Mais conteúdo relacionado
Semelhante a Internet of Things (ruSMART 2013, St. Petersburg) (20)
Mais de Rob Van Den Dam (20)
Internet of Things (ruSMART 2013, St. Petersburg)
- 1. IBM Institute for Business Value
Rob van den Dam, Global Telecommunications Industry Lead IBV
Internet of Things: The foundational
infrastructure for a smarter planet
28 August 2013
© 2013 IBM Corporation
- 2. IBM Institute for Business Value
The emerging Internet of Things attracts many players
2
IBM Confidential
© 2013 IBM Corporation
- 3. IBM Institute for Business Value
Smarter Planet is IBM’s #1 research focus
Dublin
Research
Labs
Almaden c
Big Data &
Analytics
Storage
Nanotech
Healthcare
Smarter
Cities
Industry
TSL Russia
Solutions Labs(Moscow Satellite TSL)
Web
Services
TSL China &
Research Lab
c Watson
TSL Israel &
Semiconductors
Research Lab
Systems
(Haifa Satellite TSL)
Software &
Services
TSL Europe
(La Gaude &
Big Data &
TSL North America
Montpellier, France)
Research Lab
Analytics
(Austin, Texas)
(Beijing)
Japan
Internet of Things
Smarter Cities
(Tokyo
Satellite TSL)
(New Delhi)
TSL India
(Bangalore Satellite TSL)
ASEAN
(Kuala Lumpur
Satellite TSL)
TSL LATAM
Centers of
Excellence
3
(Sao Paulo,
Brazil Satellite TSL)
Institute
for
Business Africa
South
(Johannesburg, Satellite TSL)
Value
Worldclass
Partner
Ecosystem
© 2013 IBM Corporation
- 4. IBM Institute for Business Value
Internet-of-Things (IoT) provides the foundational infrastructure for
a smarter planet ….
Hyper-growth of instrumented devices for sensing the world
The infrastructure to connect the devices, manage the sensor data
and devices, and enable intelligence at edge and in data center
Significant growth opportunities for end-to-end infrastructure and
services
50 Billion
50
40
Safety and reliability of
physical infrastructure
30
Efficient asset utilization and
management
20
15 Billion
10
Why IoT is needed?
(SERCQ)
Reduction of energy and
7 Billion
resources
Comply with regulation
2010
2015
2020
Quality of service and life
Number of Connected Devices
Multiple Sources: Intel, Ericsson, Gartner, etc.
4
© 2013 IBM Corporation
- 5. IBM Institute for Business Value
..and differs from the Web on many aspects
End Point
IP based servers/PC/mobile devices,
resource rich, human engaged on
device
Geo-distribution
Outside data center, unmanned area
Data center centric, manned
Communication
Pattern
Uplink intensive, relative fixed access
pattern, burst means important event
Downlink intensive, flexible access
pattern
Data
10-100X volume of transactional system;
numeric & multimedia; 24*7 continuous data
streams; noisy sensor data
Human generated content, Text, XML
are popular;
QoS and
Reliability
Stringent if managing mission critical
physical infrastructure
Best effort
Cyber v.s.
Physical
5
Not necessary IP based, heterogeneous
devices, constraint resource, no human
engaged
Cyber & physical, influenced by physical
world behavior model
Cyber mainly
© 2013 IBM Corporation
- 6. IBM Institute for Business Value
IBM works on a number of top challenges in Large Scale IoT systems
Sense the
impossible
before
Higher
sensitivity
Communication
Power
Sensing
Distance
Reliability
Scale
Simulation &
Data
Behavior
model
Testing
Reverse workload
pattern
Throughput
Space & Time
Processing
Internet of Things
Resource
constraint
runtime
Collaborative
intelligence
Management
Computing
Automated
Intelligent
Edge
Scalable
Distributed
Security
6
End-to-end security with insecure
components and environment
Resource constraint
© 2013 IBM Corporation
- 7. IBM Institute for Business Value
As an example: challenges in communications include spectrum,
power and inference issues
Existing spectrum allocation for smart grid
application in China
SPECTRUM CRISIS
Power
10 groups for single
direction mode
15 groups for UL
15 groups for DL
• Spectrum is SCARCE
150KHz
• Fast grow traffic requires
more spectrum
150KHz
MHz
Source: US FCC
223.525
…...
224.650
228.075
…...
228.750
230.525
…...
231.650
Frequency
POWER
• Communication is power
intensive
• No external power supply in
certain domain
RF is the BIGGEST portion of energy
consumption of wireless sensors
More frequent communication, more
percentage, energy consumed by RF
INTERFERENCE
1) Multiple Interferences
• Spectrum sharing
2) Working with certain
periods
• Large Machinery
3) Unknown signals existed
in the same band
Spectrum Environment in Shanghai
7
Large machinery in mines, oil fields, etc.
© 2013 IBM Corporation
- 8. IBM Institute for Business Value
These challenges are addressed by IBM’s wireless M2M Platform,
based on IT-based software defined radio technology
Data Center
Technical Features
IoT Infrastructure
Management & Operation
Control Center
Monitoring
& Control
• E2E SDR technology for high flexibility
Command
Center
Simu
IoT Infrastructure
Manager
Communication
Server
SDH/SONET
IoT Edge
Appliance
Substation
• Optimized WiMAX-EXT for stable wireless
communication
• Optimization for spectrum and power
consumption
• Open IT platform for services and
application injection
Substation
3G/4G
3G/4G
Wireless
Wireless
IoT Access
Appliance
IBM Research is leading the
technology of IT-based Softwaredefined radio in the world
• Collaboration with top operators, e.g. China
RS232/ETH
Mobile, SK Telecom
• Successful demo in global industry events
Feeder
Terminal Unit
Transmission
Transformer
Terminal Unit
Feeder
Terminal Unit
Distribution
Transformer
Terminal Unit
Consumption
Tech Park
Petrol/Chemical Pipe
Airport
© 2013 IBM Corporation
8
- 9. IBM Institute for Business Value
Four key Technologies IBM is working on are key for Internet of Things
Natural
Language
Analytics
DeepQA
Data
Program
Cognitive
Computing
“SyNAPSE”
Learn
Big Data: Realtime
Inference &
Knowable
Future
1,000,000X
Workload
Optimised
Systems
1,000X
Silicon
Devices
Nano
1 Billion
Scale
Transistors
1,000X
9
Exascale
Software
Defined
Environments
Nano
Systems
1 Trillion Devices
© 2013 IBM Corporation
- 10. IBM Institute for Business Value
Cognitive computing aims to evolve to entirely new computing
architectures and programming paradigms.
Current
Specific
questions
Future
Statistical
analytics
Rich problem
scenarios
Interactive
dialogue
Continuous
learning
Evidence
profiles
This poet wrote to a
friend, “We are by
September and yet my
flowers are as bold as
June. Amherst has
gone to Eden.”
Batch training
10
Statistical
ranking
© 2013 IBM Corporation
- 11. IBM Institute for Business Value
Cognitive computing will be central to many future applications
Watson
Five-in-Five
Cognitive Sensing
Technology:
• Hearing and voice recognition
• Extracting knowledge from
pixels
• Sniffing for healthiness
• Haptic technology for retail
• Healthier molecular based
recipes
Sensor Networks / Internet of Things
Infrastructure
Grids
Buildings
Hearing
Sight
Smell
Taste
Touch
Social Business
Human and
knowledge
capital
analytics
TM
Metering
Vehicles
Billions of end points
100K+ elements,10ms latency
Multiple feedback time-scales
11
© 2013 IBM Corporation
- 12. IBM Institute for Business Value
From research - to battling humans on Jeopardy! - to changing the
way the world thinks, acts and operates
IBM
Research
Project
(2006 – )
Jeopardy!
Grand
Challenge
(Feb 2011)
Watson
for
Healthcare
(Aug 2011 –)
Watson
for Financial
Services
(Mar 2012 – )
Watson
Industry
Solutions
(2012 – )
Cross-industry
Applications
Expansion
Commercialization
Demonstration
R&D
12
© 2013 IBM Corporation
- 13. IBM Institute for Business Value
Watson infrastructure comprises a platform with capabilities to
address specific smarter planet areas
ASK Services
NLP & Machine
Learning
Source
13
Big Data
Watson For
Financial Svcs.
Model
Cloud
Train
Institution
Retirement
DISCOVER Services
Analytics
Watson for
Industry
Advisor Solutions
Institutional
Technical
Knowledge
Help Desk
Call Center
Advisor Solutions
Diabetes
Cardiac
Oncology
Utilization
Advisor Solutions
Watson for Client
Engagement
Banking
Watson for
Healthcare
DECISION Services
Mobile
Workload Optimized
Systems
Learn
© 2013 IBM Corporation
- 14. IBM Institute for Business Value
In healthcare, for example, contextual services combine many
sources of data to benefit patients and providers
Anne’s daughter contacts her inApp PA and requests guidance.
From smart device, PA sees
Anne’s diabetes, recent long
flight and the local weather,
determines she is seriously
dehydrated, recommends IV
rehydration therapy and notifies
Wellness program provider.
Anne, a 76 year old
grandmother, becomes dizzy and
faints during an outdoor event in
midsummer heat and humidity
Wellness program provider
engages OTTER to evaluate
Anne’s entire medical history.
Watson MD notices a cardiac
event and congestive heart
failure (CHF) 10 years prior.
Recommends rehydration
therapy only at a cardiac care
center due to risk of triggering
recurrence of CHF.
MD
Wellness program provider
authorizes fast-tracking Anne
through triage directly to the
cardiac care unit for
rehydration therapy. Nearest
cardiac care center location
(given current traffic) is
forwarded to Anne’s
daughter’s vehicle for
transport.
OTTER
Context
Value
Context
Value
Context
Value
Context
Value
Name
Anne
Chronic Illness
Diabetes
Medical History
EMR
yes
Age
76
Travel
3 hrs flight today
History of CHF
yes
Cardiac protocols
authorized
Location
outdoors,
Miami, FL
Weather conditions
97 degrees,
90% humidity
Last CHF episode
Aug. 9, 2002
Fast-track triage
authorized
yes
dizzy, fainting
Diagnosis
Dehydration
Closest Cardiac
Care location
201 Ridge Rd,
Miami, FL
Time
3:40 PM
Treatment
Recommended
IV rehydration
therapy
Risk of CHF
recurrence without
Cardiac protocols
60%
Symptoms
Dedicated
Traffic delay
estimate
10 minutes
Wellness Program
Provider
StayWell Health
Insurance
Family Support
Network
Cardiac Care
Estimate Time of
Arrival
4:30 PM
Healthcare PoA
Daughter, On file
Treatment Location
Recommended
$1500
yes
Nearest Urgent
Care center
Estimated Cost
CC Staff notified
Treatment Location
Recommended
CHF Cost Risk
$100,000
Cost Averted
$85,000
Estimated Cost
$300
14
© 2013 IBM Corporation
- 15. Today’s environment are making workloads (and Networks) more
volatile
Volatile workload characteristics result from changing business requirements
Traditional
Current
Future
C
C
W1
W2
W3
W4
V1
V2
V3
V4
V5
…
Vn
V1
R1
R2
V3
V4
V5
V5
...
….
Vn
R3
Few, stable, and well known
workloads
Fixed System hardware,
manual scaling
Hardwired workload, minimal
configuration
15
V2
Source: IBM Global Technology Outlook 2013
Workload virtualized, to cloud
Diverse workload, limited
patterns
Homogeneous resource
pooling
Expert configuration and
mapping of workload
Rapidly changing workloads,
dynamic patterns
Dynamic automatic
composition of heterogeneous
system
Autonomic and proactive
management
© 2013 IBM Corporation
- 16. Software Defined Environments provides abstractions of workloads,
services and infrastructure and an end-to-end mappings
Workload Abstraction
Functional and non-functional requirements
that may be discovered as well as specified
Resource Abstraction
Semantically rich abstractions of
heterogeneous resource capabilities and
system components
Mapping to resource
Map requirements to potential system
architectures. Proactively orchestrate
.
infrastructure and workload
Continuous Optimization
Autonomously construct available system
architecture to optimize workload outcome
Agility
Consumability
16
Efficiency
© 2013 IBM Corporation
- 17. IBM Institute for Business Value
Summary
COGNITIVE
SYSTEMS
BIG DATA
INTERNET
OF
SDE
THINGS
CLOUD
17
© 2013 IBM Corporation
- 18. IBM Institute for Business Value
Thank you
Rob van den Dam
www.ibm.com/iibv
Global Telecom Industry Leader
IBM Institute for Business Value
rob_vandendam@nl.ibm.com
18
© 2013 IBM Corporation