Loading presentation...

Present Remotely

Send the link below via email or IM

Copy

Present to your audience

Start remote presentation

  • Invited audience members will follow you as you navigate and present
  • People invited to a presentation do not need a Prezi account
  • This link expires 10 minutes after you close the presentation
  • A maximum of 30 users can follow your presentation
  • Learn more about this feature in our knowledge base article

Do you really want to delete this prezi?

Neither you, nor the coeditors you shared it with will be able to recover it again.

DeleteCancel

Make your likes visible on Facebook?

Connect your Facebook account to Prezi and let your likes appear on your timeline.
You can change this under Settings & Account at any time.

No, thanks

IoT - HOWTO Drink From a Firehose

No description
by

simon Lemay

on 9 April 2015

Comments (0)

Please log in to add your comment.

Report abuse

Transcript of IoT - HOWTO Drink From a Firehose

IoT - HOWTO Drink
From a Fire hose

Welcome
Simon Lemay -
John Schloman -
Melvin Ramos -
This is not About
How to connected your toilet to your refrigerator
A technical in depth evaluation of a specific technology
What we make at our Company
Selling you technologies we like
This
is About Data

IP Traffic 2004-2016
The Fire hose
Keep Calm and Reread the Spec
How to deal with the Exchange of Data in the IoT
How to Aggregate Data in the IoT
So, this Talk is About Data
The Qualities of
an IOT platform
Interconnectivity
Interoperability
Responsiveness
Ownership
The way a device connect should not matter in the grander scheme of things when data are being aggregated, it as to be treated as data in a generic form

The device type does not matter when information is shared

There is no place for proprietary protocol

Security, Authorization and Access

This is both physical and conceptual

A device like a sensor could be provide access to many users in many ways simultaneously for different periods of time

Objects can own objects

Devices can be claimed if unowned, and this can be done in a secure fashion
Device type doesn't matter on all sides of a exchange. Both smart and constrained devices (e.g, $0.05 thermometer, legacy* hardware).

Unified object model

Headless understanding of operations
Low bandwidth, low power, low processing, battery-friendly

Respectful in both requests and handling responses (i.e., don't spam the device on GSM)
Exchange of
Data

http://xkcd.com/927/
IoT Standards?
Standards
Exchange Data
Device to Device
Device to Cloud
Human to Device
Many Example
Communication
CoAP
MQTT
REST
DDS
Management,
application and
Object model
LwM2M
SNMP
OMA DM
Aggregate the Generated Data
IoT + BigData = Fog Computing
What is the trend of my sensors over
time?

What is the current status of my device?
Operational cost of my devices vs. maintenance?

Can I afford to have my process be affected of a broken device?

Is my device due for maintenance?
Business Questions
How much is each device consuming on overall data bandwidth? If its mobile can we reliably tell that we are not excessively depleting it?

What is the trend of device in terms of power usage?

Is it taking long processing a command? Is the device able to send sensors in timely manner? What is the average time it took to execute a simple command? How about the total network round trip?

Technical Questions
Logistics BigData Concept
Composition of Potential Data
A typical device can send “ping” or heartbeat [H] message every 20-30
seconds.

It can send send a sensor [S] information as soon as its available (event) or
periodic. 30-60 seconds.

A device owner can send messages to a device arbitrarily, but it can also be
set as a timer.

Upward Minute Trend = No. of Device * (2*[H] + [S])

Downward Minute Trend = No. of Device * X Commands

Hours? Days? Weeks? Quarterly? Or Yearly?

Intelligence Empowers!
http://en.paperblog.com/fellow-travelers-big-data-and-the-internet-of-things-639289/
Compress and aggregate
Compress the data, by aggregating data as
soon as data is received.

Represent the data on aggregated data buckets, and split
granularities overtime.

Days = 24 Hours [24 hours is the granularity]

Group the data logically if possible.

Use a NoSQL database to store your data.

Devices can learn common commands to be executed base on time.

A Cloud application can define maintenance window where devices are least used.

Devices can be segmented and group together by purpose or data.

Summary of device activity for Month, Quarter or Year.

Device maintenance schedule and/or usage.

Settings and status of device on month-to-month.

Historical trends of device activity.

Visualization and logical grouping of devices
Reporting
Machine Learning
Questions?
Thank You!
We are also hiring talented developers please approach us if you are interested...
With the number of devices projected, we cannot achieve the end goal communicating through silos.
Era of Collaboration
The strength is to be able to exchange data freely to offer ease access to the IoT.
Data gains a lot of value when mixed with the ecosystem.
How are They Helping
Put all device on the same level
Offer a comprehensive unified transaction model
Offer an extensible Data and object model
Many IoT standards already exists
Favor a healthy exchange of data
Very Portable
Low Bandwidth
Low CPU consumption
Offer alternative communication semantics for more constraint device
Extensible
Open
Defines a norm
Can you imagine having a different power plug per appliance manufacturer
Interface between production phase
Interface between producer
Focus on core business
Don't want to be the "orphan device"
...Data?
Source: http://www.discovery.org/a/3869
"Things will exchange data"
In 2004 monthly global Internet traffic
passed 1 exabyte.


Cisco Visual Networking Index estimates 2016 monthly traffic will hit 110.3 exabytes...
Here we are going to start adding those 50-75 Billion IoT devices.
Simon Lemay
Twitter: @si_lem
email: slemay@zebra.com
John Schloman
Twitter: @jfschl
email: jschloman@zebra.com
Melvin Ramos
Twitter: @mramosjcb21
Email: mramos@zebra.com
follow us on twitter
"Greenfield" is probably more dangerous right now than "Peak Internet"

A lot has been learned in the last
decade about what data at scale means.
Those same lessons will get us through the next one.

Technical Lead
Senior Software Engineer &
Device stack Manager
Senior Software Engineer &
Big Data Manager
Full transcript