources

• Real-time• Reliablly captured• Available

• Not everything can be easilymeasured

• How do we assess anddocument data quality?

• Can we deal with sensor drift and other deviations?

• How to callibrate?

• How can we ensureinteroperability s.t. different data sets can be combined?

• Can we prevent ‘messy’ integrations?

• Can we combine data sets and discover relations?

14

Models, Algorithms, AI

Open data, context data and data standards

THE INTERCONNECTED INTERNET OF THINGS

INSPIRE & THE SEMANTIC WEB

15

INTEROPRABILITY OF INSPIRE DATA

• INSPIRE: major achievement in uniform monitoring• Next step: Use semantic Web and linked data principles to unlock the

full potential of the data

• Digital twins face a huge integration challenge• Open linked data sources can soften the burden significantly

https://www.sciencedirect.com/science/article/abs/pii/S1570826815000852

THE RIVER SCHELDT

Flows across France, Belgium, Netherlands Flooding case: water quantity Environment:

Water quality Air quality

16

INTEROPERABILITY

They both measure air quality… and expose them trough an open APIBut when trying to exchange results, they are in for a surprise …

17

TODAY’S APIs

Meet Raf & Philippe…

DIFFERENT LEVELS OF

INTEROPERABILITY

TODAY’S APIs

Raf’s results

18

DIFFERENT LEVELS OF INTEROPERABILITY

Philippe’s results{

"location": {"x": 4.4008622,"y": 51.2248885

},"time": "2019-05-20T20:00:00.000",

"measurements": [{

"property": "temp","value": 19

},{

"property": "humidity","value": 79

},{

"property": "PM10","value": 40

}]

}

<?xml version="1.0"?><Measurements time="2019-05-20T18:00:00.000+02:00">

<location lat="51.189445" long="4.3188912" /><temperature>20.5</temperature><humidity>0.78</humidity><PM10>19.46</PM10>

</Measurements>

DUET KNOWLEDGE GRAPH

19

THREE DIMENSIONAL INDEXenvironment

DUET

mobility

…

Air Quality

Water Quality

…

Traffic

…

Data Set 1

Data Set 2

DIMENSIONS Semantics Time Geography

STRUCTURED DATA THAT MACHINES CAN READ

Stable ‘cool’ identifiers for information (dereferenceable HTTP URIs as name for things)

Provide useful information, using the standards Reuse existing building blocks Often competing vocabularies

Add links, so people and machines can discover more data

20

SHARE AND REUSE

TRANSLATING DATA IN A WORLD OF STANDARDS

Connecting data sources to digital twins Standards (NGSI-xx, OGC, …) Non-standards map

Problem: Competing standards Missing features …

Approach: Internal definition in the platform for standards Unified data model (not a standard!) Onthology based mapping

21

ONTHOLOGY BASED MAPPING

AQ 1 AQ DT

NO2

O3

PM 10

PM 2,5

…

PM 10

PM 2,5

NO2

O3

KNOWLEDGE GRAPH

Registering data sources With context semantic annotations Explanation of the data source Other attributes s.a., quality, service level, license, etc.

Build a semantic inventory (the DUET catalog) that helps in searching for relevant data sources (see next slide)

Present search results in the form of most relevant data sources Make the taxonmy machine readable, explore possibilities of existing standards

such as DCAT and check for machine based discovery with comunica

22

REGISTER DATA SOURCES AND BUILD A SEMANTIC SEARCH GRAPH

BUILDING DIGITAL TWINS

23

BUDI APPROACH

Solution accelerator: A combination of a data source With a modelling component that analyses the data And one or more visualizations

YTT (yesterday/today/tomorrow) Yesterday: historical exploration Today: real-time monitoring Tomorrow:

Predictive What-if / simulations

Loosely couple components around scalable queueing technology

24

SOLUTION ACCELERATOR

25

DUET (HORIZON 2020)

26

DIGITAL URBAN EUROPEAN TWINS

Platform approach Builds upon BUDI architecture & components

Focus on EBP (evidence-based policies)

Platform approach Security Data broker Interoperable API’s Open data standards

(DUE)T-CELL ARCHITECTURE

27

event broker(pub/sub)

data catalog

master data

CITY FLOWSDATA

CITY GEO DATA

INTERACTIONCLIENT

CITY VISUALIZATIONCLIENT

MODEL

data adapters

DUET Management Module

Access & authorization Knowledge graph City Geo-data API IoT Data API Timeseries data API Simulation Model API Subscription API Data access API Interaction API Visualization API

PUBLIC