IoT state of the art - survey of IoT and its impact on Big Data - Presented for the Big Data Florida / Central Florida Machine Learning event at Florida Institute of Technology on Feb 13th 2017 by Karl Seiler
2. “…in retrospect it looks like
the rapid growth of the
WorldWideWeb may have
been just the trigger charge
that is now setting off the
real explosion, as things
start to use the Net.”
Neil Gershenfeld
“WhenThings Start to
Think”
3. "In the next century,
planet earth will don an
electronic skin. It will use
the Internet as a scaffold
to support and transmit
its sensations.
Neil Gross in Business
Week
5. WHAT
IS
IOT?
The Internetworking of physical devices / "connected
devices"/ "smart devices”
Embedded with electronics, software, sensors, actuators, and network
connectivity that enable these objects to collect and
exchange data
Allows objects to be sensed and/or controlled remotely
Creating opportunities for more direct integration of the physical
world into computer-based systems
Improved efficiency, accuracy and economic benefit in addition to
reduced human intervention
Expected to usher in automation in nearly all fields
6. WHY IT
MATTERS
If everything / person / place generates
fine-grained streams of
information…
More information = larger the
problems we can solve
We will see what could never be seen
before
7.
8. WHY IT
MATTERS
THE BIGTHEORY
Data on everything
Visibility over time
Patterns
Reactive optimization
Smarter = faster, cleaner, healthier, safer
11. HOW
BIG
IS
IT?
20-50
billion
connected
devices by
2020
$6trillion
spent on IoT
In next 5
years
bottom line impacts
lowering costs
increasing
productivity
new markets
GOV improving
their citizens’
quality of
life
$15
Billion spent
on smart home
2015
$2trillion
Industrial impact
by 2020
Rides on 5G,
the next
generation of
wireless
50-100
Connected
devices
per smart home
$500B/y
driverless
market
$1 trillion/y
Smart
city
market
$100B/y
smart
office
$2T/y
smart factory
30. Karl Seiler | President
PIVIT –TURN SMART
karl@piviting.com
Piviting.com
@pivitguru
Notas del editor
Welcome everyone. Thanks to our hosts Big Data Florida and the Florida Institute of Technology.
Intro self
Karl Seiler founder of PIVIT – helps companies incorporate Artificial Intelligence, Machine Learning and IoT product capabilities into product lines and business processes
Am also VP of ZONTAL a new startup focused on the Internet of things for R&D labratories
VP of Big Data Florida
Lead the Central Florida Machine Leaning Group
This a session on the state of the nation of the evolving Internet Of Things technical innovation wave and its relationship to Big Data
The most important thing I want you to leave with tonight is a better appreciation of the scope of change in our lives that IoT represents.
Certainly the Internet has changes our lives in many was, how we are informed, communicate, transact, meetup, run our businesses and so on.
One point of view is that the Internet / The World Wide Web has just been a trigger event, that is starting to set off the real explosion, now that we have things starting to use the Net.
Another insight from Neil Gross of Business Week is that our planet earth is donning an electronic skin, that will use the Internet as it scaffold to support and transmit is sensations.
So lets talk about what IoT is. It has gone by many names. It is not that new an idea.
Internet of everything
Internet appliance
The industrial internet
The smart car, smart city, smart watch, the connected fleet, the connected supply chain
Web 3.0
Or ubiquitous computing
So What is the internet of things. It is the internetworking of physical devices, connected device, and smart devices.
These devices have embedded sensors, actuators, network connectivity, in some cases built in security, local storage, local compute power.
Sometimes these devices are their own web servers. The point of the connectivity of these objects is to collect and exchange sensory data.
This in turn can allow the devices to be remotely controlled from our other computer systems connected to the larger Internet.
All with the goal to improve knowledge, efficiency, accuracy, precision, leading to less need for hands-on human intervention.
This is expected to result in new automation in nearly every field.
So if we logically follow.
That if every thing person and place generates fine-grained streams of information about the world.
This wealth of information, should empower us to work out and solver larger, subtler, harder problems.
We will see patterns that we never could before and that we never knew existed before.
Such as. Does anyone know what pattern this is a picture of.
The threads represent strings of time.
Which are overlaid on a map of an apartment floorplan.
The height of the bumps represent the frequency of occurrence.
The text above are words and strings of phrases.
This is a map of a toddler learning to speak. Word acquisition over time over space.
The Lesson here is that physical context matters matters in language acquisition, perhaps more than repetion.
So why IoT matters, in short. The big theory of its value reads like this:
Data on everything, provides more visibility into what is really happening over time, from which we detect patterns, big, sooner, better, in new ways, from which we can classify and optimize and react more quickly or more intelligently, so our new world our next future is faster, cleaner, healthier, safer.
However, of course our increased power of can just as easily be used for bad as for good.
Here is a view of the projected scale of IoT build out
It is already estimate that by the end of 2017 we will be at ~28 billion deployed IoT devices. Leading to 50 billion at work by 2020 a mere 3 years from now.
Please note that Estimates range from 20 to 50 B devices by 2020
An important point here is that this ramp up runs up steeply as we continue into the future.
Another view of the scale up. A little older. But the point here is the relative ratio of adoption. The IoT wedge ramps faster and gets bigger then any previous technology adoption curve.
Some more interesting growth and adoption data:
As we said 20-50 billion connected devices by 2020.
$6 trillion to be spent on IoT in the next 5 years
Companies are looking at IoT to lower costs, improve productivity and open new markets
Governments are looking for IoT to improve their citizens quality of life
$15 billion was already spent of smart home products in 2015
Each fully tricked-out smart home is expected to host 50-100 connected devices
The industrial impact is expected to be $2 trillion
We will need 5G wireless connectivity to proliferate for all this traffic to ride on less IoT will clog the pipes.
Projected income of $500 billion per year across the driverless vehicle markets
$1 trillion per year in the smart city market
$2 trillion per year in the smart factory market
And $100 billion per year in the smart office market
Big stuff
Lets explore the working parts of IoT. It is a layer cake.
From the bottom up.
1 - The things, devices, sensors and controllers. Low power, small, compute strong, sensitive, ubiquitous
2 – connectivity. The hop. Bluetooth, Wi-Fi, hubs, etc.
3 – The edge. Local processing and analysis for faster reaction time that can not cycle up to the cloud and back in time. The fly by wire car is the best example. Also, called fog computing.
4 – data accumulation – where does the data go, cloud, NAS, SAN, public / private models, hybrids, elastic capacity
5 – Data abstraction – your data, my data, cohorts of data, control, access, security of the stored asset, backup recovery, retention policies
6 – the application layer, trending, categorization, classification, machine learning, decision automation, control commands
7 – Collaboration and processes – integration with business processes, transactions, other systems integration across the enterprse, etc.
So lets tour some IoT market sectors.
Wearables.
We think fitbits, garmins, apple watches.
But the scope is much vaster. From business operations for identity control, event access, stock management
To biometric identification, presence / absence detection
To a myriad of healthcare monitors in hospital, clinic, ALS and home
To body augmentation
To wellness control for weight, injury, overwork, energy management, to sleep hacking
To sport and fitness performance improvement, in shoe, glove, helmet, ball, bike, track
To lifestyle AR / VR , optimized learning
To communication, voice, gesture controls
To glamour, mood clothes, ambient lights, pattern changing clothes
What’s ahead
Health monitoring earrings
GPS location tracking in your buttons
Visual assistants in your contact lens
Fingernail impacts for gestures
Power capture shoes to drive the charging of you other wearables
…and emotion detection shirts
Shifting gears to the transportation sector
The driverless car is oft thought of as the leading edge of Iot, capturing everyones attention. It is here now. Adapative cruise control, blind spot warnings, lane drift minders, auto bumper to bumper traffic hands free
But pay attention to the trucks. IoT impacts their bottom line directly. For a conservative industry they are certainly early adopters.
You will son see close convoying of tight flying formations of trucks on off hours. Save fuel to the tune 15% , making better use of the time cycle of the driver, less fatigue, rest while in transit, and less accidents.
The self-driving vehicle revolution has started and rolls out something like this:
Era 1 – out to 2025 – fully autonomous vehicles for consumers
Era 2 – out to 2040 – car insurance shifts to covering fleets and bugs, supply chain and logistics are redefined
Era 3- out to 2050 – autonomous vehicle dominate the landscape, people get 50 min / per day back, parking collapses, collisions are down by 90%, this in turn drives the adoption of autonomous robots in out sectors
Personally owned driver driven plummets
Shared-driver driven rises
Personally owned autonomous stays small
New dominance is Shared autonomous – think on-demand public transport by local fleets
Another angle is drones. Commercial activity and the legalization of drones leads to their use for
Law enforcement
Construction surveys
Rescue mission support
Firefighting by delivering fast HD imagery
Bridge and road inspections
Snow and avalance surveys
Etc
Containers get smart
They know if you are taking your pills
If the box got too hot or too cold in transport
Of if the shipping container is lost, compromised, and the stuff inside is still fresh
The smart home
Show of hands of how many have some smart home device now? Lights, door bell video, Amazon Alexa / Dots, Gggole Home,…
Smarthome features focus on entertainment, water controls, lighting, heating/cooling, security and smart appliances
I love this smart vent idea. Sold out already. Set the target temp per room and the thing opens and closes as needed. Communicates and collaberates with the smart thermostat.
Simple, smart, awesome
The smart office is also on its way.
Dynamic lighting as needed when needed
Locks
Climate control by presence / absence lead to big savings
Automatic blinds, windows
Access controls with remote and local controls. Who can get in when with biometric IDs
Smart home and smart office grow up to the even bigger savings in cities the smart big building. With energy recapture. Co-generation. Local reclamation and recycling. Close integration with autonomous transport ingress egress.
The smart building aggregates up in part to the push for smart cities. Cities start to vie with other cities to attract talent and business due to the optimized transport, low cost of energy, integrated health access systems, responsive security systems, flexibility designed in for events, etc.
So lets also explore the impact all this wonderful new smart stuff has on our approaches to big data.
IoT devices generate lots of data, from different sources, in different formats, over longer periods of time, all the time
So…its lots more data, fine-grained, streaming at us, 24x7
The implications are for massive storage
Flexible access I sync my wearables while on business in China and it is available to my wellness advisor up the street where I live
Ever increasing need for analytics for pattern detection, classification, trend analysis, alerts
Ever smarter algorithms reacting to the analysis to do somethings, warn someone, buy or sell or move something.
All this circles around and is part of the big data ecosystem needed to support iot
.
So it is a cyclic ecosystem revolving around things and their data, the massive / fast / flexible data stores and the analytics engines driving the applications, which in turn feedback to and control the devices and environments
And remember not all decision making can be done in the cloud. Decision making is needed at the edge for faster response. Also, not all data needs to round trip up to long term retention. The edge computing can refine, filter, distill and transform to only post upstairs what is really needed.
I tend the thinks over the long haul of history and technical innovation that life improves. However, fair warning. This connected world of things has its dark side. Such as:
Hackers seizing control of your stuff and hurting you, more thing watching and monitoring inevitably invades your privacy.
Like fake news we will have to be warry of fake data injected into flows to cause harm.
And all the advantages can and do tend to flow to the wealth nations / groups / people before the needy who could get the most benefit.
And it is all so damn creepy. Being sounded by mechanical eyes and ears. And smart things that go bump in the night. Rumbas are neat but sort of creepy.
And finally when everything is optimized who gets left behind. The path to ruin is short and step in this faster paced, finely tuned world.
I heard this the other day and it just seemed to fit. The utimate IoT geek of cool stuff.
So I leave you with always be yourself
Unless you can be be batman
Then always be batman