8. Session title
8
THE RISE OF THE
PERVASIVE SENSOR
Credit: Eric Fischer
Complete set: https://www.flickr.com/photos/walkingsf/sets/72157624209158632/
San
Francisco
10. Session title
10
SOME FACETS TO CONSIDER
• Consumer vs. Industrial
• No IP address vs. IP address
• How heavy is the client?
• How is the client connected?
• What does the client connect to?
• How and where is data processed?
14. Session title
14
• Connected things not new, but…
• Value from greater connectivity options and
how data used to improve process
capabilities and automate decisions
• Integration with existing IT nascent
• High potential value—but also requires high
process and culture change
GARTNER: INTERNET OF THINGS FOR
MANUFACTURING OPERATIONS
15. Session title
15
THE INDUSTRIAL IOT
“We need the IT side to understand the
business value and priorities of control
integration, and to take more of a
leadership role. The controls side needs to
understand the need to partner and work
more closely with IT to take advantage of
these technologies.”
--Dan McGrath, Panduit
20. Session title
20
• The Thread Group
• IoT-GSI Open Interconnect Consortium
• The AllSeen Alliance
• HyperCat
• HomeKit
• Industrial Internet Consortium *
• OASIS IoT/M2M
• Weightless
MANY STANDARDS ORGS WITH DIFFERENT GOALS
Much overlapping membership
22. Session title
22
SOME IOT MESSAGING PROTOCOLS: NEW HTTP?
Message Queue Telemetry Transport (MQTT)
• Pub/sub over TCP (one-to-one, one-to-many and many-to-one)
• Lightweight M2M
• Originally IBM—open standard
Constrained Application Protocol (CoAP)
• Document transfer protocol (like HTTP) but designed for constrained devices
• Built on UDP
• Primarily one-to-one (client/server)
XMPP (Jabber)
• Mostly data is going between distant, mostly unrelated points (like IM)
• Tends to be for consumer apps
23. Session title
23
MORE IOT MESSAGING PROTOCOLS
Data Distribution Service (DDS)
• Direct device-to-device bus communication
• Data-centric middleware standard with roots in high-
performance defense, industrial, and embedded applications.
• Can efficiently deliver millions of messages per second to many
simultaneous receivers
Advanced Message Queuing Protocol (AMQP)
• Strictly reliable point-to-point messages between servers using
TCP
• The standard for high-performance messaging in finance
• In IoT context, AMQP most appropriate for the control tier or
server-based analysis functions
24. Session title
24
IOT: STRINGENT TECH CHALLENGES FOR IT
Scalability
• IDC expects the global installed base of IoT endpoints to reach 212
billion connected devices by 2020 (Source: Worldwide Internet of Things
(IoT) 2013–2020 Forecast: Billions of Things, Trillions of Dollars, IDC,
October 2013)
• An individual intelligent system might gather and analyze billions of data
objects from millions of distinct endpoints
25. Session title
25
IOT: STRINGENT TECH CHALLENGES FOR IT
Reliability
• Many intelligent systems will be employed for mission-critical applications
where system downtime can result in diminished productivity, dissatisfied
customers, or lost revenue
• Some intelligent systems will be used in safety-critical applications where
system downtime can lead to loss of life or property or cause significant
environmental or health hazards
• Examples: emergency services, process control, real-time systems
control, health applications
26. Session title
26
IOT: STRINGENT TECH CHALLENGES FOR IT
Security
• Inherent exposure to public cloud and network services
• Security systems and practices must be extended to protect
against data loss, service theft, and increasingly sophisticated
denial-of-service attacks in a scalable manner
• Support cohesive authentication, authorization, and auditing
capabilities to establish trust, govern access to resources, and
ensure compliance
• Strong encryption often needed
27. Session title
27
THREE TIER ARCHITECTURE FOR IOT: DEVICE
• Wide variety of intelligent endpoints
(including relatively traditional client
computers/OSs)
• Standards-based wired and wireless
networking protocols for connectivity
• Standards-based data transport and
messaging mechanisms are used to forward
raw data and exchange control information
• Up to 100s of thousands or even millions
28. Session title
28
THREE TIER ARCHITECTURE FOR IOT: CONTROLLER
• Intermediary that facilitates communications,
offloads processing functions, and drives
action
• Analyzes tactical data and executes business
rules and issues control information
downstream
• Collects, summarizes, and stores device data
and forwards it upstream
• Consider “data gravity” and network
bandwidth/latency effects
• Thousands of instances
29. Session title
29
THREE TIER IOT ARCHITECTURE IOT: DATACENTER/CLOUD
• Large-scale data computation and repository
for data storage and strategic analysis
• Virtual environment for instantiating
applications, storing data, and executing
complex event processing, distributed
computing, and business analytics functions
• Collects and examines aggregated data from
the control tier and disseminates business
rules downstream
• Hundreds of instances
31. Session title
31
SUMMING UP
• IOT isn’t a single thing and the details vary a
lot
• Different IOT examples are at different
maturity points
• Some may never mature—technology, use
case, cost
• Data architectures and approaches matter
• Intelligent systems architectures matter
32. Session title
32
ABOUT ME
• Red Hat Cloud Product Strategy
• Twitter: @ghaff
• Google+: Gordon Haff
• Email: ghaff@redhat.com
• Blog: http://bitmason.blogspot.com
• Formerly: Illuminata (industry analyst), Data
General (minicomputers/Unix/NUMA/etc.)