• No results found

Genera&ng Value from Big Data in the Internet of Things

N/A
N/A
Protected

Academic year: 2021

Share "Genera&ng Value from Big Data in the Internet of Things"

Copied!
18
0
0

Loading.... (view fulltext now)

Full text

(1)
(2)

Genera&ng  Value  from  Big  Data  

in  the  Internet  of  Things  

THT10421  

Cheng  Kian  Khor  

Global  Industry  Solu&on  Leader  -­‐    IoT/M2M  for  CSPs  

Communica&ons  &  Media  Industry  Solu&ons  Group  

 

 

 

(3)

Safe  Harbor  Statement  

The  following  is  intended  to  outline  our  general  product  direc&on.  It  is  intended  for  

informa&on  purposes  only,  and  may  not  be  incorporated  into  any  contract.  It  is  not  a  

commitment  to  deliver  any  material,  code,  or  func&onality,  and  should  not  be  relied  upon  

in  making  purchasing  decisions.  The  development,  release,  and  &ming  of  any  features  or  

func&onality  described  for  Oracle’s  products  remains  at  the  sole  discre&on  of  Oracle.  

(4)

The Internet

of Things

(5)

Moving  from  M2M  to  the  Internet  of  Things  

Device Cloud & Horizontal Enablers

Focus on

C

o

n

n

ec

ti

vi

ty

D

iffe

re

n

ti

ate

d

Va

lu

e

In

n

o

va

ti

ve

Se

rv

ic

es

IoT Ecosystem

Enablers

Tr

an

sfo

rm

In

d

u

str

ie

s

Process Enablers & Information

Mashup Enablers

Ve rtic al A pp lica tio ns Ve rtic al A pp lica tio ns Ve rtic al A pp lica tio ns Ve rtic al A pp lica tio ns Ve rtic al A pp lica tio ns Ve rtic al A pp lica tio ns Ve rtic al A pp lica tio ns Ve rtic al A pp lica tio ns Ve rtic al A pp lica tio ns Ve rtic al A pp lica tio ns

Reduced  Time  and  Cost  to  Market   with  Common  PlaWorm  

Rich  Process  Interac&on,   Informa&on  Aggrega&on  

Vert

ical

App

lica

tions

Vert

ical

App

lica

tions

Vert

ical

App

lica

tions

Vert

ical

App

lica

tions

Vert

ical

App

lica

tions

Vert

ical

App

lica

tions

Large  scale  ver&cal   applica&ons  

Deep,  transforma&ve  industry   func&onality  and  ecosystem    

(6)

IoT  Analy&cs  –  Genera&ng  Value  from  Big  Data

 

 UTILITIES  

 

 

 

 

 

 

AUTOMOTIVE  

 

 

 

 

 

 

HEALTHCARE  

 

 

 

 

 

 

MANUFACTURING  

 

 

 

 

 

 

Reduced  break-­‐fix    

&me  “from  90  to  30  days”  

“75%  fewer  truck  rolls”  to  fix  the  same  number  of  broken  

meters    

USD  6M  Recovered  in  losses  in  one  year  

Proac&ve  iden&fica&on  of  problema&c  meters  

50%  Reduc&on  in  field  crew  cost  

Predict  wear  and  tear  replacement  intervals  

Remote,  pro-­‐ac&ve  and  guided  diagnos&cs  

Traffic  /  Driving  Behaviour  analy&cs  

Usage  Based  Insurance  (UBI)  

Remote  asset  fix  rate      

“greater  than  50%”  

Reduced  field  service  dispatch  

Reduced  down-­‐&me  through  pro-­‐ac&ve  detec&on  of  

equipment  failures,  saving  “…hundreds  of  thousands”    

 

Reduced  healthcare  /  insurance  costs  

Improved  adherence/compliance  to  treatment  and  

monitoring  e.g.  CPAP  for  OSA  

Analy&cs  for  pa&ents/healthcare  providers  /  popula&on  

healthcare  managers  

 

 

(7)

Ac&onable  

Events  

Streaming  Engine  

Data  Reservoir  

Enterprise  Data  &  Repor&ng  

Discovery  Lab  

Ac&onable  

Informa&on  

Ac&onable  

Data  Sets  

Input  

Events  

Execu&on  

Innova&on  

Discovery  

Output  

Data  

Structured  

Enterprise    

Data  

Oracle  Approach  -­‐  Conceptual  View  

(8)

Real-­‐Time  Analy&cs  -­‐  Event  Processing  with  Spa&al  /  

Loca&on  /  Geofencing  

(9)
(10)
(11)

Analyze,  Visualize  and  Mone&ze  Automo&ve  Big  Data  

Example  Automo&ve  Dataset  from    

Telema&cs  Data:  

 

Number  of  trips  recorded  

Number  of  kilometers  logged  

Travel  by  Loca&on  

Petrol  consump&on    by  model    

Diagnos&c  Trouble  Codes  by  model    

Driver  Aggression  Profile  (e.g.  RPM  Profile)  

Distance  by  model  /  holiday  travel  

Traffic  Palern  Visualiza&on  

(12)

Smart  Metering  Analy&cs  -­‐  Collect,  Analyze  and  Act  to  

Increase  Efficiency  /  Reduce  Cost  

Monthly  register  reads  don’t  easily  

reveal  slowing  consump&on…  

 

 

Daily  data  reveals  there  is  a  trend,  but  

is  it  unusual,  or  weather  driven?  

(Yellow  =  temperature)  

 

 

Comparison  to  rate  class  behaviour  

(Red  –  rate  class  aggregate)  reveals  

that  the  palern  is  specific  to  the  meter  

 

Inclusion  of  meter  flag  /  event  data  

seals  the  deal:  meter  is  highly  likely  to  

be  failing  

(13)

Smart  City  -­‐  Parking  Data:    Collect,  Analyze,  Act  and  Expose  as  a  

Service  

SFpark: Putting Theory Into Practice / 89

Lessons learned

Creating the technical infrastructure for SFpark’s data needs has been a large undertaking. The following lessons have emerged thus far:

tDon’t do it yourself. Most internal IT organizations do not maintain the staffing levels or skill sets to implement the technology necessary for a SFpark-style program. Bring on an experienced team to build the technical infrastructure and integrate it with existing systems.

tMake sure your technology implementation team is involved in the first stages of the project management life cycle, beginning with contracting and procurement, long before it comes time to purchase servers. Have that team work with your existing IT team to ensure that technology choices fit in with your organization’s existing IT standards and direction.

tDon’t let product vendors (sensors and meters) determine the technical infrastructure. Create a data system that can interface with multiple vendors and will provide maximum control over how the data is managed and turned into information. Insist that project plans be expressed in terms of business deliverables. Vendors will want to give you a construction plan, but you want a feature implementation plan.

tExpect to spend more time in requirements discovery, business process reengineering, and off-plan work than expected. None of the vendors has ever done this type of project before, so workarounds and detours are commonplace.

tThe technological maturity of vendor products is much less than was anticipated.

tMost vendors do not have mature software

development, testing, and change control procedures.

Business intelligence tool automated report example

94/ Ch. 5: Parking technology

Mobile applications

SFpark provides on- and off-street parking availability and rate information via an iPhone app and soon an Android app.

Screen shots of iPhone app

(14)

Oracle  Big  Data  Management  Architecture  

SO U RC ES  

DATA  RESERVOIR  

DATA  WAREHOUSE  

Oracle  Database  

 

Oracle  Industry  

Models  

 

Oracle  Advanced  

Analy@cs  

 

Oracle  Spa@al  &  Graph  

Big  Data  Appliance  

Apache    

Flume  

GoldenGate  

Oracle  

Oracle  Event    

Processing  

Cloudera  Hadoop  

 

Oracle  NoSQL  

 

Oracle  R  Advanced  

Analy@cs  for  Hadoop  

 

Oracle  R  Distribu@on  

Oracle  Database  

 

In-­‐Memory,  Mul@-­‐tenant  

 

Oracle  Industry  Models  

 

Oracle  Advanced    

Analy@cs  

 

Oracle  Spa@al  &  Graph  

Exadata  

Oracle  

GoldenGate  

Oracle  Event  

Processing  

Oracle  Data  

Integrator  

Oracle  Big  Data  

Connectors    

Oracle  Data  

Integrator  

(15)

Ac&onable  

Events  

Streaming  Engine  

Data  Reservoir  

Enterprise  Data  &  Repor&ng  

Discovery  Lab  

Ac&onable  

Informa&on  

Ac&onable  

Data  Sets  

Input  

Events  

Execu&on  

Innova&on  

Discovery  

Output  

Data  

Structured  

Enterprise    

Data  

Oracle  Approach  –  Enterprise  IoT  Analy&cs  

(16)

Ac&onable  

Events  

Streaming  Engine  

Data  Reservoir  

Enterprise  Data  &  Repor&ng  

Discovery  Lab  

Ac&onable  

Informa&on  

Ac&onable  

Data  Sets  

Input  

Events  

Execu&on  

Innova&on  

Discovery  

Output  

Data  

Structured  

Enterprise    

Data  

IoT  Analy&cs  As  A  Service  –  An  Emerging  Opportunity  

Internet  of  Things  (IoT)  

ENTERPRISE  IOT  ANALYTICS  

IOT  DATA  STORAGE  AND    ANALYTICS  AS  A  SERVICE  

(PROVIDED  BY  CSP  or  OTT  PROVIDER)    

(17)

Genera&ng  Value  from  Big  Data  in  IoT    

                                       BI  &  DW  

Big Data Analytics

Platform

“Data Driven

Decision Making for

Competitive

Advantage”

Information as a

Service

“Monetize your Data”

Analytics as a

Service

“Competitive

Advantage for your

Customers”

“Up the M2M and IoT

(18)

References

Related documents

As discussed in Section III and illustrated in Figure 3, the human user typically has a number of interaction points with an information fusion system, generally in terms of

The loadings for each maturity with respect to the various factors (see equation 17) are shown in Figures 7 and 8. A number of observations can be drawn from Figure 8: i) a clear

In addition to capturing information about the impact of hemophilia on the physical and mental health of patients, these instruments also assess the detrimental impact of the

The main purpose of the study is to determine the ultimate load of the beam with and without compacted soil, as well as to investigate the relationship of the bolt spacing

Comparison of tumor size and location detection rates (for homogeneous phantom) with other studies using Equation (1).. Tumor size detection rate Tumor location

Develop a Sales Promotion Idea (Marks 2) Develop a Direct Marketing Plan (Marks 2) Develop a Event Sponsorship Plan (Marks 2) Develop a Social Media Plan (Marks 2).. Mini

1) Is there any significant difference between monolingual and bilingual English language learners in terms of Language Learning Strategies? H01: there is no significant

The Community Development Committee Report for June 9, 2004, was approved as presented and ordered placed on file in the City Clerk’s Office upon motion by Council Member