What is IoT architecture?

IoT architecture comprises several IoT building blocks connected to ensure that sensor-generated
data is collected, transferred, stored, and processed in order for the actuators to perform their
designated tasks.

What are the different elements of IoT architecture?

For the sake of brevity, we will only explore the ten most important parts of an IoT architecture.

1- Sensors

IoT devices are equipped with sensors that gather data, which will be transferred over a network.
The sensors do not necessarily need to be physically attached to the equipment. In some
instances, they are remotely positioned to gather data about the closest environment to the IoT
device. Some examples of IoT devices include:

● Temperature detectors
● Smoke detectors
● Cameras and CCTVs

2- Actuators

Actuators are devices that produce motions with the aim of carrying out preprogrammed tasks,
for example:

● Smart lights turning on or off

● Smart locks opening or closing
● Thermostat increasing or decreasing the temperature

3- Gateways

Gateways serve as entry and exit points within a network. They allow for the movement of data
from devices to a network and vice versa. Internet Protocols (IP) are an example of gateways.

4- Cloud gateways

Cloud gateways are a specific type of gateways, solely made for data compression and entry
from field gateways to the cloud.

5- Control applications

Previously, we mentioned how actuators make “actions” happen. Control applications are a
medium which, through them, it’s possible to send out the relevant commands and alerts which
will make actuators function. An example of a control application could be soil sensors signaling
a dryness in the lawns, and consequently, the actuators turning on the sprinkles to start irrigation.
6- User applications

They are software components (e.g. smartphone apps) of an IoT system that allow users to
control the functioning of the IoT network. User applications allow the user to send commands,
turn the device on or off, or access other features.

7- Data lake

Not all collected data produce immediate decisions by the control applications or user apps. Data
needs to be stored for future analysis.

A data lake is a data storage space that stores all sorts of structured and non-structured data such
as images, videos, and audio, generated by IoT devices, which will then be filtered and cleaned
to be sent to a data warehouse for further use.

8- Data warehouse

For meaningful insight, data should be extracted from the data lake to the data warehouse, either
manually, or by using data warehouse automation tools. A data warehouse contains cleaned,
filtered, and mostly structured information, which is all destined for further use.

To learn more about data lakes and how they are different than data warehouses, click here.

9- Data analytics

Data analytics is the practice of finding trends and patterns within a data warehouse in order to
gain actionable insights and make data-driven decisions about business processes. After having
been laid out and visualized, data and IoT analytics tools help identify inefficiencies and work
out ways to improve the IoT ecosystem.

10- Machine learning

Machine learning gives the opportunity to create complex models for control applications. ML
models pick up on patterns in order to predict future outcomes, processes, and behavior by
making use of historical data that’s accumulated in the data warehouse. Once the applicability
and efficiency of the new models are tested and approved by data analysts, new models are
adopted.

What is a real-life example IoT architecture?

Let’s consider a simple system managing lights in a city:

1. Sensors take relevant data, such as daylight or people’s movement.

2. The lamps are equipped with actuators to switch the light on and off.
3. If the sensor stops sensing sufficient levels of light, an on-board control
application could decide to send a signal to the actuator to turn on the light.
 4. Field gateways or cloud gateways process the data from sensors and send them to
applications.
5. A city official could use the control application to
o change the threshold for turning on the light and could push this update to devices.
o place lights in neighborhoods with limited foot traffic on power-saver mode to conserve
energy.
o Feed data from nearby sensors to lights to help them make better decisions. For example,
a light with a malfunctioning sensor could be prevented from turning on in the morning.
6. A user could apply to turn lights on and off to celebrate an event via a user application.
The application could also provide information on the locations of lights, when they
would be turned on etc.
7. The data lake would store these raw data coming from the sensors.
8. An on-prem or cloud-based data warehouse could use this data to enrich a larger dataset
that includes the inhabitants’ behavior on various days of the week, energy costs, and
more.
9. Data analytics applications could be used to create visualizations for reporting or
identifying patterns. These applications could also include machine learning modules
that can detect patterns such as
o Reduction in crime with increased lighting to suggest improvements.
o The inhabitants not leaving home before 6 am and coming back not after 11 pm. The
lights could there be programmed to be on power saver mode when the streets are empty.
These smart adjustments would, reduce the need for human intervention and support a
better user experience.

What is the difference between IoT ecosystem and IoT architecture?

IoT ecosystem is the encompassing term attributed to the five general components: 1) devices, 2)
communication protocols, 3) the cloud, 4) monitoring systems, and 5) the end-user in the IoT
system.

IoT architecture is the breakdown of the inner workings of these building blocks to make the
ecosystem function.

Iot & M2M Value Chains:

Companies with ambitions to succeed in the provision of IoT need to identify the roles that they
can play in the IoT value chain and the types of partnerships that they should forge to deliver
solutions. They also need to understand who they are competing with.

What is the IoT value chain?

An IoT solution is formed of several building blocks or components, and each of these building
blocks forms part of the IoT value chain. 1 The IoT value chain illustrates how the different
components, in combination with one another or separately, add value to the overall IoT solution
and, in turn, for the end user. Furthermore, each component is developed by a range of
 companies, some of which play several roles in the IoT value chain. The following components
form part of the IoT value chain.

● Devices. This category includes existing devices such as smart meters or vehicles in which the
connectivity component has been integrated into the product design. This could also include new
devices that would not have existed without IoT, such as pet trackers. Such a device must have a
sensor and an actuator, as well as communications hardware (described in more detail below),
but it will also have other elements (for example, a power source such as a battery or mains
electricity). In addition, depending on the type of device, it may have a screen and other ways for
the user to interact with it directly (such as buttons or a keyboard).
o Sensor and actuators are connected to the device. Sensors are able to capture data from the
environment (for example, temperature). Actuators respond to instructions and make changes in
the device (for example, adjusting the temperature on a thermostat). The instructions for an
actuator can come from sensors on the same device, or from other sources (for example, a
thermostat can be activated by mobile phone while the homeowner is on their way home). A
device can have sensors, actuators, or both.
o Communications hardware enables the device to connect to the network to send the data from
the sensors to the backend systems. This can include hardware for connecting wirelessly via
Bluetooth, Wi-Fi, ZigBee, LoRa, cellular (for example GSM, 5G, NB-IoT, LTE-M) or a number
of proprietary technologies, or over a fixed network. Some devices will have hardware to
connect to multiple types of network.
● The connectivity network, which can be cellular, fixed or satellite, delivers the data from the
sensors over the internet or a private network connection to the user’s backend systems.
In addition, many different software components, which can be loosely grouped as applications,
deliver additional value to the end user.

● Backend systems include the servers to collect and analyse the data coming from the sensors
and from other sources (for example, weather forecast data). These backend systems can be
found in the public or private cloud, or on on-premises hardware. For very simple systems, the
backend can be a standard PC.
● Software platforms such as device management, security and data analytics ensure that IoT
devices are functioning correctly and have not been compromised. Such platforms also
include data analytics software to make sense of the data and improve business processes, as
well as data bases to store the data.
● This area also includes services such as billing and customer support.
Other parts of the value chain for IoT can include the systems integrators (SIs) or developers that
design, build and manage IoT services. The physical IoT device will often need to be installed
and maintained. Depending on the service, this installation process can account for a high share
of the overall value (for example, in smart metering projects, where the installation process may
cost more than the device).
 Figure: Components of the IoT value chain

Understanding the M2M value chain

A value chain is associated with the creation of value it is an instantiation of exchange by a

certain set of companies within an ecosystem. It is a useful model to explain how markets create
value and how they evolve over time. While a market space composed of only competing value
chains will eventually see the overall market value decrease (as they will compete only on price),
in an ecosystem, the value chains will complement one another. It is therefore critical to be
intentional with the moving from value chains to ecosystems.

Inputs

Inputs are the base raw ingredients that are turned into a product. Examples could be cocoa beans
for the manufacture of chocolate or data from an M2M device that will be turned into a piece of
information. Production/Manufacture: Production/Manufacture refers to the process that the raw
inputs are put through to become part of a value chain. For example, cocoa beans may be dried
and separated before being transported to overseas markets. Data from an M2M solution,
meanwhile, needs to be verified and tagged for provenance.

Processing

Processing refers to the process whereby a product is prepared for sale. For example, cocoa
beans may now be made into cocoa powder, ready for use in chocolate bars. For an M2M
solution, this refers to the aggregation of multiple data sources to create an information
component something that is ready to be combined with other data sets to make it useful for
corporate decision-making.

Packaging

Packaging refers to the process whereby a product can be branded as would be recognizable to
end-user consumers. For M2M solutions, the data will have to be combined with other
information from internal corporate databases, for example, to see whether the data received
requires any action. This data would be recognizable to the end-users that need to use the
information, either in the form of visualizations or an Excel spreadsheet.

Distribution/Marketing

This process refers to the channels to market for products. An M2M solution will have produced
an Information Product that can be used to create new knowledge within a corporate
environment examples include more detailed scheduling of maintenance based on real-world
information or improved product design due to feedback from the M2M solution. As mentioned
previously, M2M value chains are internal to one company and cover one solution. IoT Value
Chains, meanwhile, are about the use and reuse of data across value chains and across solutions.

M2M value chains are internal to one company and cover one solution. IoT Value Chains,
meanwhile, are about the use and reuse of data across value chains and across solutions.

Similarities between IoT & M2M:

The Internet of Things (IoT) and Machine-to-Machine (M2M) have many similarities,
including:

Data exchange: Both technologies allow devices to communicate and share data with
each other.

Automation: Both technologies can automate processes and allow devices to interact
without human intervention.
Remote control: Both technologies allow devices to be remotely monitored and
controlled.

Scalability: Both technologies are designed to support large-scale deployments.

Cloud integration: Both technologies often use cloud computing to store and analyze
data.

Industry applications: Both technologies are used in many industries, including

healthcare, manufacturing, transportation, agriculture, energy, and smart cities.

Data collection: Both technologies use sensors and other methods to collect data.

Data analysis: Both technologies require a way to analyze the collected data.

DIFFERENCES BETWEEN IoT & M2M

VALUE CHAINS:

Definition of M2M
Wikipedia: Machine to machine refers to technology that allowed both
wireless and wired systems to communicate with other devices of the same
type.
Digi: Machine to Machine (M2M) technology allows organization to gather
data from the edge of the enterprise and apply it in way that positively
impact the business.
Orange: exchange the information between machine that is established
between central control system (server) and any type of equipment, through
one or several communication networks.

Global value chains

A value chain describes the full range of activities that firms and workers
perform to bring a product from its conception to end use and beyond,
including design, production, marketing, distribution, and support to the final
consumer
Analyzing an industry from a global value chain (GVC) perspective permits
understanding of the context of globalization on the activities contained
within them by “focusing on the sequences of tangible and intangible value-
adding activities, from conception and production to end use. GVC analysis
therefore provides a holistic view of global industries both from the top down
and from the bottom up”.

M2M value chain

IoT value chains
The Information-Driven Global
Value Chain(I-GVC)
● Fundamental roles within the I-GVC
● Inputs:
● Sensors, RFID, and other devices.
● End-Users.
● Data Factories.
● Service Providers/Data Wholesalers.
● Intermediaries.
● Resellers.
The Information-Driven Global Value Chain.

Inputs to the information-driven global

commodity chain
● Sensors and other devices (e.g. RFID and NFC).
● End-users
->Both of these information sources input tiny amounts of data into the I-GVC
chain, which are then aggregated, analyzed, repackaged, and exchanged
from value chain.
-> As a result, sensor devices and networks, RFIDs, mobile and consumer
devices, Wi-Fi hotspots, and end-users all form part of a network of
“subcontractors” in the value chain, all contributing to the increased value of
the information products

Sensors and radio frequency identification

● Sensors and RFID devices are working as inputs to the I-GVC through the
capture and transmission of data necessary for the development of
information products.
 ● Smart phones have also been developed that allow mobile devices to interact
with sensors and RFID.
● This allows for a two-way interaction between a mobile terminal and the
sensor technology.
● The data is used as one part of the input to the commodity chain, which uses
it to create the information products that are eventually exchanged.
● In this sense, the sensor networks, and NFC and RFID technologies may be
viewed as subcontractors Resellers Intermediaries Data Factories Service
Providers/ Data Wholesalers End User Information Domestic and overseas
input from human and sensor-based ‘subcontractors’ to the I-GVC, workers
that constantly gather data for further processing and sale.

End-users
● End-users that choose to use and participate within the digital world are now
deeply embedded into the very process of production.
● Every human that enters a search query into a search engine,
● Every human that agrees to allow the mobile broadband platform to inform a
service of their location,
● Every human that uses NFC to allow a bank to establish and confirm their
identity are also functioning as subcontractors to the global information
systems that form the basis of the I-GVC.

Production processes of the information – driven

global value chain
● Data factories
● Data factories are those entities that produce data in digital forms for use in
other parts of the I-GVC.
● Service providers/data wholesaler:
● Service Providers and Data wholesalers are those entities that collect data
from various sources worldwide, and through the creation of massive
databases, use it to either improve their own information products or sell
information products in various forms.
● Example: Twitter, Facebook, Google

Intermediaries
● In the emerging industrial structure of the I-GVC, there is a need for
intermediaries that handle several aspects of the production of information
products.
● The development of databases such as the ones created by Google,
Facebook, and Twitter may therefore require the creation of entities that are
able to protect individuals’ privacy rights in relevant regional settings.
● Intermediary is to reduce transaction costs associated with the establishment
of a market for many different companies to participate in.
●
Resellers
● Resellers are those entities that combine inputs from several different
intermediaries, combine it together, analyze, and sell it to either end-users or
to corporate entities.
● These resellers are currently rather limited in terms of the data that they are
able to easily access via the converged communications platform, but they
are indicative of the types of corporate entities that are forming within this
space.