WELCOME WILL IT DOMO? MQTT TRANSACTIONS TO GET ANY IOT DATA INTO - - PowerPoint PPT Presentation

welcome will it domo
SMART_READER_LITE
LIVE PREVIEW

WELCOME WILL IT DOMO? MQTT TRANSACTIONS TO GET ANY IOT DATA INTO - - PowerPoint PPT Presentation

WELCOME WILL IT DOMO? MQTT TRANSACTIONS TO GET ANY IOT DATA INTO DOMO Ryan Wells 3 WILL IT DOMO? Ryan Wells Jared Schumacher Trey Hoffman Vice President of Information Enterprise Systems Enterprise Applications Technology Manager


slide-1
SLIDE 1
slide-2
SLIDE 2

WELCOME

slide-3
SLIDE 3

WILL IT DOMO?

MQTT TRANSACTIONS TO GET ANY IOT DATA INTO DOMO Ryan Wells

3

slide-4
SLIDE 4

WILL IT DOMO?

Ryan Wells

Vice President of Information Technology

Angelica

Trey Hoffman

Enterprise Applications Manager

Angelica

Jared Schumacher

Enterprise Systems Manager

Angelica

4

slide-5
SLIDE 5

5

slide-6
SLIDE 6

WHAT WE WERE TOLD ABOUT IOT

  • That it stands for “Internet of Things”
  • It was going to be “the future”
  • Thing “X” will talk to thing “Y” and that will

somehow order milk, call your mom, lower your blood pressure, etc.

6

slide-7
SLIDE 7

A NARRATIVE REALITY OF IOT

  • I have an individual app for my water meter, my electrical

breaker box, my light switches, my Amazon Echo, my thermostat, my camera system, my bathroom scale and my

  • fridge. Which, by the way…won’t tell me if the door was left
  • pen
  • What I really want to know is:
  • Is my 2008 GE Freezer…well, freezing?
  • If not, can something notify me?
  • A new smart freezer is 5x the cost of a regular one; worth it?
  • Which member of my household keeps leaving the lights on?

My Light app doesn’t talk to my camera or location app.

7

slide-8
SLIDE 8

THE REALITY OF IOT - SYNTHESIS

  • The destination of IOT data is usually proprietary
  • The source of the IOT data must be purpose built
  • There is no standard for IOT communications

supported by multiple vendors

  • If you’re lucky, you get an API that you have to

develop custom code to use

  • Then you get to set up a server and database to

collect and synthesize the information before sending to a reporting system

8

slide-9
SLIDE 9

I REJECT YOUR REALITY AND SUBSTITUTE MY OWN

  • Make or retrofit IOT devices inexpensively
  • Use common communication

protocols/languages

  • Interface those devices into a single,

common output

9

slide-10
SLIDE 10

THE BUSINESS “ASK”

  • Find a low cost way to collect cycle data from

legacy linen equipment (the “thing”)

  • Use that same mechanism to also collect
  • perator efficiency, machine effectivity, quality,

temperature and humidity information

  • Synthesize and compare the information

without having to build another database

10

slide-11
SLIDE 11

CONSTRAINTS

  • There was not a repeatable way to get IOT data

into a reporting system

  • Vendors that specialized in this field, would only

provide data to their own platform or “API coming spring two- thousand and never”

  • Many machines would have to be completely

replaced to gain IOT functionality and recurring costs were significant

11

slide-12
SLIDE 12

PARALLEL TECHNOLOGY – THE AHA! MOMENT

  • Many of us had been doing machine integrations in our homes

for years, so we knew that our vision was technically possible

  • Some parallel systems
  • Temperature from hot tub
  • Temp and humidity of legacy deep freeze
  • Lights and other switches in our homes
  • All these devices, natively speak MQTT protocol and run

common hardware (ESP8266)

  • We could potentially parse this output and ingest it into DOMO

12

slide-13
SLIDE 13

IOT HOT TUB: WILL IT DOMO?

13

slide-14
SLIDE 14

IOT FREEZER: WILL IT DOMO?

14

slide-15
SLIDE 15

IOT LIGHTS: WILL IT DOMO?

15

slide-16
SLIDE 16

WE DON’T CELEBRATE MEDIOCRITY

  • The Manager: “This is all really great but how to I

combine this data into something meaningful?”

  • The Developer: “I see the data outputs, and I can

ingest them into DOMO but that will require development work!”

  • The Infrastructure team: “It will require another

server and database, don’t you have enough already?”

16

slide-17
SLIDE 17

The Missing Link

slide-18
SLIDE 18

MQTT

  • MQTT – Message Queuing Telemetry

Transport

  • MQTT is a machine-to-machine (M2M)/"Internet of Things"

connectivity protocol. It was designed as an extremely lightweight publish/subscribe messaging transport

  • It uses a broker to handle transactions. It does not require that

thing “X” trust or even know about thing “Y” and vice versa

  • It is fault tolerant, messages do not leave queue until they’re

consumed by a subscriber to their specific “topic”

18

slide-19
SLIDE 19

GETTING THE DATA TO DOMO

Domo developed a direct-connect MQTT Broker We changed the outputs of the MQTT devices to point directly at DOMO, tied it to a dataflow, created some cards and dropped them in a page

19

slide-20
SLIDE 20

LINEN MACHINE: WILL IT DOMO?

Technically, a linen machine is a heap of metal, switches and temperature sensors This was no different than what we’d already proven to work

20

  • Project “Tempo” was born
  • Initial prototype was built by Jared and I
  • We included some additional hardware

to provide real-time information

  • Initial pilot was done
  • Install was 10% cost of the next best
  • ption
  • The board asked us to make 400 more
  • The question ceased to be “Will it

DOMO?” and became “Will it Scale?”

  • Additional production was outsourced
  • Total cost delivered is $210/each
slide-21
SLIDE 21

LINEN MACHINE: WILL IT DOMO?

21

  • The Operator Interface
  • Touch Screen on

Raspberry pi

  • Barcode Scanner
  • Information
  • Current pace
  • 5 Min average
  • Quality check

system

  • Machine problem

“Andon” tools

slide-22
SLIDE 22

LINEN MACHINE: WILL IT DOMO?

Would we be talking about it if it didn’t?

22

slide-23
SLIDE 23

KEY TAKEAWAYS FROM TEMPO

23

  • Our Tempo implementation is more complex than needed for

many applications and might make this technology seem overly complex

  • If there is a way to measure qualities in the real world, it will

DOMO

  • Many different hardware/firmware platforms can be used

(Nodemcu, Sonoff, Raspberry Pi, Tasmota, ESPhome, ESPeasy, MQTTpubsub)

  • If it can speak MQTT, you can get the data into DOMO.
slide-24
SLIDE 24

COTTAGE CHEESE: WILL IT DOMO?

24

  • Let’s domo something together – Cottage Cheese
  • Why cottage cheese?
  • Because that’s weird
  • It’s “hard” to DOMO
  • I happen to have some right here
slide-25
SLIDE 25

COTTAGE CHEESE: WILL IT DOMO?

25

  • What do we want to measure?
  • Weight? - HX711 sensor
  • Moisture? I2C sensor
  • Temperature? AM2301, DHT11
  • Hits with a hammer? load cell, binary switch
  • For simplicity, let’s go with Temperature and Hammer

strikes

  • Temperature is an analog signal (converted to digital)
  • Hammer strikes are a digital signal
slide-26
SLIDE 26

COTTAGE CHEESE: WILL IT DOMO?

26

  • Buy a Node MCU and a temp sensor (DS18B20) - amazon <$20
  • Buy some breadboard jumper wires - amazon - get a kit of M-F, M-

M, F-F <$6

  • Buy a resistor kit – amazon – will last your lifetime <$12
  • Download Arduino IDE – https://www.arduino.cc
  • Download the Tasmota firmware package

https://github.com/arendst/Tasmota

slide-27
SLIDE 27

COTTAGE CHEESE: WILL IT DOMO?

27

  • Open sonoff.ino in the

Arduino IDE

  • Edit my_user_config.h to set

WIFI, Topic, MQTT and IP information

slide-28
SLIDE 28

COTTAGE CHEESE: WILL IT DOMO?

28

  • Plug the

NodeMCU into a USB port

  • Change the

board and firmware info as shown

  • Click checkbox to

compile and upload the firmware to the NodeMCU

slide-29
SLIDE 29

COTTAGE CHEESE: WILL IT DOMO?

29

  • Unplug the NodeMCU to wire it
  • Wire a switch between GND and

D5

  • Wire sensor signal (yellow) to D0
  • Wire sensor power (red) to 3v3
  • Wire sensor Ground (black) to

GND

  • The pin numbers will become

clear when we look at the software interface

  • If you think this is too complicated,

I’ll introduce you to the 12 year old girl that did this one

slide-30
SLIDE 30

COTTAGE CHEESE: WILL IT DOMO?

30

  • Plug in the

NodeMCU to USB for power and configure

  • Test it
  • Wait for

DOMO to collect data

  • Visualize it
slide-31
SLIDE 31

DEMONSTRATION

  • Node MCU
  • Connection to Broker
  • Temperature
  • Binary Switch

31

slide-32
SLIDE 32

WE DON’T CELEBRATE MEDIOCRITY - SOLVED

  • The Manager: “This is all really great but how to I combine this

data into something meaningful?” – INJEST INTO DOMO

  • The Developer: “I see the data outputs, and I can ingest them

into DOMO but that will require development work!” – USE MQTT INSTEAD OF CUSTOM APPS

  • The Infrastructure team: “It will require another server and

database, don’t you have enough already?” – ALL DATA AND INFRASTRUCTURE EXISTS AT DOMO

32

slide-33
SLIDE 33

THANK YOU