Reply Bootcamp Rome - Mastering AWS - IoT Bootcamp

AWS IOT
BOOTCAMP
MASTERING AMAZON
WEB SERVICES
Emiliano Pecis, Andrea Mercanti, Stefano Rosini, Gabriele Stella
Storm Reply

MASTERING AWS SERIES
AGENDA
AWS AI as a Service
AWS IoT Foundations
AWS Database-as-a-service
AWS IoT Advanced
Bootcamp "Mastering AWS IoT»
October 27th in Rome

AWS IOT BOOTCAMP
OVERVIEW
AWS IOT
PLATFORM
IoT
topic
IoT
MQTT
protocol
IoT
shadow
IoT
certificate
SNS
IoT
rule
IoT
policy
Elastic Search
KinesisLambda
Machine
Learning
DynamoDB
Grove Sensors
Raspberry Pi + Grove Board
AWS IoT Button Python IoT SDK

AGENDA
AWS IoT
LAB 1-2: IoT HighTemp Alarm
Lunch Break
LAB 3: Streaming IoT
Coffee Break
LAB 4: Predictive Maintenance
Wrap-up
09:00 – 10:00
10:00 – 13:00
13:00 – 14:00
14:00 – 16:30
16:30 – 16:45
16:45 – 17:50
17:50 – 18:00

STORM REPLY
THE REPLY’S COMPANY FOCUSED ON AWS
70 Experts
CloudArchitects, DevOps, Microservices Architects
50 Top Brand Customers
Enel, Vodafone, Ferrero, Volkswagen, ePrimo, Grohe, etc
AWS Premier Consulting Partner
The only Italian company to have the highest level of
certification

STORM REPLY
THE REPLY’S COMPANY FOCUSED ON AWS
Design Services
Design and development of
solutions in both traditional
environment, supplying them
as a service-in-the-cloud
(providing and managing
IaaS), and directly on the
cloud, using service
architecture (PaaS) and cloud
native applications
Advisory Services
Supply of consulting services
for IaaS, SaaS & PaaS
solutions: analysis, setup,
development, customization,
maintenance and, above all,
migration from on premises to
the cloud. Storm Reply is able
to provide both a vision and a
strategy for the cloud
Managed Services
Storm Reply provides Managed
Services on top of Cloud
Services: System Monitoring,
Security Monitoring, System
Management, Patch & Change
Management. Everything
customers need to manage and
control can be done by Storm
Reply

WHAT IS AWS?
AMAZON WEB SERVICES

AWS GLOBAL INFRASTRUCTURE
AWS CLOUD OPERATES 44 AVAILABILITY ZONES WITHIN 16
GEOGRAPHIC REGIONS AROUND THE WORLD
EUROPE
Ireland
Frankfurt
London
Paris
Stockholm

ARTIFICIAL INTELLIGENCE ON AWS
LEARN, SEE, SPEAK, WRITE
MACHINE
LEARNING
REKOGNITION POLLY LEX

DATABASES ON AWS
STORE, ORDER, RETRIEVE, MIGRATE
RDS DYNAMODB ELASTICACHE REDSHIFT
DATABASE
MIGRATION
SERVICE

INTERNET OF THINGS ON AWS
HEAR, FEEL, CONTROL, COMMUNICATE
IOT GREENGRASS

IOT INTRODUCTION
WHAT IS IOT?
“The Internet of Things
(IoT) is a system of
interrelated computing
devices, mechanical and
digital machines, objects
with the ability to transfer
data over a network
without requiring human
interaction“

IOT INTRODUCTION
“A thing, in the Internet
of Things, can be a
person with a heart
monitor implant, a farm
animal with a biochip
transponder, an
automobile or any other
natural or man-made
object that can be
connected.“
WHERE IS IOT?

IOT INTRODUCTION
“Cloud computing and the IoT have a complimentary
relationship. The IoT generates massive amounts of data,
and cloud computing provides all the computational
resources to store and elaborate this big amount of data“
WHY IOT AND CLOUD?

DEVICE SDK
Set of client libraries to
connect, authenticate and
exchange messages
DEVICE GATEWAY
Communicate with devices via
MQTT and HTTP
AUTHENTICATION
AUTHORIZATION
Secure with mutual
authentication and encryption
RULES ENGINE
Transform messages
based on rules and
route to AWS Services
AWS Services
- - - - -
3P Services
DEVICE SHADOW
Persistent thing state
during intermittent
connections
APPLICATIONS
DEVICE REGISTRY
Identity and Management of
your things
AWS IOT PLATFORM OVERVIEW

MQTT PROTOCOL
MQ Telemetry Transport:
 Lightweight protocol designed to optimize the use of resources.
 Publish-Subscribe Topic based protocol.
 Client-Broker based protocol.
OVERVIEW
Note:
“MQ” does not stand for “Message Queue”. It comes from “MQseries”, the product
name of IBM original project.

MQTT PROTOCOL
FEATURES OVERVIEW – 01 0F 02
Defined Methods
 Connect: connect to the broker.
 Disconnect: disconnect from the
broker.
 Subscribe: subscribe to a topic.
 Unsubscribe: unsubscribe from a topic.
 Publish: publish a message on a topic.
Additional Features
 Keep Alive: keep the connection to the
broker alive
 Session Persistence: restore session
and topic subscription on a reconnection

MQTT PROTOCOL
FEATURES OVERVIEW – 02 0F 02
Quality of Service
 QoS Level 0: At most once.
It guarantees a best effort delivery.
 QoS Level 1: At least once.
It guarantees that the message will be
delivered at least once.
 Qos Level 2: Exactly once.
It guarantees that the message will be
delivered only once.
Topic Wildcards
 Topic path.
home/livingroom/groundfloor/temperature
 “+”: Single-level wildcard.
home/+/groundfloor/temperature
 “#”: Multi-level wildcard.
home/#
 “$”: Topic reserved for the broker
$aws/things/groudfloor/shadow/get

DEVICE SDK
exchange messages
AUTHENTICATION
AUTHORIZATION
Secure with mutual
RULES ENGINE
Transform messages
based on rules and
AWS Services
- - - - -
3P Services
DEVICE SHADOW
during intermittent
connections
APPLICATIONS
DEVICE REGISTRY
your things
AWS IOT DEVICE GATEWAY
DEVICE GATEWAY
MQTT and HTTP

DEVICE GATEWAY
Powerful Message Broker
 Standard Protocol (No Lock-In)
MQTT, HTTP, WebSockets
 Elastic Pub-Sub Brocker
from one to one bilion of connections
 Long Live connections
support for long-live connection
 Secure by default
x509 Certs, TLS 1.2, Client Mutual Auth
AWS IOT MESSAGE BROKER
Subscribers
Publishers

DEVICE SDK
exchange messages
AUTHENTICATION
AUTHORIZATION
Secure with mutual
DEVICE GATEWAY
MQTT and HTTP
AWS Services
- - - - -
3P Services
DEVICE SHADOW
during intermittent
connections
APPLICATIONS
DEVICE REGISTRY
your things
AWS IOT RULES ENGINE
RULES ENGINE
Transform messages
based on rules and

Select DATA from TOPIC where FILTER
SQL LIKE CODE
Functions:
 String manipulation with regex
 Mathematical operations
 Context-based helper function
 Crypto support
 UUID, timestamp, rand, etc.
Json Collection:
 get(array, int) - get item at index
 get(string, int) - get char at position
 get(object, key) - get value of key
Nested Query:
 Select (select v from e) From …

SEND DATA TO OTHER SYSTEMS

DEVICE SDK
exchange messages
AUTHENTICATION
AUTHORIZATION
Secure with mutual
RULES ENGINE
Transform messages
based on rules and
AWS Services
- - - - -
3P Services
APPLICATIONS
DEVICE SHADOW
during intermittent
connections
DEVICE GATEWAY
MQTT and HTTP
AWS IOT REGISTRY
DEVICE REGISTRY
your things

AWS IOT REGISTRY
Devices connected to AWS IoT are
represented by things in the thing registry
Thing
 Representation of a specific IoT Device.
Thing Type
 Representation of a family of IoT Devices.
STATIC REPRESENTATION OF THINGS
Thing Example
{
"version": 3,
"thingName": "MyLightBulb”,
"thingTypeName": "StopLight",
"attributes": {
"model": "123",
"wattage": "75"
}
}

DEVICE SDK
exchange messages
RULES ENGINE
Transform messages
based on rules and
DEVICE REGISTRY
your things
AWS Services
- - - - -
3P Services
APPLICATIONS
DEVICE SHADOW
during intermittent
connections
DEVICE GATEWAY
MQTT and HTTP
AWS IOT SECURITY
AUTHENTICATION
AUTHORIZATION
Secure with mutual

AWS IOT SECURITY
STRONG THING IDENTITY
Mutual Auth TLS
 One-Click Certificate Creation
 Bring Your Own CSR
 Bring Your Own CA

DEVICE SDK
exchange messages
AUTHENTICATION
AUTHORIZATION
Secure with mutual
RULES ENGINE
Transform messages
based on rules and
AWS Services
- - - - -
3P Services
APPLICATIONS
DEVICE REGISTRY
your things
DEVICE GATEWAY
MQTT and HTTP
AWS IOT DEVICE SHADOW
DEVICE SHADOW
during intermittent
connections

DYNAMIC REPRESENTATION OF THINGS
{
"state" : {
"desired" : {
"lights": { "color": "RED" },
"engine" : "ON”
},
"reported" : {
"lights" : { "color": ”BLUE" },
"engine" : "ON”
},
"delta" : {
"lights" : { "color": "RED" }
}
}
DEVICE
Report its current state to one or multiple
shadows and retrieve its desired state
from shadow.
SHADOW
Shadow reports delta, desired, and
reported states.

DEVICE SHADOW UPDATE FLOW - STEP 07 0F 07

AWS IOT PRICING
 Pay as you go, no minimum fees.
 5$ per million messages published or
delivered.
 Free Tier of 250K messages per month
free
(for the first 12 month)
PRICING MODEL
AWS IOT
PLATFORM

AWS IOT BOOTCAMP – LAB 1
OVERVIEW
W E L C O M E T O
T H E S T O R M
S M A R T H O M E

OVERVIEW
“We want to Configure an alarm from different
temperature and humidity sensors placed in every room,
to alert us when the temperature of a certain place goes
over a configurable threshold. A buzzer will inform us if
we are close and a message will be sent with a
notification to our device“

OVERVIEW

OVERVIEW
AWS IOT
PLATFORM
IoT
topic
IoT
MQTT
protocol
IoT
certificate
SNS
IoT
rule
IoT
policy
Grove Sensors

LAB GOALS
1. Assemble the IoT Demo Architecture
2. Familiarize with the AWS IoT Console
3. Create the Thing on AWS IoT
4. Configure the Raspberry Pi
5. Test the environment
6. Understand the Python Code
7. Create Rule on AWS IoT Rules Engine
8. Create SNS Topic
9. Make it Work!

1. ASSEMBLE THE IOT DEMO ARCHITECTURE
1. Assemble the Grove Sensors as
follows:
• A1: TEMP & HUMIDITY SENSOR
• D2: RED LED
• D3: BUZZER
• D4: BUTTON
2. Connect the GrovePi+ to the
Raspberry Pi
3. Configure your LAN interface
• 192.168.0.100/24
4. Ping the Raspberry via LAN cable
• 192.168.0.X0 (X IS YOUR GROUP)

2. FAMILIARIZE WITH THE AWS IOT CONSOLE
1. kj
https://stormroma.signin.aws.amazon.com/console - Users have been provided to each group
Select AWS
IoT service
Go to
Ireland
regionIoT usage
dashboard
Quickstarts
Manage things
and types
Certificates
and policies
Manage rules
MQTT Client
test

1. Click on Manage  Things
2. Thing Name
• 000x (x = group id)
3. Thing Type
• Raspberry_pi_bootcamp
4. Thing Attributes
• DEVICE_ID
• GROUP
5. Create Thing!
3. CREATE THE THING ON AWS IOT

6. Click on Security tab
7. Create Certificate

8. Download:
• CERTIFICATE
• PRIVATE KEY
• PUBLIC KEY
9. Rename the Files in (X = group id):
• 000x.cert.pem
• 000x.private.key
• 000x.public.key
10. Activate the certificate
11. Click on Attach Policy

12. Copy the number of the Certificate
13. Attach Policy
• IOT-POLICY
14. CLICK DONE!
This is a very
open policy!
Usually, you
would provide
more rescrictive
rules

4. CONFIGURE THE RASPBERRY PI
1. SSH Connect to your Raspberry using the LAN cable
• IP: 192.168.0.X
• USER: pi
• PASSWORD: raspberry
2. Copy certificate, public & private key to:
• /home/pi/bootcamp_iot/certs/
3. Set Date On Raspberry (Cause It Doesn’t Have A HW
Clock)
• sudo date -s "27 OCT 2017 10:00:00"

5. TEST THE ENVIRONMENT
1. Go back to the Console
2. Click on Test
3. Subscribe to topic
• iot/000x/# (# is the wildcard)

4. SSH Connect to your Raspberry using the LAN cable
• IP: 192.168.0.X
• USER: pi
• PASSWORD: raspberry
5. Navigate to /home/pi/bootcamp_iot/
6. Start the Publish Service
• ./start.sh 000X (x is your group)

7. After few seconds, Check the
presence of MQTT messages
published in the topics:
8. iot/000X/dht
• DATA ABOUT TEMPERATURE
AND HUMIDITY
9. iot/000X/stats
• STATS ABOUT YOUR DEVICE

6. UNDERSTAND THE PYTHON CODE
• It uses the AWS IoT Python SDK
• Create the AWS IoT Endpoint
o Publish in 3 Topic: iot/000x/dht, stats, button
o Subscribe in 1 Topic: iot/000x/listen
o Define Actions for interact with sensors
• BUZZER
• BUTTON
• BLINKREDLED

1. To test the logic behind the Python code, go to Test and try to publish a message to the topic:
• iot/000x/listen
• «action»: «BLINKREDLED» The raspberry is subscribed to the topic
/listen and will react to your message!

We will now learn how to create a
Rule to interact with the sensors
2. Subscribe to Topic iot/000x/button
3. Go to Act and Create Rule to Republish
the Topic to iot/000x/listen
4. Push the Button!

7. CREATE RULE ON AWS IOT RULES ENGINE
1. Go back to the Console
2. Click on Act  Create
• NAME: Tempalarmgroup6
• ACTION: REPUBLISH TO iot/000x/listen
• STATEMENT:

7. CREATE RULE ON AWS IOT RULES ENGINE
BONUS PART
1. Click on «Add Action»
• TYPE: SNS
Note: Test this feature
only if you’re on-time!

8. CREATE SNS TOPIC
2. Create a new Topic specific to your group on SNS Service

8. CREATE SNS TOPIC
3. Subscribe to Topic by:
• E-MAIL (Your E-mail)
• SMS (39xxxxxx)
4. Check your e-mail and confirm the
subscription

8. CREATE SNS TOPIC
5. Configure The Action
• SNS TARHET: Your New Sns Topic
• MESSAGE FORMAT: RAW
• IAM Role: raspberry-iot-bootcamp-role
6. Add the rule
TEST THE SENSOR INCREASING THE
TEMPERATURE AND…
… PAY ATTENTION TO YOUR EARS! 

Continue to LAB 2
WELL DONE!
target
11:30

OVERVIEW PART 2
“We want to store the data gathered from the temperature
and humidity sensors, to create a dashboard with real-time
and historical data and the main metrics of our sensors.
We will use Kibana to create appealing Dashboards.“

OVERVIEW PART 2
AWS IOT
PLATFORM
IoT
topic
IoT
MQTT
protocol
IoT
certificate
IoT
rule
IoT
policy
Elastic Search
Grove Sensors

LAB GOALS
1. What is ElasticSearch
2. From IoT to ElasticSearch
3. Rules and Actions
4. Meet Kibana
5. Create your first Dashboard!

WHAT IS ELASTICSEARCH?
• Elasticsearch is a popular open-source search and analytics engine
It is often coupled with Kibana, its natural visualization tool
• Amazon Elasticsearch Service (Amazon ES) is a managed service
that makes it easy to create a domain and deploy, operate, and
scale Elasticsearch clusters in the AWS Cloud.

FROM IOT TO ELASTICSEARCH
• We need to create three different IoT Rules to manage three Topics:
• TEMPERATURE /iot/000x/dht
• HUMIDITY /iot/000x/dht
• RASPBERRY STATISTICS /iot/000x/stats
3 ELASTICSEARCH INDEXES

TEMPERATURE RULE
1. Go back to IoT Console
• NAME:
000x_iot_to_es_temperature
• ATTRIBUTE: id, timestamp,
temperature
• TOPIC FILTER: iot/000x/dht

TEMPERATURE RULE ES ACTION
3. Click on Add Action  Send
Message to Amazon
Elasticsearch
• DOMAIN NAME: bootcamp
• ENDPOINT: https://search-
bootcamp-
dubcwlclatdifvomglc3kmq3ze.eu-
west-1.es.amazonaws.com
• Id: ${newuuid()}
• Index: iot-dht
• Type: temperature
• IAM Role Name: iot-device
4. Click on Add Action

HUMIDITY RULE
5. Do the same for the other two
rules…
• NAME: 000x_iot_to_es_humidity
humidity
• TOPIC FILTER: iot/000x/dht

HUMIDITY RULE ES ACTION
Message to Amazon
Elasticsearch
bootcamp-
• ID: ${newuuid()}
• INDEX: iot-dht
• Type: humidity

RASPBERRY STATS RULE
• NAME: 000x_iot_to_es_stats
cpuPercent00, cpuPercent01,
cpuPercent02, cpuPercent03,
cpuPercentAvg, cpuFreq,
memAvail, memTotal, memUsed,
memPercent, netBytesRecv,
netBytesSent, netConnections,
bootTime
• TOPIC FILTER: iot/000x/stats

RASPBERRY STATS RULE ES ACTION
Message to Amazon
Elasticsearch
bootcamp-
• ID: ${newuuid()}
• INDEX: iot-stats
• Type: stats

MEET KIBANA
1. Open your browser and go to
URL:
• https://search-bootcamp-
dubcwlclatdifvomglc3kmq3ze.
eu-west-
1.es.amazonaws.com/_plugin/
kibana/
2. Set the time filter on the top-
right to  this_month
3. Discover: shows your
indexes
4. Visualize: browse your
charts or create new ones
5. Dashboard: aggregate your
charts and information

MEET KIBANA
6. Click on Visualize  + (Create
New Visualization)  Line
7. Select iot-dht index
8. Choose the metric for Y axis, in
our case the average
temperature
9. Choose the metric for X axis, in
our case the timestamp (Date
Histogram)
10. Click on Play Button and Voilà

CREATE YOUR FIRST DASHBOARD
• Dashboards can hold many
different charts, giving you a wide
insight about your system
• For our IoT system we can keep
track of:
• Temperature and Humidity
Changements Over Time
• Average Temperature and
Humidity Values
• Max Humidity and Temperature
Values
• Total Number of Samples in Our
Analysis

You’re almost half-way!
GOOD JOB!
target
13:00

OVERVIEW
“Now that we have raw metrics from sensors, we would
need to enrich the data with information from other
sources, like a NoSQL Database. We can do that by
adding a Kinesis stream that triggers a pre-processing
lambda function where our business logic is
implemented. We want also to use our custom IoT button
to interact with the shadow of our Raspberry“

OVERVIEW PART 3
AWS IOT
PLATFORM
IoT
topic
IoT
MQTT
protocol
IoT
shadow
IoT
certificate
IoT
rule
IoT
policy
Elastic Search
KinesisLambda
DynamoDB
AWS IoT Button
Grove Sensors
Python IoT SDK

LAB GOALS
1. What is the IoT Button
2. Configure our IoT Button
3. Configure the rule
4. Play with the Button
5. What is DynamoDB
6. Create Kinesis Delivery Stream
7. Create the IoT Rules
8. Review the Kibana Dashboard!

AWS IOT BUTTON
PROGRAMMABLE VERSION OF AMAZON DASH BUTTON

AWS IOT BUTTON
IoT Complete Hardware Device
Characteristics:
 Code in the Cloud
 Learn IoT Rules, Shadow and Topics
Features:
 Wi-Fi Connectivity
 One multi-click Button
 One RGB Led
SAMPLES WITH AND WITHOUT BULLETED TEXT
Device message:
{
"serialNumber": "GXXXXXXXXXXXXXXX",
"batteryVoltage": "mV",
"clickType": "type"
}
Click Type:
 Single
 Double
 Long

AWS IOT BUTTON
AWS IOT BUTTON EXAMPLE WORKFLOW
Private Key
& Certificate
Thing/Device
SDK
AWS Services
Execution Role PolicyPermissionPolicy Rule
AWS IoT
Action
Lambda Function
Rules
Engine
SNS Topic

To configure the IoT Button:
• Create the Thing on the IoT console and Download the certificates
• Press and hold the button until a blue blinking light appears (around 6 sec)
• The button acts as a Wi-Fi access point, so when your computer searches for Wi-Fi
networks, it will find one called Button ConfigureMe - XXX where XXX is a three-
character string generated by the button. Use your computer to connect to the
button's Wi-Fi access point
• The first time you connect to the button's Wi-Fi access point, you will be prompted for
the WPA2-PSK password. Type the last 8 characters of the device serial number
(DSN)
• You can navigate to http://192.168.0.1/index.html in a browser to configure your
button.
CONFIGURE THE IOT BUTTON

MANUAL SETUP PANEL

CONFIGURE THE RULES
1. Create a Rule for each feature of
the IoT Button
• SINGLE CLICK
• DOUBLE CLICK
• LONG CLICK
• NAME:
000x_iot_button_to_RedLedON
• ATTRIBUTE: «000x» as deviceId,
«redled» as sensor, «ON» as status
• TOPIC FILTER: iotbutton/xxxxxx
• CONDITION: clickType=«SINGLE»
• ACTION: Invoke a Lambda function

CONFIGURE THE RULES
3. Select the Lambda
function:
• NAME: bootcamp-
lambda-shadow
4. Check the Code of
the Lambda to
understand the
scope

CONFIGURE THE RULES
5. Repeat for Double and Long Click
• NAME: 000x_iot_button_to_RedLedOFF
«redled» as sensor, «OFF» as status
• CONDITION: clickType=«DOUBLE»

CONFIGURE THE RULES
function:
• NAME: bootcamp-
lambda-shadow

CONFIGURE THE RULES
• NAME: 000x_iot_button_to_Buzzer
«BUZZER» as action
• CONDITION: clickType=«LONG»

CONFIGURE THE RULES
function:
• NAME: bootcamp-
lambda-topic
10. Check the Code of
the Lambda to
understand the
scope

You’re almost done!
TAKE A BREATH
AND KEEP GOING!
target
15:00

AMAZON DYNAMODB
FAST AND FLEXIBLE NOSQL DATABASE SERVICE
Partition key Sort key

AMAZON DYNAMODB
STRUCTURE OF A TABLE
Items
PARTITION KEY
• Mandatory
• Key-value
access pattern
• Determines data
distribution
SORT KEY
• Optional
• Model 1:N
relationship
• Enables rich
query capabilities
(==, <, >, contains,..)

AMAZON DYNAMODB
EXAMPLE OF ITEMS IN MUSIC TABLE
{
TableName: "Music",
Item: {
"Artist": "No One You Know",
"SongTitle": "My Dog Spot",
"AlbumTitle":"Hey Now",
"Price": 1.98,
"Genre": "Country",
"CriticRating": 8.4
}
}
{
TableName: "Music",
Item: {
"Artist": "The Acme Band",
"SongTitle": "Still In Love",
"AlbumTitle":"The Buck Starts Here",
"Price": 2.47,
"Genre": "Rock",
"PromotionInfo": {
"RadioStationsPlaying":[
"KHCR", "KBQX", "WTNR", "WJJH"
],
"TourDates": {
"Seattle": "20150625",
"Cleveland": "20150630" },
"Rotation": "Heavy«
}
}
}
{
TableName: "Music",
Item: {
"Artist": "The Acme Band",
"SongTitle": "Look Out, World",
"AlbumTitle":"The Buck Starts Here",
"Price": 0.99,
"Genre": "Rock"
}
}

KINESIS FIREHOSE STREAM TO ES
1. Starting from our Data Source
(AWS IoT), we can create stream
with Amazon Kinesis Firehose.
2. Data can be processed with
Lambda functions before being
send to Elasticsearch
3. Elasticsearch will hold our
enriched data and show them
through Kibana

TEMPERATURE FIREHOSE STREAM
1. Select Kinesis Service in AWS
Console
2. Choose Firehose (Go to the
Firehose Console)
3. Click on Create Delivery Stream
4. Choose stream name
(bootcamp-000x-temperature-
stream)
5. Iot Service will put data directly
in our Stream so choose as
source Direct PUT or other
sources

6. Enable record transformation to
allow data enrichment
7. Choose the appropriate lambda
function bootcamp-lambda-
firehose
8. Select $LATEST version of
lambda function

9. Choose our stream destination:
Elasticsearch Cluster
 Domain: bootcamp
 Index: iot-kinesis-dht
 Type: temperature
 Retry Duration: 300

10. Setup S3 bucket that will hold
data sent to Elasticsearch in
case of failures
 Backup Mode: Failed record
Only
 Backup S3 bucket: bootcamp-
iot-kinesis-firehose
 Backup S3 Bucket Prefix:
000x/temperature/

11. Select the Role
 bootcamp-0001-temperature-
stream-role

12. Check all details of your new
Kinesis Firehose stream
13. You can jump to our data
processing lambda and to your
S3 bucket

HUMIDITY FIREHOSE STREAM
14. Repeat same steps to create
humidity stream named
bootcamp-000x-humidity-stream
15. Check details on the Detail page
to ensure everything is as
expected

STATS FIREHOSE STREAM
16. Repeat same steps to create
humidity stream named
bootcamp-000x-stats-stream
17. Check details on the Detail page
to ensure everything is as
expected

TEMPERATURE RULE FH ACTION
1. Go back to IoT Console  Click
on Acts  Select
000x_iot_to_es_temperature rule
Message to an Amazon Kinesis
firehose Stream
• STREAM NAME: bootcamp-000x-
temperature-stream
• SEPARATOR: n (newline)

RASPBERRY STATS RULE FH ACTION
on Acts  Select
000x_iot_to_es_stats rule
firehose Stream
stats-stream

HUMIDITY RULE FH ACTION
on Acts  Select
000x_iot_to_es_humidity rule
firehose Stream
humidity-stream

But you’ve not finished yet!
YOU’RE NOW AN
IOT MASTER!
target
16:30

PREDICTIVE
MAINTENANCE
- LAB 4 -

WHAT IS PREDICTIVE MAINTENANCE?
“Predictive maintenance is one of many appealing use
cases for the Internet of Things (IoT). Using sensors to
predict the health of a fleet of machines in the field can
prevent down time without conducting unnecessary
maintenance. Further, predictive maintenance allows for
maintenance to be conducted at the most cost effective
time. Allowing you to shift your operations from being
reactive to proactive.“

OVERVIEW
AWS IOT
PLATFORM
IoT
topic
IoT
MQTT
protocol
IoT
certificate
IoT
rule
IoT
policy
COLLECTED
DATA
Machine
Learning

LAB GOALS
1. Create the Datasource
2. Verify the Schema
3. Select the target to evaluate
4. Create the ML Model
5. Configure the ML model
6. Review Model Performances
7. Create the Real-time Endpoint
8. Configure Alarms and Start the Prediction!

AMAZON MACHINE LEARNING (ML)
CREATING MACHINE LEARNING (ML) MODELS WITHOUT HAVING TO
LEARN COMPLEX ML ALGORITHMS AND TECHNOLOGY.
Will the customer buy this Product?
Is this email spam?
What will the temperature be tomorrow?
Is this book a romance, thriller or adventure story?

MACHINE LEARNING PROBLEM TYPES
BINARY
CLASSIFICATION
MULTICLASS
CLASSIFICATION
REGRESSION
28

STEPS TO SOLVE A ML PROBLEM
STEP 1: PREPARE YOUR DATA AND LOADING THE DATASET
STEP 2: SPLITTING DATA INTO TRAINING AND TESTING SUBSETS
STEP 3: CREATING ML MODEL USING TRAINING SET
STEP 4: TESTING ML MODEL AGAINST THE TESTING SET
STEP 5: EVALUATING MODEL PERFORMANCE
STEP 6: USE THE ML MODEL TO GENERATE PREDICTIONS
AWS
TASKS

CREATE THE DATASOURCE

VERIFY THE SCHEMA

SELECT THE TARGET TO EVALUATE

CREATE THE MODEL FROM THE DATASOURCE

CONFIGURE THE ML MODEL

REVIEW THE PERFORMANCES OF THE MODEL

CREATE REAL-TIME ENDPOINT

CONFIGURE ALARMS ON CLOUDWATCH
aws cloudwatch put-metric-alarm --alarm-name maintenanceNotification --metric-name maintenance-status --
namespace "pm-metrics" --statistic Maximum --period 60 --evaluation-periods 1 --threshold 1 --
comparison-operator GreaterThanOrEqualToThreshold
aws cloudwatch put-metric-data --metric-name maintenance-status --namespace "pm-metrics" --value 0
To create an alarm based on our predictions, we first need to create an Amazon CloudWatch
metric, then attach an alarm to that metric to let us know when a unit is in need of maintenance.
This can be done through the console or through the AWS CLI (Command Line Inteface)
The alarm will go off if at least one of our units is predicted to need maintenance. It is possible to
configure the alarm to notify via text or email when status is set to ALARM

CONFIGURE ALARMS ON CLOUDWATCH
aws iam create-role --role-name cw_role --assume-role-policy-document file://cw_role.json
Next, we need to create a role with permission to write to Amazon CloudWatch metrics. This role
gives the AWS IoT Rules Engine rule permission to forward our AML predictions to Amazon
CloudWatch.
We create the policy and attach the role to the policy
Finally, we create the Topic Rule, to route the observations to ML Endpoint
aws iot create-topic-rule --rule-name cwAlarm --topic-rule-payload file://cw_rule.json
aws iam create-policy --policy-name cw_policy --policy-document file://cw_policy.json
aws iam attach-role-policy --role-name cw_role --policy-arn arn:aws:iam::192878198312:policy/cw_policy

REVIEW DATA ON CLOUDWATCH

AWS IOT GREENGRASS
SAMPLES WITH AND WITHOUT BULLETED TEXT
Respond to Local Events
in Near Real-time
Operate Offline Secure Comunications
Simplified Device Programming
with AWS Lambda
Reduce the Cost of
Running IoT Applications

AWS IOT GREENGRASS
GREENGRASS COMPONENTS
GreenGrass is software, not hardware.
You can use your standard hardware
Components:
 AWS GreenGrass Core
 AWS IoT Device SDK
 AWS GreenGrass Cord SDK

AWS IOT GREENGRASS
GREENGRASS CORE FEAUTRES
GreenGrass Core is responsible for:
 Local Lambda execution
 Local messaging
 Local Device Shadow
 Local Security
 Interacting with the AWS Cloud

AWS IOT GREENGRASS
GREENGRASS CORE REQUIREMENTS
GreenGrass Core requirements:
 Min 1 Single-Core 1GHz
 Min 128MB RAM
 x86 or ARM architecture
 Linux OS

AWS IOT GREENGRASS
GREENGRASS CORE REQUIREMENTS
GreenGrass Core v1.1.0
 x86_64
 ARMv8
 ARMv7
GreenGrass SDK v1.0.0:
 Python 2.7
 Java 8
 Node.js 6.10

THANK YOU
www.reply.com
STORM REPLY IT
Torino – Milano – Roma

STORM REPLY IT
Torino – Milano – Roma

Reply Bootcamp Rome - Mastering AWS - IoT Bootcamp

Reply Bootcamp Rome - Mastering AWS - IoT Bootcamp

More Related Content

Similar to Reply Bootcamp Rome - Mastering AWS - IoT Bootcamp (12)

More from Andrea Mercanti (8)

Recently uploaded (20)

Reply Bootcamp Rome - Mastering AWS - IoT Bootcamp