Chapter 6

IoT PLATFORM
IoT Platforms
 What is an IoT platform?

◼ An IoT platform is an application or service that provides built-in tools

and capabilities to connect every “thing” in an IoT ecosystem. By
providing functions including device lifecycle management, device
communication, data analytics, integration, and application
enablement.
◼ An IoT platform harmonizes the many moving parts that contribute to
your IoT system. An IoT platform is the foundation for building IoT
solutions that deliver value to your business, your customers, your
customers’ end users, and your partners. By enabling you to maintain
visibility, security, and control over connected assets, IoT platforms
enable you to start and scale IoT projects efficiently so you can launch
customer-centric services and remain competitive in an evolving
market environment.
 WHY USE AN IOT PLATFORM?

◼ An IoT platform is critical to building an IoT ecosystem, it simplifies IoT,

making it more secure, regardless of where you are on your IoT journey to
build smart, connected products.
◼ IoT is a complex ecosystem that spans a network of devices and software
applications touching multiple parts of the physical and digital landscapes. It
is rare for an organization to maintain in-house expertise across all the
relevant domains to build a complete set of IoT capabilities. As a result, in
the “buy versus build” debate over IoT capabilities, most enterprises see
value in buying an IoT platform to provide an out-of-the-box set of key
capabilities, on top of which the business can build differentiating logic and
applications.
 IOT PLATFORM CAPABILITIES

◼ At a basic level, IoT platforms should allow you to connect and manage
your devices with ease, offer application enablement and integration tools,
and analyze your IoT data for actionable insights.
◼ IoT Platform between things/devices and business applications
◼ connect devices to gather information
◼ secure communication with devices and applications
◼ manage devices to control their behavhior
◼ analyze data, e.g., with AI
◼ build applications, also interacting with CRM/ERP/…
 Benefits of IoT Platforms

◼ IoT platforms offer a plethora of advantages for businesses across various

sectors. Some of the benefits are as follows:
1. Real-time Monitoring
Imagine having a watchful eye over the devices and sensors round the clock.
IoT platforms enable just that, allowing businesses to swiftly detect any
issues as they arise and respond in real-time. This proactive approach
minimizes downtime and ensures uninterrupted operations.
◼ 2. Enhanced Efficiency
By leveraging automation and real-time insights, IoT platforms streamline
operations, cut down on costs, and supercharge productivity. Tasks that
once required manual intervention can now be automated, freeing up
valuable time and resources for more strategic efforts.
◼ 3.Improved Customer Experience
Picture a scenario where every interaction with the customers feels tailor-
made for their needs. IoT platforms make this a reality by enabling
personalized interactions, offering better recommendations, and providing
immediate support. This level of attentiveness cultivates customer loyalty
and satisfaction, fostering long-term relationships.
4. Data-driven Decisions
With IoT platforms, businesses gain access to a treasure trove of data that
offers invaluable insights into customer behavior and preferences. Armed
with this understanding, organizations can make informed choices, fine-tune
strategies, and outpace competitors in a constantly changing market
landscape.
◼ 5. Heightened Security
In today's digital landscape, security is paramount. IoT platforms provide
robust security measures, including realtime monitoring and rapid response
to security threats. This proactive approach helps safeguard sensitive data
and ensures the integrity of operations, instilling confidence in customers
and stakeholders alike.
6. Scalability
As businesses grow and evolve, so do their needs. IoT platforms offer the
flexibility to easily accommodate additional devices and sensors, ensuring
scalability without compromising performance or efficiency. Whether it’s
expanding operations or adapting to changing market demands, IoT
platforms provide the agility needed to stay ahead of the curve.
 What are the Key components of an IoT platform?

◼ 1. Connectivity
This layer serves as the foundation for IoT platforms, facilitating seamless
communication between devices and the internet. It encompasses protocols,
gateways, and APIs that enable devices to establish connections and
exchange data reliably.
2. Device Management
Managing the lifecycle of IoT devices is crucial for ensuring their efficient
operation. This layer handles tasks such as device onboarding, provisioning,
and maintenance. It includes features like firmware updates, remote
configuration, and monitoring to ensure devices are functioning optimally
throughout their lifecycle.
◼ 3. Data Management
As IoT devices generate vast amounts of data, effective management of this
data is essential. This layer handles tasks related to data storage, processing,
and analysis. It includes features such as data ingestion, storage
optimization, real-time processing, and advanced analytics to derive
actionable insights from the data generated by IoT devices. An IoT platform
helps businesses manage data at scale
4. Application Enablement
It provides tools and APIs for developers is essential for building
applications that interact with IoT devices and data. This layer offers
development frameworks, APIs, SDKs, and other tools that empower
developers to create innovative applications tailored to specific use cases
and business requirements.
◼ 5. Cloud Computing Services
Leveraging cloud-based services is instrumental in scaling IoT platforms
and handling the massive amounts of data generated by IoT devices. This
layer encompasses various cloud services such as storage, computing, and
analytics. It includes features like cloud storage, serverless computing, and
machine learning to efficiently process and analyse IoT data in the cloud.
6. Interoperability and integration
This layer enables diverse devices, like sensors, cameras, and wearables, to
communicate seamlessly with the IoT platform. This ensures accurate data
transmission to the cloud, enhancing scalability and fostering innovation in
the IoT ecosystem.
◼ 7. Security
By utilizing the security features inherent in IoT platforms, organizations can
effectively protect the integrity and confidentiality of data within their IoT
systems. This step includes incorporating encryption, access controls, and
authentication methods to mitigate security risks and ensure the security of
sensitive information against unauthorized access.
8. Analytics
IoT platforms often come equipped with advanced analytics capabilities,
including rule engines, event stream processing, machine learning, and data
visualization tools. These analytics functionalities empower organizations to
analyse data either at the edge or in the cloud, enabling them to derive actionable
insights. Through the utilization of analytics, organizations can make informed
decisions, enhance operations, and boost efficiency within their IoT ecosystems
◼ 9. Application Development
Enabling the development of custom applications is crucial for extending
the capabilities of IoT platforms. This layer provides developers with the
necessary tools, resources, and documentation to build and deploy
applications that interact with IoT devices and data. It includes features such
as APIs, SDKs, development environments, and testing tools to streamline
the application development process.
 HOW DOES AN IOT PLATFORM WORK?

◼ An IoT platform works by supporting the connectivity and communication

between all the components in your IoT ecosystem through one application.
It connects your devices and sensors—from registering one new device to
bulk connecting thousands of devices. It allows you to remotely manage
your devices and device groups, for example, by updating device firmware
and software.
◼ An IoT platform then facilitates data transmission and applies analytics to
the collected data. It allows you to access IoT device data at both a granular
and high-level view.
◼ Finally, an IoT platform integrates this device data with other business applications and
systems of record, so you can generate insights and make impactful decisions.
◼
 Several IoT Platforms

◼ We go a bit deeper into a selection of the many available ones:

◼ Amazon Web Services (AWS) IoT
◼ Microsoft Azure IoT Hub
◼ Mindsphere by Siemens
◼ EdgeXFoundry
◼ And several others:
◼ Google Cloud Platform
◼ ThingWorx IoT Platform
◼ IBM Watson
◼ Carriots
◼ Kaa
◼ …
Amazon Web Services (AWS) IoT
AWS IoT: Architecture
 AWS IoT: Main Components (1)

◼ AWS IoT Device SDK connects devices to AWS IoT by using the
MQTT, HTTP, or WebSockets protocols
◼ the Device SDK supports C, Java, JavaScript, Arduino, iOS, Android,…
◼ Device Gateway supports (secure) communication with devices, by
using a (1:1 or 1:n) publish/subscribe model
◼ supports MQTT, WebSockets, and HTTP 1.1 protocols
◼ Authentication and Authorization with mutual authentication and
encryption
◼ authentication with native AWS (called ‘SigV4’) as well as X.509 certificate based
authentication
◼ AWS facilitates the whole certificate process management
◼ Registry assigns a unique identity to each device
◼ also supports metadata describing capabilities of devices, e.g., whether a

sensor reports temperature and if data in Fahrenheit or Celsius

◼ Device Shadows, i.e., persistent, virtual version of devices
◼ applications or other devices exchange messages and interact through
the Shadow Device
◼ Device Shadows persist the last reported state and desired future state of
each device even when the device is offline
◼ exploit it to retrieve the last reported state of a device or set a desired
future state through the API or using the rules engine
◼ Rules Engine, supporting to build IoT applications that gather, process,
analyse, and act on data generated by connected devices
◼ evaluates inbound messages published into AWS IoT and transforms and delivers
them to another device or a cloud service, based on business rules you define
◼ a rule can apply to data from one or many devices, and it can take one or many
actions in parallel
 Azure IoT Platform

◼ Azure IoT Platform is a flexible architecture to solve industrial IoT

problems that consists of many different services and modules that can
be combined together, according to specific requirements.
Azure IoT is internally organized in 3 levels of abstraction / layers
◼ Things: data is generated from devices, has a brief processing through edge
applications and is sent trough some connectivity service to the next layer.
◼ Insights: here data really get some manipulation. Can be stored in different kinds of
storage, aggregated and analyzed through stream analytics and machine learning
tools to derive some insights.
◼ Actions: Finally those insights are visualized into graphs and dashboards that can
help companies to take some actions on their plants
Azure IoT Architecture
 Azure IoT Main Components

◼ IoT Hub: manages the communication among Azure services and in

particular with IoT devices
◼ IoT Edge: service to deploy and manage applications over remote devices
in their locations
◼ Stream analytics: one of the many analytics tools offered by Azure,
focused on real-time data and complex event processing from IoT devices
◼ Machine Learning (ML): cloud-based tool to train, deploy, and manage
different ML models with support for different languages and popular ML
libraries
◼ Blob storage: object storage solution for non-structured data, optimized for
big data
◼ Azure IoT Hub
◼ Service hosted in the cloud that acts as a message hub for bidirectional
communication between IoT devices and other modules of the application.
It allows sending both the telemetry data from devices to cloud and commands to
change devices’ behaviour from the cloud.
◼ Different features:
◼ Secure communications: complete control through a specific authentication for each
device, different types (SAS, X.509)
◼ Built-in Routing functionality with rules to automatically dispatch messages
◼ Complete integration with other Azure services to build complete industrial solutions
using among the others: Azure analytics, ML and Data Lake storage
◼ Configuration of devices using metadata and state information which are stored in the
hub and can be queried from the up-stream
◼ Azure IoT Connectivity
The connection between devices and the hub is made possible trough
Azure IoT devices SDK, which are libraries that provides the methods
to connect to the platform. Many different languages and transport protocol
are supported Languages
◼ C
◼ Java
◼ Python
◼ Protocols
◼ HTTPS
◼ AMQP
◼ MQTT
◼ Azure IoT Edge
◼ Services can run at the edge of the architecture in order to enable fast data
aggregation and analytics, running it closer to the devices and making it possible to
respond quicker to anomalies and avoid sending terabytes of data up to the cloud
Three main components:
◼ Edge Modules: containers that run Azure services and apps executed at the edge in
the device locality.
◼ Edge Runtime: environment that runs on each device and manages deployed
modules.
◼ Cloud interface: to remotely monitor the devices
◼ Edge modules
◼ Smallest unit of computation in Azure IoT Edge, implemented as Docker
compatible containers. They can contain other Azure services or specific code
written and containerized for the edge. Multiple modules can be deployed over the
same device in order to communicate and create a more complex data pipeline
◼ Every module is made of 4 conceptual elements:
◼ Image: package containing the software of the module
◼ Instance: unit of computation that runs the image on the device; it is started by IoT
Runtime.
◼ Identity: information about credentials and permissions associated with each
module.
◼ Twin: JSON document that stores metadata regarding the status of a module and
configuration
◼ Edge runtime
◼ Edge runtime is made by two modules: Edge Agent that deploys and
monitors other modules and EdgeHub that is responsible for
communication. The runtime once installed turns a device into an
◼ Edge device and gives them many features:
◼ Ensures that modules are running and report module health to the cloud for
monitoring
◼ Manages communications between edge devices and normal devices, upstream with
the cloud, and downstream with machines
◼ Siemens MindSphere Platform
◼ MindSphere is a complex PaaS to help developers build industrial IoT
solution trough their services and apps.
◼ MindSphere consists of three distinct layers:
◼ Application system: environment powered by CloudFoundry, deeply integrated with
MindSphere ecosystem and services. Deployment, scaling, and monitoring are some
of the features provided.
◼ Services: support services exposed trough REST API to make development easier.
◼ MindConnect Elements: different communication possibilities to connect machines
and plants to MindSphere regardless of the manufacturer
MindSphere Architecture
◼ MindConnect Elements
◼ MindConnect is the layer that provides
connectivity to/from Mindsphere platform.
There are both hardware and software ways
to connect:
◼ MindConnect series and Simatic edge
devices are meant to provide secure,
reliable and fast connection.
◼ They can interact with popular
industrial protocols such as OPCUA, and
MQTT or use S7 proprietary protocol.
There is always the possibility to build your
own connector using specific SDK library
for a certain protocol
◼ MindSphere Gateway
◼ MindSphere provides a
gateway service between web
application clients, agents and
edge applications, and
backend services. Routing has
to be defined in the gateway
registry from the provider of
services as well as the identity
to grant the authorization.
◼ Mindsphere IoT Services
◼ File Service: provides a basic file management tools with an interface to manage,
search and download data. Each file is linked to an asset, has some properties that
describe it and can be retrieved after a connection to the IoT gateway. APIs
provide CRUD operations on files.
◼ Time Series service: a service similar to File service, but only for the data type
time series. Time series records are linked to an asset and an aspect and consist of
a timestamp, one or more values and a quality indicator for variables.
◼ Aggregation service: creates aggregated summaries of time series data and
interfaces to read them. It is particularly useful to reduce the amount of data sent
to cloud for analysis. Once chosen a time interval to aggregate provides standard
aggregate information such as min, max, average, tot number of points
◼ Mindsphere Connectivity Services
◼ Agent management service: provides API to create, update, delete and
manage the rights of agents. Agents are the primary actors that perfom
actions on data in Mindsphere, when created they are assigned an access
token which is the key to get authenticated and autorized to communicate
◼ MindConnect API: API to send data securely from the shop floor devices
to Mindsphere. Only agents that supports HTTP processing, TLS and JSON
parsing have the requirements to communicate. It allows also to set the
configuration for both the data source from which retrieve the data and the
data model to be shared
◼ OPC UA PubSub Service: specific module that provides MQTTbased API
to agents to upload data according to OPC UA specification