Event held 8th Dec 2016, Edinburgh. The evolution of Big Data analytics has been staggering: it has progressed from an underused asset to a vital source of intelligence and insight, driven by improved hardware, cloud technologies and a plethora of specialist software. These technological advances have pushed the boundaries of what is possible, driving new innovation and enabling huge strides forward in fields like AI and Cognitive Computing.
12. Boots UK
* Figures are approximations as at 31 March 2012 and include associates and joint ventures
88% of population within 10 minutes of a
Boots store
Nearly 2,500 Boots stores
60m visitors each year to boots.com
Nearly 625 Boots Opticians practices
17.8 million Boots Advantage Card members
45%order online and collect in-store
13. To be the world’s leading
pharmacy-led health and
beauty retailer
Boots mission
15. We have to really “Get Women”
Truly Customer Led
Boots Understands Women
Through Great Insight
16. Advantage card at the core
Shops on weekdays at
lunchtime in a local
store
Buys vitamins –
health conscious
Is a parent with a
young baby
3 for 2 offers
Boots Advantage
Card number
Buys into meal
deal offer
17. Every time our customer shops
Shops in large
store Saturday
mornings
Redeems
coupons
Purchases self-
selection cosmetics,
but also premium
cosmetics
Could have a partner?
Boots Advantage Card number –
same as previous receipt!
21. Context of the empowered customer
• More touch points
• More complex and faster changing
opinions
• Expectation that you use insight
• Seamless multi-channel delivery
22. Our multi-channel approach
UK’s No1 visited health and beauty website*
* excludes National Health Service, based on most recent information provided by Experian Hitwise
25. Single view of the customer
Web metrics by
device
Role of different
devices in the
same customer
journey
Impact of Advice &
Info on customer
behaviour
Identifying and understanding
the same customer online and
offline – browsing, purchasing,
sharing
How do consumers
influence one another?
And who is really
influential?
26. Single view of the customer
Cheryl – 33
Living alone no children
Active Ad Card Member
27. Drives insight driven communication
Transactional Data:
We know what they
bought
Who to speak to? About what?
Demographic Data:
We know who
the customer is
Response Data:
We know who
responds to offers
Contact Data:
We know who
received offers
28. …Becomes omni channel optimisation
Delivering ‘Let’s Feel Good’
Traditionally a direct mail
focus
Loyalty
Comms
Mail
Kiosk
Tills
ClubsEmail
App
Text
Now active via multiple
channels
40. Personalisation delivers results
Generic 3 for 2 offer
Personalised to each customers
favourite skincare products
vs
Redemption rates (%)
“I like that the coupons relate to the
products which I buy. It makes it feel
like you have gone that bit extra to
know your customers.”“If coupons are
more relevant, you
are more likely to
go out of your way
and make a
special visit”
43. Zopa – a story of growth
Vlasios Vasileiou
Head of Data Science
44. 44
Peer to Peer Lending at Zopa
• Credit & fraud risk
• ID verification
• Pricing
• Loan servicing
Retail & Institutional Lenders Borrowers
45. 45
Peer to Peer Lending at Zopa
• Credit & fraud risk
• ID verification
• Pricing
• Loan servicing
Retail & Institutional Lenders Borrowers
~3 - 6% annualized return Better interest rates
Faster, simpler service
46. 46
Zopa
Launched 2005, inventing peer to peer (P2P) lending
Largest P2P platform in Europe
56,000 active retail lenders
Lent £1.8bn of unsecured personal loans to over
230,000 UK borrowers
49. 49
Zopa during analytical renaissance (2014 – 2015)
2014, £15m investment
Board recognized need for data-driven growth
Creation of Data Science function
Systematization of Data Analytics
50. 50
Systematizing Machine Learning at Zopa (2014)
Wanted to be able to produce ML models that were:
Rapidly generated
easily-vettable
highly-predictive
easily deployable
Several considerations:
• Common codebase or personal choice of tools?
• Buy or build?
• Which language? Which package?
51. 51
Systematizing Machine Learning at Zopa (2014)
Wanted to be able to produce ML models that were:
Rapidly generated
easily-vettable
highly-predictive
easily deployable
Several considerations:
• Buy or build?
• Which language? Which package?
Common codebase
52. 52
Systematizing Machine Learning at Zopa (2014)
Wanted to be able to produce ML models that were:
Rapidly generated
easily-vettable
highly-predictive
easily deployable
Several considerations:
• Which language?
Common codebase
Built in-house
53. 53
Systematizing Machine Learning at Zopa (2014)
Wanted to be able to produce ML models that were:
Rapidly generated
easily-vettable
highly-predictive
easily deployable
Common codebase
Built in-house
54. 54
Streamlined and Automated ML Application
• Leverage PyData Tools (sklearn, pandas, xgboost, keras, …)
• 9k lines
• Used and improved by all Zopa data scientists
• Combines external toolkits + best practices in ML
Predictor – Zopa’s ML Toolkit (2014)
55. 55
First big win – our Credit Risk Model (2015)
• Credit-risk estimation: a core
component of our operations
• Pre 2015, using externally obtained
credit-risk scores & models
£0
£100
£200
£300
£400
£500
£600
£700
2005 2007 2009 2011 2013 2015
Disbursals(Millions)
Year
Annual disbursals
56. 56
First big win – our Credit Risk Model (2015)
• Q1 2015, built and deployed own
credit-risk model in-house
£0
£100
£200
£300
£400
£500
£600
£700
2005 2007 2009 2011 2013 2015
Disbursals(Millions)
Year
Annual disbursals
57. 57
First big win – our Credit Risk Model (2015)
£0
£100
£200
£300
£400
£500
£600
£700
2005 2007 2009 2011 2013 2015
Disbursals(Millions)
Year
Annual disbursals
+100%
• Q1 2015, built and deployed own
credit-risk model in-house
New model considerably more predictive
than previous one
100% increase in disbursals yoy
58. 58
2015 – 2016, Emerging Data Culture
Data-driven
wins
Commitment
to data
59. 59
2015 – 2016, Emerging Data Culture
Data-driven
wins
Commitment
to data
Accelerators
• Embedded data science
• Two-way training/outreach
• Tool sharing
60. 60
10 ML Models currently used for decisioning, more under consideration
• Borrower application pipeline (7 active models)
• Pricing (2 models)
• Marketing (1 model)
• Customer satisfaction
• Collections
2015 – 2016, Machine Learning Proliferation at Zopa
62. 62
Improving Data Governance and Federation, 2016 –
Diminishing returns of increasing
modelling sophistication
Need better & more data
63. 63
Improving Data Governance and Federation, 2016 –
Diminishing returns of increasing
modelling sophistication
Need better & more data
Data analytics only as good as your
data quality/availability
govdelivery.com
64. 64
Improving Data Governance and Federation, 2016 –
Diminishing returns of increasing
modelling sophistication
Need better & more data
Data analytics only as good as your
data quality/availability
Break down the silos!
govdelivery.com
65. 65
Improving Data Governance and Federation, 2016 –
Data warehouse with AWS Redshift
In progress
Data lake
Planned
68. About me
• Education in Physics/Astrophysics
• Researcher in Astrophysics
• Joined Zopa as a data scientist, 2014
• “Head of Data Science”, late 2015
76. “We have grown gross sales and market share across both
Waitrose and John Lewis, but our profits are down.”
- Sir Charlie Mayfield, Chairman of JLP
Maybe not…
77. Faulty Can’t be resold as
new
Up to 60% returned
*IHL Group, Retail Analysts, June
Returns cost retailers £435billion globally*
£221 billion preventable retail returns
& returns are growing faster than sales
It’s not a sale until the customer decides to keep it
79. Accessories
Purchased96 items
GrossSales= £1,144
Womenswear
Purchased104items
GrossSales= £1,845
Menswear
Purchased28 items
GrossSales= £349
Health & Beauty
Purchased8 items
GrossSales= £352
Total Net after Refunds & Costs = £63
-£45 Net after refunds& costs. -£203 Net afterrefunds & costs.
-£41 Net after refunds& costs. £352 Net
Returned 88 Returned 100
Returned 25 Returned 0
Total Sales = £3,690
A Retailer’s Dream Turned into a NightmareTurned Nightmare
81. “Feminine and figure
enhancing”
Stiff structure only
fits one particular
body type
“Berry lace panel fit…
wear with heels and a
statement clutch”
Too bright, poor design
gets caught on all
accessories
Finding the Toxic Products
82. Improving the Bottom Line
Data
Focus on
the Right
Metrics
Reward the
Right
Behaviours
Optimize
Profits
89. #scotdata
• Dr Nava Tintarev, Bournemouth University
• Ken Macdonald, ICO
• Martin Squires, Walgreen Boots Alliance
• Dr Hannah Rudman, Rudman Consulting
91. MOTIVATION
• Chartered Institute of Marketing (CIM) survey with 2500 people
• Nine in ten people have no idea what companies do with the personal
information the firms hold about them.
• Personal data policies on websites should be clearer and simpler.
Source: http://www.bbc.co.uk/news/business-37476335, published September 2016
• ESRC event ``What is the Internet Hiding FromYou’’?
• In May 2016, the EU passed a General Data Protection Regulation (effective
from 2018) which will also create a ``right to explanation'’: user can ask for
an explanation of an algorithmic decision that was made about them.
92. WHAT IS THIS IS AND IS NOT.
• This session is not about pointing fingers.
• It is about having a conversation about what happens with personal DATA.
• What users are willing to share
• … and what they should expect to receive in return.
• This is new ground, we have not been here before.
• We will need to have a lot of conversations.
93. PANEL MEMBERS
• Dr Hannah Rudman, Director, Rudman Consulting
• Ken Macdonald, Head of ICO Regions, Scotland, NI & Wales, ICO
• Martin Squires, Global Lead, Customer Intelligence and Data,Walgreen BootsAlliance
• Nava Tintarev,Assistant Professor,Intelligent Interactive Systems, Bournemouth University
96. INTRUSIVE DATA ANALYTICS
When do analytics become too intrusive?
When can we make inferences across data sources, or
inferences that users did not consent to being made when
they initially provide the data?
97.
98. TRANSPARENCY OF ANALYTICS
How should we make algorithmic biases visible to users?
How do we avoid filter bubbles like the one that happened
during Brexit? How can explanations be used to improve
transparency?
99.
100. LEGISLATION OF PRIVACY
Is there going to be a swing in the balance of power towards
individuals / consumers? How do we balance this with
businesses' need to be competitive?
103. DATA ANONYMIZATION AND RE-
IDENTIFICATION
• 87% of US residents can be uniquely identified by zip+DOB+gender
• Sent the Massachusetts Governor his own medical records based on
publically available data
• Working paper: Uniqueness of Simple Demographics in the U.S.
Population. Latanya Sweeney
104. DATA ANONYMIZATION AND RE-
IDENTIFICATION
• In 2006,AOL (America OnLine) released detailed web search logs of a large number of its
users.
• The release was intentional, and aimed at promoting academic research; however, there was no
restriction on who could see the information.
• The user information (named and usernames) was anonymized (by replacing it with a unique
number). However,AOL did not redact search query.
• Soon, it was clear that search queries were enough to identify the users:
• The NewYork Times was able to locate an individual from the released and anonymized search
records by cross referencing them with phonebook listings
Source:A Face Is Exposed for AOL Searcher No. 4417749. M. Barbaro andT. Zeller Jr.August 9, 2006
• As many search information contained sensitive details (medical, sexual orientation, …) and re-
identification was possible,AOL removed the data.
105. WHAT IS THE FILTER BUBBLE?
• Tailoring information (personalization) may result in insufficient
exposure to items outside of their existing interests: `filter bubbles'
[Pariser, 2011].
• People have a tendency to self-filter [Bakshy et al, 2015].
• This is a real risk: many online `big data' systems (e.g. Facebook) already filter what
people are exposed to, often without their awareness.
• This creates polarized views, and segregated online communities.
• Explanations can help widen user(s)’ views while justifying choices
outside the user(s)’ usual sphere of interest.
• I have a responsibility to address this as a personalization technologist!
110. Extract using Python.
1
1
0
part
of the
• Weblog data source on SFTP server.
• Create Amazon EC2/Azure VM Instance
• Sample python Script to get/copy file
sftp.get(file, local_file)
s3_client.upload_file(local_file, bucket, s3_file_base)
• Copy Python/shell script on VM
• Automate script using cron job on Linux VM.
• Sample cron job
15 * * * * /usr/bin/python <path to python script>
111. Transform using Spark Scala/Python
1
1
1
part
of the
• Apache Spark on AWS/Azure
• SBT (Source build Tool for Scala Java).
• Package up source code in a Jar file.
• Create AWS EMR cluster/Azure HDInsight Cluster of desired configuration with
Apache Spark running.
• Add an EMR step to run jar file.
• Create AWS Data Pipeline to automate the Transform process.
• If using MS Azure Orchestrate using Azure Data Factory.
112. Load into AWS Redshift db
1
1
2
part
of the
• Create Redshift cluster of desired configuration.
• Create a sample database/role/user.
• Use AWS Copy command to load spark output file into redshift db.
COPY dbo.customer
FROM 's3://EdinburghDemo/myfile.txt.gz'
CREDENTIALS 'aws_access_key_id=<>;aws_secret_access_key=<>'
delimiter ',' IGNOREHEADER 1 gzip;
COMMIT;
114. Next Steps…
1
1
4
part
of the
• Create a free account
• AWS or MS Azure
• Create EMR/HDInsight cluster.
• Copy jar file to AWS S3 or MS Blob Storage.
• Run jar file using spark step.
• Save output on cloud storage of your choice.
• Load output file into AWS Redshift db or MS Azure Sql db.
118. Harnessing the CERN network
for analysis, insight and
understanding
Dr. Ian Bird
CERN Senior Staff Scientist &
LHC Computing Project Lead
Edinburgh; 8th December 2016
119. The Mission of CERN
Push back the frontiers of knowledge
E.g. the secrets of the Big Bang …what was the matter like within the first
moments of the Universe’s existence?
Develop new technologies for accelerators and
detectors
Information technology - the Web and the GRID
Medicine - diagnosis and therapy
Train scientists and engineers of tomorrow
Unite people from different countries and cultures
8 Dec 2016 Ian.Bird@cern.ch
11
9
120. 120
CERN: founded in 1954: 12 European States
“Science for Peace”
Today: 22 Member States
Member States: Austria, Belgium, Bulgaria, Czech Republic, Denmark, Finland,
France, Germany, Greece, Hungary, Israel, Italy, Netherlands, Norway, Poland,
Portugal, Romania, Slovak Republic, Spain, Sweden, Switzerland and
United Kingdom
Associate Member States: Pakistan, Turkey
States in accession to Membership: Cyprus, Serbia
Applications for Membership or Associate Membership:
Brazil, Croatia, India, Lithuania, Russia, Slovenia, Ukraine
Observers to Council: India, Japan, Russia, United States of America;
European Union, JINR and UNESCO
~ 2300 staff
~ 1600 other paid personnel
~ 12700 scientific users
Budget (2016) ~1000 MCHF
8 Dec 2016 Ian.Bird@cern.ch
12
0
121. Science is getting more and more global
8 Dec 2016 Ian.Bird@cern.ch
12
1
122. 8 Dec 2016 Ian.Bird@cern.ch
12
2
Evolution of the Universe
Test the
Standard
Model?
Dark Matter?
Dark Energy?
Anti-matter?
(Gravity?)
123. The Large Hadron Collider (LHC)
• Largest Scientific Apparatus ever built
• World’s most powerful particle accelerator
• Two multi-purpose and two specialized
detectors
• Probes the conditions of the universe a
fraction of a second after the big bang
8 Dec 2016 Ian.Bird@cern.ch
12
3
125. Data Analysis at the LHC
The process to transform raw data into useful physics datasets
This is a complicated series of steps at the LHC (Run2)
Data
Volume
Processing
and people
HLT Reconstruction Reprocessing Organized
Analysis
Final
Selection
40k
cores
60kcores
20k
cores
30k
cores
DAQandTrigger
(lessthan200)
Operations
(lessthan100)
Operations
(lessthan100)
AnalysisUsers
(lMorethan1000)
AnalysisUsers
(lMorethan1000)
SelectedRAW
(1GB/s)
DerivedData(2
GB/s)
FromDetector(1PB/s)
AnalysisSelection
(100MB/s)
AfterHardwareTrigger(TB/s)
DerivedData(2
GB/s)
8 Dec 2016 Ian.Bird@cern.ch
12
5
126. 12
6
Tier-1: permanent
storage, re-processing,
analysis
Tier-0
(CERN and Hungary):
data recording,
reconstruction and
distribution
Tier-2: Simulation,
end-user analysis
> 2 million jobs/day
~650k CPU cores
500 PB of storage
~170 sites,
42 countries
10-100 Gb links
WLCG:
An International collaboration to distribute and analyse LHC data
Integrates computer centres worldwide that provide computing and storage
resource into a single infrastructure accessible by all LHC physicists
The Worldwide LHC Computing Grid
8 Dec 2016 Ian.Bird@cern.ch
127. 8 Dec 2016 Ian.Bird@cern.ch
12
7
Optical Private Network
Support T0 – T1 transfers
& T1 – T1 traffic
Managed by LHC Tier 0 and
Tier 1 sites
Networks
Up to 340 Gbps transatlantic
128. 8 Dec 2016 Ian.Bird@cern.ch
12
8
Asia North America
South America
Europe
LHCOne: Overlay network
Allows NREN’s to manage HEP
traffic on general purpose network
Managed by NREN collaboration
0
10
20
30
40
50
60
70
JAN FEB MAR APR MAY JUN JUL AUG SEPT OCT NOV DEC JAN FEB MAR APR MAY
131. Data distribution
8 Dec
Ian.Bird@cern.ch
13
1
CERN (CPU)
CERN (Disk)
WLCG
LHC expts
FTS
CERN (Tape)
Regular transfers of >80 PB/month with ~100 PB/month during July-October
(many billions of files) (>50 GB/s globally)
133. Compute services
Cloud compute on OpenStack at CERN
- heart of the global federated structure
Moving towards Elastic
Hybrid IaaS model:
• In house resources at full
occupation
• Elastic use of commercial
& public clouds
• Assume “spot-market”
style pricing
OpenStack Resource Provisioning
HTCondor
Public Cloud
VMsContainersBare Metal and HPC
LSF
Volunteer
Computing
IT & Experiment
Services
End Users CI/CD
APIs
CLIs
GUIs
Experiment Pilot Factories
8 Dec 2016 Ian.Bird@cern.ch
13
3
134. Archive storage – tape
Tape system – key optimisation:
per stream speed
High throughput/high latency
Largest physics data repository
worldwide: 200 PB / 500 M files
8 Dec 2016 Ian.Bird@cern.ch
13
4
LHC Raw Data Recording
11 PB in July
Total LHC Data : 160 PB
Tape technology for data repositories: TCO
media
power
density
Reliability/resilience
4 Oracle SL8500 libraries: 40k slots
3 IBM TS3500 libraries: 26k slots
104 drives; 8 & 10 TB tapes Tape stores at all Tier 1 sites
135. Disk pools – extreme performance Designed for very high performance open/read; low
latency
In-memory namespace
Highly scalable
Open source
JBOD commodity hardware
Ignore failed disks
Use replication and erasure coding for reliability and
performance
Geo-localisation (distributed Data Centre)
Tunable QoS
Choose level of reliability/cost/performance
Many protocols supported
Strong security (Kerberos, X509)
Fine grained access control, quotas
8 Dec 2016 Ian.Bird@cern.ch
13
5
136. Data distribution – file transfer
Reliability/performance are the
key
Open source low level data
mover
“Move file F from A to B”
Highly scalable
>80 PB per month
1 million files per day
Adaptive optimisation of
storage and network
Supports GridFTP, HTTP,
S3, [SRM, xrootd, ...]
8 Dec 2016 Ian.Bird@cern.ch
13
6
137. Data federations
Key: global namespace
Allows on-the-fly access to
remote data sets
Also allows remote (WAN) I/O
8 Dec 2016 Ian.Bird@cern.ch
13
7
Provides a global namespace
Unifies dCache, DPM, Lustre/GPFS, Xrootd storage
backends
Xrootd an efficient protocol for WAN access
Main Fall-back use case in production at many sites
Regional redirection network provides lookup scalability
Browser-friendly realtime scalable aggregator of
HTTP/WebDAV/S3/MS-Azure metadata sources.
Aggregates/caches/presents metadata,
redirects clients to resources for reading or
writing. Geography-aware redirections
Presentation is via WebDAV and HTML
Low latency realtime behavior, can be used
in LAN and WAN
138. 13
8
Storage federation – R&D
aka “exploring the 300 ms region”…
ASGC
AARNET
CERN
AARNET, ASGC and CERN collaboration8 Dec 2016 Ian.Bird@cern.ch
139. CERNBox
CERNBox provides a cloud synchronisation service
Synchronise files (data at CERN) and offline data access
Easy way to share with other users
All major platforms supported
Based on ownCloud integrated with EOS
• Available for all CERN users (1TB/user initial quota)
Much more than a Dropbox™ replacement!
13
9
8 Dec 2016 Ian.Bird@cern.ch
142. Hadoop and Analytics
Hadoop Production Service
New scalable data services
Scalable databases
Hadoop ecosystem
Time Series databases
Big Data Analytics
Activities and objectives
Develop projects and services with/for users
Support of Hadoop Components
Further value of Analytics solutions
Define scalable platform evolution based on
requirements
142
8 Dec 2016 Ian.Bird@cern.ch
144. Future Challenges
Raw data volume for LHC increases
exponentially and with it processing
and analysis load
Technology at ~20%/year will bring
x6-10 in 10-11 years
Estimates of resource needs at HL-
LHC x10 above what is realistic to
expect from technology with
reasonably constant cost
PB
First run LS1 Second run LS2 Third run LS3 HL-LHC
…
FCC?
2009 2013 2014 2015 2016 2017 201820112010 2012 2019 2023 2024 2030?20212020 2022 …
0
100
200
300
400
500
600
700
800
900
1000
Raw Derived
Data estimates for 1st year of HL-LHC (PB)
ALICE ATLAS CMS LHCb
0
50000
100000
150000
200000
250000
CPU (HS06)
CPU Needs for 1st Year of HL-LHC (kHS06)
ALICE ATLAS CMS LHCb
2025
CPU:
• x60 from 2016
Data:
• Raw 2016: 50 PB 2027: 600 PB
• Derived (1 copy): 2016: 80 PB 2027: 900 PB
8 Dec 2016 Ian.Bird@cern.ch
14
4
145. HEP Data cloud
Storage and compute
1-10 Tb/s
DC
DC DC
Compute
Compute
Cloud users:
Analysis
8 Dec 2016 Ian.Bird@cern.ch
14
5
Possible Model for future HEP computing infrastructure
Simulation resources
146. 8 Dec 2016 Ian.Bird@cern.ch
14
6
CERN open data portal
154. 154 Copyright 2016 FUJITSU
Smart Cities, Big Data
Michael Mooney
Smart Cities Advisor, Fujitsu
Vasilis Kapsalis
Converged Infrastructure EMEIA, NetApp
155. 155 Copyright 2016 FUJITSU
Overview
Overview of Smart City Projects: Examples of the types of Smart
Cities projects currently rolling out in the UK, Europe and Japan
Big Data Demands from Smart Cities: Are there unique challenges
coming out of IoT and Smart Cities Projects?
DM
156. 156 Copyright 2016 FUJITSU
Definitions of a Smart City
British Standards Institute: A city is smart when it displays effective
integration of physical, digital and human systems in the built environment
to deliver a sustainable, prosperous and inclusive future for its citizens.
DM
157. 157 Copyright 2016 FUJITSU
Definitions of a Smart City
BSI: A city is smart when it displays effective integration of physical, digital and human systems in the built environment to
deliver a sustainable, prosperous and inclusive future for its citizens.
Japanese Smart Community Alliance : “A smart community is a community
where various next-generation technologies and advanced social systems are
effectively integrated and utilized, including the efficient use of energy, utilization of
heat and unused energy sources, improvement of local transportation systems and
transformation of the everyday lives of citizens.”
DM
158. 158 Copyright 2016 FUJITSU
Definitions of a Smart City
BSI: A city is smart when it displays effective integration of physical, digital and human systems in the built
environment to deliver a sustainable, prosperous and inclusive future for its citizens.
Japanese Smart Community Alliance : “A smart community is a community where various next-generation
technologies and advanced social systems are effectively integrated and utilized, including the efficient use
of energy, utilization of heat and unused energy sources, improvement of local transportation systems and
transformation of the everyday lives of citizens.”
International Standards Organisation: A ‘Smart City’ is one that……
dramatically increases the pace at which it improves its social economic and
environmental (sustainability) outcomes, responding to challenges such as climate
change, rapid population growth, and political and economic instability …… by
fundamentally improving how it engages society, how it applies collaborative
leadership methods, how it works across disciplines and city systems, and how it
uses data information and modern technologies……in order to provide better
services and quality of life to those in and involved with the city now and for the
foreseeable future, without unfair disadvantage of others or degradation of the
natural environmentDM
159. 159 Copyright 2016 FUJITSU
Definitions of a Smart City
BSI: A city is smart when it displays effective integration of physical, digital and human systems in the built environment to
deliver a sustainable, prosperous and inclusive future for its citizens.
ISO: A ‘Smart City’ is one that…… dramatically increases the pace at which it improves its social economic and
environmental (sustainability) outcomes, responding to challenges such as climate change, rapid population growth, and
political and economic instability …… by fundamentally improving how it engages society, how it applies collaborative
leadership methods, how it works across disciplines and city systems, and how it uses data information and modern
technologies……in order to provide better services and quality of life to those in and involved with the city now and for the
foreseeable future, without unfair disadvantage of others or degradation of the natural environment
Japanese Smart Community Alliance : “A smart community is a community where various next-generation technologies
and advanced social systems are effectively integrated and utilized, including the efficient use of energy, utilization of heat
and unused energy sources, improvement of local transportation systems and transformation of the everyday lives of
citizens.”
Smart Cities are easier to live in.
DM
172. 172 Copyright 2016 FUJITSU
StorageGrid
NetApp Vertical IoT
ISV Integration
IoT Data Aggregators and Cleansers
IoT Sensors and Data Generators
S-DOTFlexPod Express
Edge Sensor
VM
Edge Sensor
VM
Edge Sensor
VM
FlexPod
Managed Edge Cloud
FlexPod Express
HyperScaler
Cloud with
CloudOnTap
Cloud Service
Providers with
ONTAP
Private Cloud
with ONTAP
IoT Platforms, Big Data Analytics And Predictors + archive
Wireless Devices
Edge Sensor
VM
Edge Sensor
VM
Edge Sensor
VM
Controllers – Near and Real Time
Customer owned products
Enterprise Apps/DB
IoT Framework and Ecosystem
173. 173 Copyright 2016 FUJITSU
Example - Ecosystem of IoT Partners
1
7
3
HyperscalersService Providers
Technology
Partners
179. An automatic system
developed to disseminate
information to the various sections
of any industrial, scientific, or
government organisation.
Hans Peter Luhn, IBM Researcher
1958
180.
181. Concepts and
methods to improve business
decision making by using
fact-based systems.
Howard Dresner, future Gartner Analyst
1989
193. Many organisations are looking to pursue
standardisation
Source: TDWI Research
A third of organisations plan to
standardise on a single BI tool within two
years
200. • Radically reduce the risk and
expense of standardising a
business intelligence estate
• Future-proof technology and
information for changes in
direction, leadership or from
mergers and acquisitions
• Gets the security, controls
and scalability they require
• No longer needs to learn
multiple BI systems or hunt for
information across them
• Able to make joined-up,
strategic decisions based on
one version of the truth across
the entire organisation
• Free to keep using the tool
that they know and love
IT USERS