1.

Introduction to IoT
2. Definition and characteristics of IoT, Design of IOT
3. Physical design of IOT
4. Logical Design of IOT- Functional Blocks
5. Communication models
6. Communication APIs
7. IOT enabling Technologies- Wireless Sensor Networks
8. Cloud computing
9. Big data analytics
10.Embedded systems
11.IOT Levels and deployment templates
12.IoT Hardware and Software
13.Sensor and actuator
14.Humidity sensors
15.Ultrasonic sensor
16.Temperature Sensor
17.Arduino,
18.Raspberry Pi
19.LiteOS
20.RIoTOS
21.Contiki OS
22.Tiny OS
23.Architecture and Reference Model: Introduction,
24.Reference Model and architecture
25.Representational State Transfer (REST) architectural style
26.Uniform Resource Identifiers (URIs)
27.Challenges in IoT- Design challenges
28.Development challenges
29.Security challenges
30.Other challenges
31.IOT and M2M
32.M2M
33.Difference and similarities between IOT and M2M
34.Software defined networks, network function virtualization
35.Difference between SDN and NFV for IoT
 36. Case study of IoT Applications
37. Domain specific IOTs- Home automation
38. Cities, environment
39. Energy
40. Retail
41. Logistics
42. Agriculture
43. Industry
44. Health and Lifestyles

Introduction to IoT:
The Internet of Things (IoT) is a system of interrelated computing devices,
mechanical and digital machines, objects, animals or people that are
provided with unique identifiers and the ability to transfer data over a
network without requiring human-to-human or human-to-computer
interaction.”

Definition and characteristics of IoT, Design of IOT

If you just read that and thought, “ok, so what?”, you’re not alone. Most
people don’t want to nor need to dive into the nitty-gritty of IoT. In this
chapter, we’ll provide you with a simple explanation of the Internet of
Things and what it means for you.
Before we jump in, note that “The Internet of Things” and “IoT” can and will
be used interchangeably. And a quick tip to sound knowledgeable: avoid
saying “the IoT”.

A Simple, Non-Technical
Explanation of the Internet of
Things
How are you reading this ebook right now? It might be on desktop, on
mobile, maybe a tablet, but whatever device you’re using, it’s most likely
connected to the internet.

An internet connection is a wonderful thing, it give us all sorts of benefits

that just weren’t possible before. If you’re old enough, think of your cell
phone before it was a smartphone. You could call and you could text, sure,
but now you can read any book, watch any movie, or listen to any song all
in the palm of your hand.

The point is that connecting things to the internet yields many amazing
benefits. We’ve all seen these benefits with our smartphones, laptops, and
tablets, but this is true for everything else too. And yes, we do
mean everything.
The Internet of Things is actually a pretty simple concept, it means taking
all the physical places and things in the world and connecting them to
the internet.

Confusion arises not because the concept is so narrow and tightly defined,
but rather because it’s so broad and loosely defined. It can be hard to nail
down the concept in your head when there are so many examples and
possibilities in IoT.

To help clarify, it’s important to understand the benefits of connecting

things to the internet. Why would we even want to connect everything to
the internet?

Why IoT Matters

When something is connected to the internet, that means that it can send
information or receive information, or both. This ability to send and/or
receive information makes things “smart."

Let’s use smartphones again as an example. Right now you can listen to

just about any song in the world, but it’s not because your phone actually
has every song in the world stored on it. It’s because every song in the
world is stored somewhere else, but your phone can send information
(asking for that song) and then receive information (streaming that song on
your phone).

To be smart, a thing doesn't need to have super storage or a super

computer inside of it - it just needs access to it. All a thing has to do
is connect to super storage or to a super computer. In the Internet of
Things, all the things that are being connected to the internet can be put
into three categories:

1. Things that collect information and then send it.

2. Things that receive information and then act on it.

3. Things that do both.

And all three of these have enormous benefits that compound on each
other.

1. Collecting and Sending Information

Sensors could be temperature sensors, motion sensors, moisture sensors,
air quality sensors, light sensors, you name it. These sensors, along with a
connection, allow us to automatically collect information from the
environment which, in turn, allows us to make more intelligent decisions.
On a farm, automatically getting information about the soil moisture can tell
farmers exactly when their crops need to be watered. Instead of watering
too much (which can be an expensive over-use of irrigation systems) or
watering too little (which can be an expensive loss of crops), the farmer can
ensure that crops get exactly the right amount of water. This enables
farmers to increase their crop yield while decreasing their associated
expenses.

Just as our sight, hearing, smell, touch, and taste allow us, humans, to
make sense of the world, sensors allow machines (and the humans
monitoring the machines) to make sense of the world.

2. Receiving and Acting on Information

We’re all very familiar with machines getting information and then acting.
Your printer receives a document and it prints it. Your car receives a signal
from your car keys and the doors open. The examples are endless.

Whether it’s a simple as sending the command “turn on” or as complex as

sending a 3D model to a 3D printer, we know that we can tell machines
what to do from far away. So what?
The real power of the Internet of Things arises when things can do both of
the above. Things that collect information and send it, but also receive
information and act on it.

3. Doing Both: The Goal of an IoT System

Let’s quickly go back to the farming example. The sensors can collect
information about the soil moisture to tell the farmer how much to water the
crops, but you don’t actually need the farmer. Instead, the irrigation system
can automatically turn on as needed, based on how much moisture is in the
soil.

You can take it a step further too. If the irrigation system receives
information about the weather from its internet connection, it can also know
when it’s going to rain and decide not to water the crops today because
they’ll be watered by the rain anyways.

And it doesn’t stop there! All this information about the soil moisture, how
much the irrigation system is watering the crops, and how well the crops
actually grow can be collected and sent to supercomputers that run
amazing algorithms that can make sense of all this information.

And that’s just one kind of sensor. Add in other sensors like light, air
quality, and temperature, and these algorithms can learn much, much
more. With dozens, hundreds, thousands of farms all collecting this
information, these algorithms can create incredible insights into how to
make crops grow the best, helping to feed the world.

And agriculture is just one of many applications of IoT…

The Power of IoT: Examples

and Applications
The Internet of Things (IoT) promises to bring immense value to every
organization. By continuing to connect all our things, people, and
environments, we’ll unlock tremendous organizational value and achieve
feats that will truly seem like magic. But because IoT is so broad and far-
reaching of a concept, we’ve found that many are confused about what the
potential applications for IoT are exactly. How can my business actually
implement IoT solutions? How should my city think about creating value for
residents using IoT? Below we’ll give some Internet of Things examples
and applications to clear things up.

But before we do, we should first make the distinction between consumer
IoT and enterprise IoT. Consumer IoT refers to things like wearables, smart
home devices, etc., all of which are marketed directly to consumers. In
contrast, enterprise IoT refers the use of IoT in improving an organization’s
existing systems and processes and enabling organizations to increase
operational efficiency or unlock entirely new value (e.g. by launching new
business lines or products).

We’ll be focusing on enterprise IoT in this ebook because at Leverege, we

believe that this is where the most value can be created, even if it’s not
“sexy”. Plus, this is the area in which we have deep experience to share
with you. So now let’s explore some examples and applications of IoT.

It’s helpful to think of IoT as doing one (or more) of the following:
increasing efficiency, improving health/safety, or creating better
experiences.

Increasing Efficiency
“This years’ series of Internet of Things (IoT) and Industrial Internet of
Things (IIoT) forecasts reflect a growing focus on driving results using
sensor-based data and creating analytically rich data sets…solving
complex logistics, manufacturing, services, and supply chain problems.” —
Louis Columbus, Roundup Of Internet Of Things Forecasts And Market
Estimates, 2016
Increasing efficiency means more output with the same input or the same
output with less input. Inputs could include time, energy, money, or
resources. Output could be units produced or tasks accomplished.

Efficiency is particularly important for industrial applications, because more

production at less cost means greater profit, but efficiency gains can be
realized in just about any organization. Below are some examples:

Asset Tracking
Whether the assets are big or small, fixed or mobile, attaching sensors to
them allows organizations to track real-time location, monitor performance,
improve workflows, and optimize utilization.

For example, the smart boating solution we built for Siren Marine enables

boat owners to check in on their boat(s) from afar and make sure all
systems are functioning correctly. And the car tracking solution we built for
Manheim allows personnel at the auction locations to quickly locate the
vehicle(s) they’re looking for, rather than manually search through
thousands of parked cars.

Manufacturing Efficiency
Sensors embedded in manufacturing equipment and placed throughout a
factory can help identify bottlenecks in the manufacturing process. By
addressing bottlenecks, manufacturing time and waste is reduced.

Rather than standard preventative maintenance, which means performing

maintenance on machines before they break, “predictive maintenance”
means using advanced sensing and analytics to predict exactly when
machines will need maintenance. Because predictive maintenance means
only servicing machines when they need it, this cuts total costs and the
time machines spend idle.

Energy Efficiency

People and organizations can achieve significant decreases in their energy

usage with IoT. Sensors monitor things like lighting, temperature, energy
usage, etc. and that data is processed by intelligent algorithms to
micromanage activities in real-time. This is how Google cut 15% of its
energy expenditure in its data centers.

Agricultural Efficiency
For outdoor agriculture, an example could be sensing soil moisture and
taking weather into account so that smart irrigation systems only water
crops when needed, reducing the amount of water usage.
For indoor agriculture, IoT allows monitoring and management of micro-
climate conditions (humidity, temperature, light, etc.) to maximize
production.

Inventory Management
By placing tags on individual products, the exact location of single items in
a large warehouse can be shared, thus saving search time and lowering
labor costs.

Another example is in a retail setting. By knowing exactly what’s in-stock

and what isn’t, the store can order new products only when needed. This
reduces the cost of keeping extra inventory in the back. Also, smart
inventory management eliminates the need to manually check what’s on
the shelves, reducing labor costs.

Improved Health and Safety

IoT enables heightened surveillance, monitoring, and detection, which all
combine to improve health and increase safety. This is particularly
interesting for organizations like local or city governments, which need to
ensure the health and safety of their residents, but also extends to large
businesses supporting their employees.
Disaster Warning
Sensors can collect critical information about the environment, allowing for
early detection of environmental disasters like earthquakes, tsunamis, etc.,
thus saving lives.

Law Enforcement
Better surveillance and tracking tools will allow authorities to detect when
crime has occurred and respond much faster, keeping citizens safer. Also,
law enforcement will even be able to predict crime, stopping it from
happening in the first place.

Care giving
Patient surveillance can be life-saving; automatically detecting when
someone falls down or when they begin to experience a heart attack so
that emergency care can be sent immediately.

Environmental Quality
Sensors can also detect radiation, pathogens, and air quality so that
dangerous concentrations can be identified early, allowing people to
evacuate.
Better Experience
“We are stuck with technology when what we really want is just stuff that
works” — Douglas Adams, The Salmon of Doubt

The Internet of Things will allow our world to increasingly shape itself to our
needs and our wants, creating a better experience. Rather than just
passively providing information and reacting to our inputs, much of the
value of IoT will come from anticipating and addressing needs
automatically.

For example, if your building(s) are equipped with smart building

management systems, they can adjust temperature in real-time in response
to occupancy (how many people are in which areas of the building),
environmental factors (certain rooms might be getting more sunlight and in
need of further cooling), and contextual factors (today might be a holiday so
no one will be in the office).

While you might group the above example under energy efficiency because
it would in fact be saving energy, this would undervalue the better
experience provided to all your employees to make sure everyone has the
optimal temperature for comfort and performance.
Hopefully you’re beginning to grasp the potential of IoT, and some ways
that you can apply it to or within your own organization. But how does an
IoT system work exactly?

How an IoT System Actually

Works
As you saw in the previous chapter, the applications for IoT extend across
a broad variety of use cases and verticals. However, all complete IoT
systems are the same in that they represent the integration of four distinct
components: sensors/devices, connectivity, data processing, and a user
interface.

We’ll outline what each one means in the sections below and how they
come together to form a complete IoT system. Each of these sections will
also serve as the organizational structure of the rest of this ebook, and we’ll
dive into these components more deeply in the chapters to follow.

1) Sensors/Devices
First, sensors or devices collect data from their environment. This data
could be as simple as a temperature reading or as complex as a full video
feed.

We use “sensors/devices,” because multiple sensors can be bundled

together or sensors can be part of a device that does more than just sense
things. For example, your phone is a device that has multiple sensors
(camera, accelerometer, GPS, etc), but your phone is not just a sensor
since it can also perform many actions.

However, whether it’s a standalone sensor or a full device, in this first step
data is being collected from the environment by something.

2) Connectivity
Next, that data is sent to the cloud, but it needs a way to get there!

The sensors/devices can be connected to the cloud through a variety of

methods including: cellular, satellite, WiFi, Bluetooth, low-power wide-area
networks (LPWAN), connecting via a gateway/router or connecting directly
to the internet via ethernet (don’t worry, we’ll explain more about what
these all mean in our connectivity section).
Each option has tradeoffs between power consumption, range, and
bandwidth. Choosing which connectivity option is best comes down to the
specific IoT application, but they all accomplish the same task: getting data
to the cloud.

3) Data Processing
Once the data gets to the cloud (we’ll cover what the cloud means in our
data processing section)), software performs some kind of processing on it.

This could be very simple, such as checking that the temperature reading is
within an acceptable range. Or it could also be very complex, such as using
computer vision on video to identify objects (such as intruders on a
property).

But what happens when the temperature is too high or if there is an intruder
on property? That’s where the user comes in.

4) User Interface
Next, the information is made useful to the end-user in some way. This
could be via an alert to the user (email, text, notification, etc). For example,
a text alert when the temperature is too high in the company’s cold storage.
A user might have an interface that allows them to proactively check in on
the system. For example, a user might want to check the video feeds on
various properties via a phone app or a web browser.

However, it’s not always a one-way street. Depending on the IoT

application, the user may also be able to perform an action and affect the
system. For example, the user might remotely adjust the temperature in the
cold storage via an app on their phone.

And some actions are performed automatically. Rather than waiting for you
to adjust the temperature, the system could do it automatically via
predefined rules. Rather than just call you to alert you of an intruder, the
IoT system could also automatically notify security teams or relevant
authorities.

In Summary
An IoT system consists of sensors/devices which “talk” to the cloud through
some kind of connectivity. Once the data gets to the cloud, software
processes it and then might decide to perform an action, such as sending
an alert or automatically adjusting the sensors/devices without the need for
the user.
But if user input is needed or if the user simply wants to check in on the
system, a user interface allows them to do so. Any adjustments or actions
that the user makes are then sent in the opposite direction through the
system: from the user interface, to the cloud, and back to the
sensors/devices to make some kind of change.

That’s how an IoT system works at a high level. Now we’ll take a deeper
dive into each of these components to explain what they are, how they
work, and important considerations for your organization as you consider
building and/or implementing IoT solutions.

Chapter 2 : Sensors and Devices

Hardware Capabilities
As we established in the previous chapter, sensors/devices are a critical
piece of the Internet of Things, serving as a system's "senses" by
interacting with the world. Although no complete IoT solution can be built
without some kind of hardware, the sensors/devices are often an
underappreciated aspect of the system. Choices about the hardware affect
everything downstream, from the connectivity you choose, to the analytics
you’re able to provide, and to the interactions and interfaces that you
enable for end-users.
There are too many sensors/devices to possibly provide an exhaustive list
here. Ultimately, the choices you make on hardware stem directly from the
needs the specific application your organization is interested in pursuing.
So instead of exploring all the possible sensors/devices there are out there
for you to use, we’ll be providing some important factors you need to
consider.

Battery Considerations
One of the first considerations is whether or not the sensors/devices that
you’ll be using will have power available. If you’re designing, say, a smart
agriculture application with hundreds of sensors spread across broad, rural
areas, you’re going to have to rely on battery. If your application takes
place in a building and only involves a few devices, power may not be an
issue.

The reason this is one of the first considerations is because this directly
translates into how much you can do with the sensor/device. If your
hardware is battery-powered, you want it to last on battery power for a long
time - hopefully years (as we’ll cover in the next section, replacing and
managing batteries at scale is a huge operational burden).
But to get to multi-month or multi-year battery life, the device can’t be
constantly active. Power hungry operations on the device, such as using
GPS or sending and receiving messages over the network, must be used
judiciously.

To get a GPS fix, a device must “listen” for signals from at least 4 GPS
satellites that are in orbit (here’s how GPS works). Depending on the
terrain (which can block line-of-site to satellites) and the position of the
satellite constellation, this can take 30 seconds, 60 seconds, or more. All of
that time is time that the GPS unit is draining precious battery.

The same goes for connectivity. To receive messages over the network,
the sensors/devices must be in “listening” mode, and that means battery
drain. So when does your device listen? Are there situations where you
need to push a critical message down to the device, such as triggering an
alarm? Well that’s going to significantly impact your battery life. And if not,
how often do you want the device to check in?

A “heartbeat” message is a periodic message from the device where it

essentially tells you that it’s still alive and functioning. If you make its
heartbeat once-per-week, you’re significantly reducing battery drain, but
that also means a device might die and be offline for a week or two before
you notice. If it’s every 10 minutes, that’s much closer to real-time but also
1000x the battery drain relative to the once-per-week.

All of these battery considerations also influence the kind of connectivity

you choose, which we’ll cover next in the connectivity section. The point is
that hardware decisions are extremely important and there are many
considerations to take into account, which all stem directly from your
specific use case.

Over-the-Air (OTA) Firmware

Updates
However, regardless of application, having support for over-the-air
(OTA) firmware updates is essential. At a high level, “firmware” is the
program that’s put on the hardware of the device, basically telling it how to
function and perform. As the name implies, it’s between hardware (which
you can’t change once it’s been manufactured) and software (which you
can update with relative ease).

While firmware can be updated (and should be), it’s a non-trivial process.

Having devices that can receive updates to their firmware over-the-air
(meaning that that you can update them over the network, rather than
needing to have them physically in your hands) is critical, especially in
applications where you have devices spread over large areas.

OTA firmware updates are a critical tool in addressing issues that may
come up as you learn and refine, but they can also cause major issues. If
you’re considering pushing a firmware update to all of your
sensors/devices, make sure that you’ve tested extensively on a small
subset that are actually out in the field. We’ve had experiences where
firmware updates caused unforeseen issues and drained the batteries of
hundreds of devices before new firmware update to fix it could be sent
OTA.

Key Takeaways
We’ve given a few examples of important hardware considerations above,
but IoT applications vary so widely in their requirements that a
comprehensive exploration of all considerations isn’t within the scope of
this ebook.

One of the key takeaways from our experience is that no matter how much
you think through your application, inevitably you’re going to run into things
you didn’t foresee. So if your application has been done before, you may
be able to find hardware that has been purpose-built directly for your
application, which you can and should leverage.

If you’re pursuing something new, make sure to work with an experienced

hardware partner and listen closely to their expertise. Every little decision
matters, and there are likely many considerations you may not have
thought of.

‍ Scaling & Operations

Like the considerations around the capabilities of hardware, which we just
explored in the previous chapter, another underappreciated aspect of a
large-scale IoT solution is the operational component.

It’s one thing to deal with a prototype that only has a few pieces of
connected hardware, or a pilot with just a few hundred sensors/devices at
one location. It’s quite another when you scale up to a full production
system with potentially millions of sensors/devices, and there are critical
considerations purely on the operations side that you need to address.

More Battery Considerations

In the last chapter, we covered how the needs of the application influence
hardware decisions like whether or not you use battery power in your
sensors/devices. If you go with battery power, what happens when you
need to replace the batteries? First, you need to figure out how you’re
going to access or collect the sensors/devices.

Are they spread over miles of terrain? That means you’ll need to know
where they all are, because if they run out of battery power they won’t be
using GPS.

Or maybe they don’t even have GPS because they’re just soil moisture
sensors and don’t need it. Do you go get them one at a time as the
batteries die? Or do you wait and do it in batches?

How do you replace the batteries? If you’re replacing, don’t forget the cost
of all the additional batteries you’ll need!

Or are they rechargeable? This means you don’t need to replace batteries,
which is nice, but now you need to figure out how they’re recharging. If they
can charge wirelessly that’s great, you can just put them on a wireless
charging matt and decrease hassle. But if they need to be plugged in, that
means either dedicated charging stations for multiple sensors/devices at
once, or a ton of wires. Or maybe the devices are cheap enough that you
just replace them completely; in that case make sure to consider how
you’re going to dispose of them!

Sensor/Device Association
Say you’re doing an asset tracking application. Great. That means you
have some way of tracking the location of the asset (probably GPS if
outdoors and over large areas, probably bluetooth if indoors and within
relatively confined areas) by putting a device on it. But how do you know
that a specific device is associated with a specific asset?

Sensor/device association (for example, device A12B3 goes with this car)

can be a big hassle. It’s critical, because otherwise you just know the
location of the sensors/devices but not the specific assets they’re attached
to. This means that, at some point in the process, you need to make that
association. Does someone have to manually enter the information? Is
there a barcode on the asset that you can scan? For example, all cars have
a VIN (vehicle identification number) which can be scanned. That’s
awesome, but now you need to sync the information you already have in
your system (the VINs of your cars) with the new location data you’re
getting, and that’s going to require some integration.
Sensor/Device Errors
As much as we all hope for things to work perfectly, inevitably there are
going to be errors. These could stem from a defect in manufacturing for a
specific sensor/device or could be due to a bug in the firmware.

Regardless, you need a process for how you handle when a sensor/device
has an error. How do you 1) identify the error in the first place and 2)
address the issue once you find out?

The ability to predict and proactively address errors is absolutely critical.

Dealing with errors in sensors/devices after they happen can be a huge
operational burden. Just imagine trying to find a single sensor/device over
square miles of parking lot when the error means that you can’t get GPS
location anymore. Not fun.

Key Takeaways
The above considerations may not matter for many applications, whether
that’s because you don’t need that many devices or because you’re not
relying on battery power. Again, we bring up these examples simply to get
you thinking and realizing that there are many purely operational factors,
beyond technology, to consider when you go from prototype to pilot to full
scale.

This is why many IoT deployments fail. Not because the underlying
technology isn’t good enough, but because the sheer operational burden
makes the return-on-investment not worth the effort. This is also why it’s so
critical that you have a clear, measurable impact you’re trying to achieve
with your solution.

Manufacturing & Shipping

During the prototype and pilot phases, you can get by with using hardware
that’s been hacked together from off-the-shelf parts. But when you go to
deploy a full IoT system, you’re likely going to need to work with a
manufacturer.

Manufacturing could be an entire ebook in and of itself (and maybe it will

be, let us know if you’re interested!), but here we’ll give you some of the
important considerations for manufacturing production-grade
sensors/devices. Some of the points below won’t apply when you’re using
sensors/devices that have already been built and manufactured, and
whether you need to build entirely new sensors/devices will (as always)
depend heavily on your specific use case.
Manufacturing: It’s Going to Take
a While
The most important consideration, is that it’s going to take a while. Do not
expect to go from a few prototypes to multiple thousands of production
units within a couple months. We’ve done it in four or five months, but that’s
like trying to sprint a marathon. It’s possible, but it’s more than likely you’ll
burn out and fail.

Expect several months to a year to go through the entire process. And

scale also factors in here; a manufacturer will be much more willing to
move quickly if they know they’ll be producing hundreds of thousands of
units than they will to produce just a few thousand.

This process takes a considerable amount of time for a number of reasons.

With software, if there are bugs or areas of improvement, you can update
the system after it’s live. And with the connectivity layer, you might be using
proven standards and existing infrastructure that carry relatively low risk
(like WiFi or Cellular). Even if you’re using a relatively new standard,
making updates to the network can also be performed after-the-fact.
With hardware, you are not making changes to your sensors/devices once
they’ve been produced. This is why having a small scale pilot is absolutely
critical, it allows you to rigorously test your prototypes to identify any bugs
or weaknesses in the hardware so that the final units can as purpose-built
for the specific use case as possible. And manufacturing usually involves
building injection molds, which help reduce per-unit costs but have massive
upfront fixed-costs, meaning that it’s extremely costly and time-intensive to
make changes.

Also, for any sensor/device that’s communicating wirelessly, you’ll need to

get FCC certifications (if you’re in the US that is; for any other country you’d
get certifications from the relevant governing body). Wireless means that
the device is communicating using electromagnetic waves over a certain
spectrum, so these certifications are to ensure that your device isn’t
harmful to people nor infringing on licensed bands.

Another reason that the manufacturing process can take a while, is that you
need to source all the materials. For a given sensor/device, there may be
dozens or into the hundreds of individual components necessary to build
the full device.

The manufacturer will need to set up production lines too. This involves
setting up all the equipment and assembly lines to actually manufacture the
sensors/devices, as well as  setting up the testing processes for each
important stage. You want your sensors/devices to be tested throughout
the entire process to ensure that you don’t produce an entire batch of units
with defects.

Shipping: Unexpected Issues

Finally, once the units have been produced, they’ll need to be shipped to
wherever they’re needed (perhaps directly to you or to the location where
they will be deployed). If you’re manufacturing in China, which is likely, this
means that the devices will first need to pass customs because they’re
being imported from abroad. This process can be several days or weeks
and doesn’t have a set time, so make sure there’s a bit of cushion in your
delivery schedule.

And shipping itself is an important consideration. If units have lithium

batteries, there are regulations that may prevent them from being
transported on airplanes (due to possibility of combustion).

Also, it’s critical to consider device behavior (which ties into our earlier
chapter on hardware capabilities). Do your devices “know” they’re being
shipped? Do they stay in sleep mode? We once made the mistake of
shipping units that, because of movement, “woke up” and began trying to
find a network. Since they were in transit, they couldn’t find a network and
ended up completely draining their batteries trying to connect. Most were
dead when they arrived on location.

Key Takeaways
It’s no accident that hardware has “hard” in its name. Here at Leverege, we
develop software solutions and act as the overarching systems integrator
on end-to-end IoT solutions; we don’t manufacture hardware. However,
we’ve been fortunate to have hardware partners with extremely deep
experience and high aptitude and recommend that you listen closely to the
words of any hardware partners you work with.

Gateways
As mentioned previously, our goal in this section of the ebook isn’t to go
into detail about specific sensor/device types. However, nearly every IoT
system needs some way to connect its sensors/devices to the cloud so that
data can be sent back-and-forth between them. So in this chapter we’ll be
exploring a particular type of IoT hardware called the gateway, which
makes that connection to the cloud possible.
Gateways act as bridges between sensors/devices and the cloud. Many
sensors/devices will “talk” to a gateway and the gateway will then take all
that information and “talk” to the cloud.

But you may be wondering, what benefit is there to taking that extra step
between the sensors/devices and the cloud? There are several benefits:

Battery life
As you may be noticing, battery life tends to be a critical consideration for
many IoT systems. For example, take an IoT solution that operates in a
remote area. To get data from sensors/devices to the cloud, there will need
to be a long-range connection, usually provided by satellite. As will be
explained in greater depth in the connectivity section, longer range typically
means increased power consumption (and costs); this can be a problem for
small sensors/devices with limited battery life.

If you’re doing Smart Agriculture, you want your field sensors to last years,
not months or weeks. By using an elevated gateway installed near the top
of an outbuilding or grain silo, the sensors/devices only have to send data a
relatively short distance to the gateway and the gateway can then send the
data to the cloud through a single higher bandwidth connection like
satellite.
Gateways allow sensors/devices to communicate over shorter distances,
boosting battery life.

Varying Protocols
A complete IoT application might involve many different kinds of sensors
and devices. Using Smart Agriculture again, you might want sensors for
temperature, moisture, and sunlight and devices such as automated
irrigation and fertilizer systems.

All of the different sensors and devices can use varying transmission
protocols (basically, the rules and format for the information being
transmitted). Protocols include LPWAN, Wi-Fi, Bluetooth, and Zigbee,
among many others.

Gateways can communicate with sensors/devices over varying protocols

and then translate that data into a standard protocol such as MQTT to be
sent to the cloud.

Unfiltered Data
Sometimes, sensors/devices can generate so much data that it’s
overwhelming to the system or extremely costly to transmit and store. Often
in such cases, only a small fraction of the data is actually valuable. For
example, a security camera doesn’t need to send video data of an empty
hallway.

Gateways can pre-process and filter the data being generated by

sensors/devices to decrease transmission, processing, and storage
requirements. There are also techniques that can be employed on the
sensor processor itself (if there is enough processing power) to limit the
amount of unfiltered data sent to the gateway or directly across the
network.

High Latency
Time can be critical for certain IoT applications; the sensors/devices can’t
afford to transmit data to the cloud and wait to get a response before taking
action. This is true for life-or-death situations in the medical realm or for
fast-moving objects like cars.

Higher latency can be avoided by processing the data on the gateway or on

the sensor itself and giving commands locally. However, many
sensors/devices in IoT applications are too small and too power
constrained to do the processing themselves.
Gateways can reduce latency in time-critical applications by performing
processing on the gateway itself rather than in the cloud.

Security
Every sensor/device that is connected to the internet becomes vulnerable
to being hacked. Hacked sensors/devices are bad. Not just for the owner,
but for everyone else too.

Gateways reduce the number of sensors/devices connected to the internet

because the sensors/devices are only connected to the gateway. However,
this makes gateways themselves targets and also the first line of defense.
This is why security needs to be a priority for any gateway.

Key Takeaway
Not all IoT applications will need a gateway, but they’re an important class
of hardware that’s often a requirement for certain use cases because
they’re needed to provide the connectivity to the sensors/devices.

In the next section we’ll explore connectivity as a whole for IoT as well as
specific connectivity and network standards that you may need to consider.
Chapter 3: Connectivity

An Introduction to
Connectivity
When it comes to connecting the Internet of Things, there are a seemingly
overwhelming number of options. Cellular, satellite, WiFi, Bluetooth, RFID,
NFC, LPWAN, and Ethernet are just some of the possible ways to connect
a sensor/device. And within each of these options there can be different
providers (e.g. for cellular there’s T-Mobile, Verizon, AT&T, Sprint, etc.).

Trade Off Between Power

Consumption, Range, and
Bandwidth
The perfect connectivity option would consume extremely little power, have
huge range, and would be able to transmit large amounts of data (high
bandwidth). Unfortunately, this perfect connectivity doesn’t exist.

Each connectivity option represents a tradeoff between power

consumption, range, and bandwidth. This allows us to segment the various
connectivity options into three major groups, which you’ll find below.
However, these groups should serve more as a framework for thinking
about connectivity than a definitive classification, as there can be
connectivity standards that sit more on the borders of these groups.

1) High Power Consumption, High Range, High

Bandwidth
To wirelessly send a lot of data over a great distance, it takes a lot of
power. A great example of this is your smartphone. Your phone can receive
and transmit large amounts of data (e.g. video) over great distances, but
you need to charge it every 1–2 days.  Connectivity options in this group
include cellular and satellite.

Cellular is used when the sensor/device is within coverage of cell towers.

For sensors/devices that are, say, in the middle of the ocean, satellite
becomes necessary.

2) Low Power Consumption, Low Range, High

Bandwidth
To decrease power consumption and still send a lot of data, you have to
decrease the range. Connectivity options in this group include WiFi,
Bluetooth, and Ethernet.
Ethernet is a hard-wired connection, so the range is short because it’s only
as far as the wire length. WiFi and Bluetooth are both wireless connections
with high bandwidth and lower power consumption than cellular and
satellite. However, as I’m sure you’ve experienced just walking around your
home, the range is limited.

3) Low Power Consumption, High Range, Low

Bandwidth
To increase range while maintaining low power consumption, you have to
decrease the amount of data that you’re sending.  Connectivity options in
this group are called Low-Power Wide-Area Networks (LPWANs).

LPWANs send small amounts of data which allows them to operate at very
low power with ranges in miles rather than feet. For example, a moisture
sensor for agricultural purposes doesn’t need to send a lot of data, perhaps
just a single number (the moisture level) every few hours. You also don’t
want this sensor to consume a lot of power because it needs to run on
battery (plugging it into an outlet in the middle of a field just isn’t realistic).
And since agriculture covers a wide area, WiFi and Bluetooth lack the
range.

LPWANs are extremely useful for many IoT applications. They allow tons of
sensors/devices to collect and send data over broad areas while lasting
years on battery life. Although they can’t send much data, most sensors
don’t need to. However, these kinds of application often need IoT gateways
to work, which we explored in the previous chapter.

When to Skip Connectivity

The Internet of Things is made up of connected sensors/devices, so by
definition an IoT system needs some kind of connectivity, especially if it
uses the cloud.

However, there are certain cases where the data processing or the
interaction with the sensor/device through the user interface can take place
without any data first being transferred over an external network.

Why Skip the Connectivity?

One reason is latency. Latency refers to how long it takes for a packet of
data to get from the start point to the end point. Although latency doesn’t
matter in the vast majority cases, for some IoT applications latency is
critical.

Imagine you’re in a self-driving car and suddenly somebody loses control of

their car in front of you. Would you want to wait for the self-driving car to
send data to the cloud, have that data processed, then have instructions for
what to do sent back to the car? No! Those milliseconds could mean life or
death.

Even if you’re the one driving the car, you want the user interface (i.e. the
steering wheel) directly hooked up to the device (i.e the car) rather than
waiting for your input to be transmitted externally, processed, and then sent
back.

Another reason is that sending lots of data can become really expensive.
Some IoT applications collect a ton of data but only a small fraction is
actually important. Local algorithms can restrict what gets sent thus
lowering costs.

A good example is a security camera. Streaming video takes a lot of data,

but the vast majority of the footage might be of an empty hallway.

So How Do You Skip the Connectivity?

Rather than send data over a network for it to be processed in the cloud, an
alternative approach is to process the data on a gateway or on the
sensor/device itself. This is called either fog computing or edge computing
(because you’re bringing the cloud “closer to the ground” and the
computing is taking place at the edges of the IoT system rather than the
center).

For the security camera, it could use machine vision to “watch” for anything
abnormal and only then send that footage to the cloud.

For the self-driving car, the data processing all takes place in the onboard
computer which allows for faster decision-making.

Key Takeaway
Every IoT system combines the four components we’ve outlined and begun
to detail, Sensors/Devices, Connectivity, Data Processing, and User
Interface. However, an IoT system can combine these components in
different ways and can use very different forms of connectivity. It all comes
down your specific application and organizational need.

In the coming chapters we’ll take a deeper dive into some of the
connectivity options we’ve mentioned here.
LPWAN
As the name implies, Low-Power Wide-Area Networks (LPWANs) allow for
low power consumption over a wide area, aka long range. So how is this
accomplished?

Messages sent over LPWAN must be small and simple. Because of their
simplicity, these messages can be communicated over the distance without
a large power source. For machines, decreasing the amount of data sent
(the bandwidth) means lower energy at range.

This is what LPWANs do, they send and receive small packets of
information at infrequent intervals. Sensor/devices can send data over
miles of range instead of feet and can last for years on battery instead of
weeks or months.

However, LPWANs aren’t without downsides. Messages that are

transmitted over LPWAN sometimes aren't received by the gateway (called
packet loss). This can usually be overcome by sending multiple messages
or by adding additional gateways to the network, but these solutions have
power and financial costs respectively.
Despite certain disadvantages, LPWANs play an essential role in the
Internet of Things.

Key Takeaways
IoT applications can vary greatly, but many applications need tons of
sensors spread over big areas.

There are many ways for these sensors/devices to communicate, each with
varying pros and cons. When you have thousands of sensors spread over
a big area, you need wireless communication with long range and low
power consumption. After all, and as we covered previously, it’s a
tremendous operational burden to replace the batteries in thousands of
sensors on a frequent basis.

Also, it costs money to send messages and connectivity options like

cellular are expensive. Imagine having to pay your phone bill not just for
one device, but for thousands. Yikes.

LPWAN technology thus plays a crucial role in enabling the Internet of

Things. These networks make it possible to have many thousands of
sensors/devices collecting and sending data at lower cost, over longer
range, and with better battery life than other connectivity options. Some use
cases among many include:

 A car auction lot or dealership — sensors are placed on vehicles to

track their location and status in real time.

 A parking garage — sensors detect when spots are open, sending a

simple Yes or No message only when that value changes.

 A school building — battery-powered locks can be remotely

activated or deactivated, helping with general security and crisis
situations.

 A city — waste containers throughout a city can send alerts when

they’re close to being full, allowing for more efficient garbage
collection.

It’s important to note that LPWAN is a general term, and there are many
different competing standards and technologies under that umbrella. The
competing LPWAN standards and technologies include but are not limited
to: LoRa, SIGFOX, Ingenu, Weightless, and SymphonyLink. For the
purposes of this ebook we won’t go into detail on these specific
technologies, but at Leverege we have hands-on experience with most of
them which we’re always happy to share if you’re weighing them for an
application.
Cellular
Cellular networks provide the backbone for much of what we know and
love, allowing us to access the internet, send messages, and connect with
friends. In addition to the personal benefits we're all familiar with, cellular
networks also serve a critical role in many Internet of Things applications.

As we’ve discussed in previous chapters, there will always be a tradeoff

between power consumption, range, and bandwidth. Cellular connectivity
has been focused on range and bandwidth at the expense of power
consumption, meaning that it can send lots of data over long distance but
drains battery rather quickly.

This is fine for devices that are connected to an electricity source or that
can be recharged often (i.e. your phone), but a no-go when it comes to IoT
applications that require remote sensors/devices to last months or years on
battery.

As such, cellular connectivity is usually reserved for backhaul (i.e. a

gateway might use LPWAN to talk to all the sensors/devices but use
cellular to connect to the cloud and pass along that data) or for
sensors/devices that need to send a lot of data and/or don’t have concerns
about battery life.
However, that's not the full story when it comes to cellular. You've probably
heard names like 2G, 3G, and 4G (which refer to different generations of
cellular networks), but new cellular technologies like Cat-1, LTE-M, and
NB-IoT are aimed specifically at IoT applications. As with LPWANs, we
won’t go into the details of these cellular technologies, but the main
takeaway is that they are either upgrades to existing networks or new
infrastructure entirely, and all aimed specifically at reducing data costs per
sensor/device and power requirements.

Some of these cellular technologies are currently available, and others are
promised but yet to come. This also includes 5G, which will also
have significant implications for IoT applications and enable high-
bandwidth, high-speed applications like Ultra-HD (4K) streaming, self-
driving car connectivity, or VR/AR applications.

There’s also discussion around supporting IoT devices with 5G-IoT

networks. However, all these are just speculations as 3GPP (the
standards organization for cellular technologies)) will finalize the
specifications in 2019. The commercial rollout target year is 2020.

Key Takeaway
It’s important to understand that these different options do not have to be
mutually exclusive. This extends to other connectivity options as well, like
the LPWANs.

IoT covers a broad spectrum of applications. Sometimes you need high

bandwidth, like with real-time surveillance. For asset tracking, data
throughput is small, but there are inevitably many handovers as objects
move. Smart meters and many smart city use-cases require small data
transfer once or twice a day. This means that no one technology (even 5G)
may fit the specific needs of your particular IoT solution and may use a
combination.

Satellite
As the name implies, this form of connectivity uses satellites to connect
sensors/devices to the cloud. The first artificial communications satellite
was launched in 1960, and served merely as a giant reflector for signals
beaming between different places on the earth’s surface. Today’s
communications satellites are much more robust and featured.

The importance of satellite connectivity for the Internet of Things comes

from its incredible coverage. A single network of satellites is capable of
providing coverage to effectively the entire planet. This means that a single
device moving around the world can stay on a single network and use only
a single connectivity type.

Satellite’s incredible range give it an advantage in remote areas that other

communication types such as cellular or Wi-Fi cannot reach, and in places
that have underdeveloped infrastructure or none at all, such as the middle
of the ocean.

Satellite connectivity has two major configurations with respect to

connectivity: direct and backhaul.

Direct
The first major type of configuration, direct, is broken down into the two
sub-categories: dual mode and satellite only.

Dual mode satellite connectivity is connectivity that uses cellular data as

much as possible and uses satellite when necessary. This gives a best-of-
both-worlds connectivity option that leverages the lower cost and higher
bandwidth of cellular when possible, but makes use of satellite
connectivity’s greater coverage to fill in spaces where cellular data
connections are sparse or unreliable.
The best example of this connectivity is container ships, which use cellular
when in port or near coastlines, but make use of satellite when on the open
ocean.

Satellite-only connectivity is exactly what is sounds like, a data connection

that uses purely satellite connectivity to transmit data. This is typically for
large, immobile resources like oil and gas equipment, that are sending
large amounts of data from locations that have no cellular or other
connectivity options.

Backhaul
The second major type of configuration, backhaul, uses a main tower that
connects directly to a satellite and then a different kind of connectivity (e.g.
an LPWAN) to connect with the sensors/devices in the area. This
connectivity option is typically used when you have many low bandwidth
sensors/devices in remote areas.

Satellite requires high power usage, and can require larger pieces of
equipment such as dishes for connectivity. This raises the cost for
individual sensors/devices, and can make direct connection infeasible for
groups of sensors/devices that don’t use much data.
One example of this is a farm that uses a set of moisture sensors to collect
soil data. All of those sensors may use an LPWAN to connect to a main
tower that then transmit the data over a satellite connection. This saves on
battery life and lowers the overall cost of the sensors.

Key Takeaways
Satellite has excellent coverage, but with it comes larger equipment and
higher battery usage than other connectivity options. Satellite also has
good bandwidth, but can be expensive at scale. As such it fills a niche
where a single tower can be used to service a group of sensors/devices,
sensors/devices are larger and higher costs are acceptable, or
sensors/devices are so remote that satellite is the only means of
transmitting data to the cloud.

In these instances, satellite is an excellent connectivity option because a

single network can encompass the entire globe, and connectivity can be
reliable in places no other options can reach, even in the middle of the
ocean.

‍
WiFi
WiFi has a few notable differences from other wireless technologies. For
example, WiFi transmits at frequencies of 2.4 GHz or 5 GHz. These
frequencies are much higher than the frequencies used for cellular
transmission. Higher frequency means that signals can carry more data.

However, as you now know well, all forms of wireless communication

represent a tradeoff between power consumption, range, and bandwidth.
So in exchange for high data rates, WiFi consumes a lot of power and
doesn’t have a lot of range.

In fact, the longest range WiFi has ever transmitted data is 260 miles. The
Swedish Space Agency transmitted data to an overhead stratospheric
balloon 260 miles away, but they used non-standard WiFi equipment and 6
watt amplifiers to achieve this.

For your average WiFi router, ranges are much, much shorter and depend
on a number of factors. Range can depend on the antenna, reflection and
refraction, and radio power output. A range of about 100 ft is common, so if
you have thousand of sensors out in a field, WiFi isn’t a great option.
WiFi can be good for IoT applications that don’t have to worry about power
drain (e.g. devices that are plugged into an outlet), that need to send a lot
of data (e.g. video), and that don’t need high range. A good example would
be a home security system.

Types of WiFi
Like LPWANs and cellular connectivity, there are several versions of WiFi
including, 802.11a, 802.11b, 802.11g, 802.11n, and 802.11ac.

Each of these standards comes with various pros/cons related to data

speed, signal interference from outside sources, and cost. Cost is a factor
because different hardware is needed for different standards, though newer
versions are made to be backwards compatible with older versions.

So, while WiFi currently isn’t great for many IoT applications, there are two
WiFi standards that have been developed, or are being developed,
specifically for IoT; WiFi HaLow (802.11ah) and HEW (802.11ax).

WiFi HaLow was ratified in 2016 and aimed at addressing range and power
concerns for IoT applications. HEW (High Efficiency Wireless) is an
upcoming standard that builds on HaLow to add additional IoT-friendly
features.
Key Takeaway
As always, it all comes down to your specific application. One of the
advantages of WiFi is that it’s a proven and standardized technology that’s
already present in many buildings and public areas. However, current
instantiations of WiFi lack the necessary  range and consume too much
power for many IoT applications.

Bluetooth
Invented by Ericsson in 1994, Bluetooth was intended to enable wireless
headsets. Bluetooth has since expanded into a broad variety of
applications including Bluetooth headsets, speakers, printers, video game
controllers, and much more.

Bluetooth is also important for the rapidly growing Internet of Things,

including smart homes and industrial applications. It is a low power, low
range, high bandwidth connectivity option. When Bluetooth devices
connect to each other (for example, your phone and your wireless
speaker), it follows the parent-child model, meaning that one device is the
parent and other devices are the children. The parent transmits information
to the child and the child listens for information from the parent.
A Bluetooth parent can have up to 7 children, which is why your computer
can be connected via Bluetooth to multiple devices at the same time. When
devices are connected together via Bluetooth, it’s called a “piconet”.

Not only can a device be a parent in one piconet and a child in a different
piconet at the same time, but the parent-child relationship can also switch.
When you put your Bluetooth device in pairing mode to connect it, it’s
temporarily becoming the parent so that it can establish a connection and
proceeds to connect as the child.

In contrast to WiFi, which we explored in the previous chapter, Bluetooth

was meant for portable equipment and related applications therefore excels
when you need to connect two devices with minimal configuration. Also,
because Bluetooth uses weak signals, there’s limited interference and
devices can communicate in “noisy” environments.

In the Industrial Internet of Things, machines often need to send short

bursts of data in extremely noisy environments. With potentially hundreds
of sensors and devices sending data, WiFi poses too much hassle to set
up.

A drawback of Bluetooth is lower bandwidth, but for many industrial

applications this higher bandwidth simply isn’t needed.
Bluetooth is also useful in a smart home setting. Again, many devices in
the smart home don’t need high bandwidth connections and it’s much
easier to set up Bluetooth.

Furthermore, newer versions of Bluetooth can create a self-healing mesh

network which means that individual devices can still communicate even if
one device runs out of power or is disconnected. If your door locks, HVAC
system, washer, dryer, fridge, and lights are all connected, you certainly
wouldn’t want them all to fail just because one goes down.

Bluetooth Version 5
The Bluetooth Special Interest Group officially adopted Bluetooth 5 as the
latest version of Bluetooth back in December 2016.

“With Bluetooth 5, Bluetooth continues to revolutionize how people

experience the IoT. Bluetooth continues to embrace technological
advancements and push the unlimited potential of the IoT.”

—Bluetooth 5 Now Available

As is clear from Bluetooth SIG’s announcement, Bluetooth 5 is specifically

aimed at the Internet of Things. It boasts quadruple the range, double the
speed, and boosts broadcast messaging capacity by 800%. It also
introduces the mesh networking capability mentioned above.

Bluetooth 5 is backwards-compatible with previous versions of Bluetooth,

but new hardware is required to take advantage of the new benefits listed
above. So it might be awhile until we see all the benefits that Bluetooth 5
has to offer, but it’s an exciting development as the Internet of Things
continue to gain traction.

Key Takeaway
In addition to the capabilities explored above, Bluetooth can also provide
indoor asset tracking by using multiple Bluetooth beacons and using their
relative signal strengths to triangulate position. GPS is great for outdoor
applications but has inherent accuracy limitations and fails indoors when
sensors/devices can’t receive the signal from the GPS satellites.

Together with the advantages in noisy environments and the ease of setup,
Bluetooth is therefore a strong option for many indoor Internet of Things
applications.
Chapter 4: Data Processing

Introduction to the Cloud

So far we’ve covered the sensors/devices that are out in the world
collecting data, and the connectivity technologies that enable those
sensors/devices to pass that data up to the cloud for processing. But what
is the cloud? And what happens when that data is received?

Back in the 1970s, it was popular for businesses to rent time using big,
mainframe computer systems. These systems were extremely large and
expensive, so it didn’t make sense financially for businesses to own the
computing power themselves. Instead, they were owned by large
corporations, government agencies, and universities.

Microprocessor technology allowed for great reductions in size and

expense, leading to the advent of the personal computer, which exploded
in popularity in the 1980s. Suddenly, businesses could (and did) bring
computation in-house.

However, as high-speed connections have become widespread, the trend

has reversed: businesses are once again renting computing power from
other organizations. But why is that?
Instead of buying expensive hardware for storage and processing in-house,
it’s easy to rent it for cheap in the cloud. The cloud is a huge,
interconnected network of powerful servers that performs services for
businesses and for people.

The largest cloud providers are Amazon, Google, and Microsoft, who have
huge farms of servers that they rent to businesses as part of their cloud
services.

For businesses that have variable needs (most of the time they don’t need
much computing, but every now and then they need a lot), this is cost
effective because they can simply pay as-needed.

When it comes to people, we use these cloud services all of the time. You
might store your files in Google Drive instead of on your personal
computer. Google Drive, of course, uses Google’s cloud services.

Or you might listen to songs on Spotify instead of downloading the songs to

your computer or phone. Spotify uses Amazon’s cloud services.

Generally, something that happens “in the Cloud” is any activity that takes
place over an internet connection instead of on the device itself.
The Internet of Things and the
Cloud
Because activities like storage and data processing take place in the cloud
rather than on the device itself, this has had significant implications for IoT.

Many IoT systems make use of large numbers of sensors to collect data
and then make intelligent decisions.

Using the cloud is important for aggregating data and drawing insights from
that data. For instance, a smart agriculture company would be able to
compare soil moisture sensors from Kansas and Colorado after planting
the same seeds. Without the cloud, comparing data across wider areas is
much more difficult.

Using the cloud also allows for high scalability. When you have hundreds,
thousands, or even millions of sensors/devices, putting large amounts of
computational power on each sensor/device would be extremely expensive
and energy intensive. Instead, data can be passed to the cloud from all
these sensors and processed there in aggregate.

For much of IoT, the head (or rather, the brain) of the system is in the
cloud. Sensors/devices collect data and perform actions, but the
processing/commanding/analytics (aka the “smart” stuff), typically happens
in the cloud.

So Is the Cloud Necessary for IoT?

Technically, the answer is no. The data processing and commanding could
take place locally rather than in the cloud via an internet connection. Known
as “fog computing” or “edge computing”, this actually makes a lot of sense
for some IoT applications.

However, there are substantial benefits to be had using the cloud for many
IoT applications including:

 Decreased costs, both upfront and infrastructure.

 Pay-as-needed for storage/computing.

 High system scalability and availability.

 Increased lifespan of battery-powered sensors/devices.

 Ability to aggregate large amounts of data.

 Anything with an internet connection can become “smart”.

There are legitimate concerns with cloud usage though:

  Data ownership. When you store data in a company’s cloud service,
do you own the data or does the cloud provider? This can be hugely
important for IoT applications involving personal data such as
healthcare or smart homes.

 Potential crashes. If connection is interrupted or the cloud service

itself crashes, the IoT application won’t work. Short-term inoperability
might not be a big deal for certain IoT applications, like smart
agriculture, but it could be devastating for others. You don’t want
applications involving health or safety crashing for even a few
seconds, let alone a few hours.

 Latency. It takes time for data to be sent to the cloud and commands
to return to the device. In certain IoT applications, these milliseconds
can be critical such as in health and safety. A good example is
autonomous vehicles. If a crash is imminent, you don’t want to have
to wait for the car to talk to the cloud before making a decision to
swerve out of the way.

The Internet of Things is a broad field and includes an incredible variety of

applications. There is no one-size-fits-all solution so you need to consider
your organization’s specific application when deciding whether the cloud
makes sense.

‍
Introduction to IoT Platforms
Whether you’re new to IoT or a seasoned veteran, you’ve probably heard
the term “IoT Platform” before. After all, there were over 300 IoT
platforms as of 2016 and this number continues to quickly grow (I’ve heard
there are now over 700). The IoT platform market is growing at a
compound annual growth rate (CAGR) of 33% and is expected to reach
a $1.6 billion market size in 2021.

IoT platforms are a critical component of the IoT ecosystem, but

we’ve found that for many people, it’s not clear what an IoT platform
is exactly or the differences between them.

So What Is an IoT Platform

Exactly?
IoT platforms are the support software that connects everything in an IoT
system. An IoT platform facilitates communication, data flow, device
management, and the functionality of applications.
With all the varying kinds of hardware and the different connectivity options
that you just read about in the previous section, there needs to be a way of
making everything work together and that’s what IoT platforms do.

IoT platforms help:

 Connect hardware.

 Handle different communication protocols.

 Provide security and authentication for devices and users.

 Collect, visualize, and analyze data.

 Integrate with other web services.

When Should Your Organization

Use an IoT Platform?
Because IoT is a system of systems, rare is the organization that has
expertise across all the relevant domains. IoT platforms exist to help
businesses overcome technical challenges without the need to figure it all
out in-house.
For example, your organization might be really good at building hardware
and decide that you want to make your hardware “smart”. Instead of the
expensive and time-intensive process of hiring software developers
to build everything in-house, you can instead use an IoT platform to
get up and running quickly and more cost-effectively.

However, there is a tradeoff. IoT platforms that save you time may cost
more in the long run depending on how they’re priced. This is because they
charge use-based and/or subscription fees that can add up over time. But
you still get the benefit of significantly lower up-front costs (no CapEx).

IoT platforms that are inexpensive up front will likely cost you time. This
comes back to the same point in bold above, the less you spend the more
work you’ll have to do on your own, which takes time.

In the next chapter we’ll explore how to choose an IoT platform.

‍
Choosing an IoT Platform
How do you know which IoT platform is the best? As much as we’d like to
give you a simple answer, as with most things, it depends. O’Reilly Media
puts it nicely:

“Each industry vertical — healthcare, manufacturing, energy, and banking,

to name a few — will present its IT and OT specialists with particular
conditions and problems to solve. Municipal police and fire departments,
for example, will depend on a platform that ensures communication
between field operations and command centers. Energy and transportation
companies will search for ruggedized solutions that will protect field assets
from harsh environmental conditions. Banking IoT platforms will
demonstrate robust encryption and security features that protect internal
and consumer communications and transfers.”

—Evaluating and Choosing an IoT Platform

That being said, there’s a key distinction between enterprise IoT platforms
and consumer IoT platforms. Consider:  is your application enterprise (such
as oil and gas, manufacturing, or asset management) or meant for
consumers (such as smart home applications or wearables)?
Enterprise IoT platforms and consumer IoT platforms can differ significantly
due to their different needs. For enterprise IoT platforms, a failure in the
system can be extremely high-stakes, perhaps costing millions of dollars or
even lives. For consumer IoT platforms, a failure might simply be an
inconvenience to the end-user.

And even within enterprise or consumer segments, applications can have

very different platform needs. However, despite the great variance in IoT
applications, there are some common elements that are critical to consider
when evaluating the best IoT Platform for your application:

1. The Stability of the Platform

With so many platforms out in the market, it’s likely that some will fail. It’s
important to choose a platform that’s likely to be around for several years,
otherwise your investment might go to waste if the platform provider folds.

Ask about current and past customers. If they don’t have any, that’s
probably not a good sign.

2. The Scalability and Flexibility of the

Platform
Your needs are going to change with time. Make sure that the platform
works when you’re small and just beginning, but will also work when you’re
(hopefully) large and growing fast.

In addition to being scalable, the platform should be flexible enough to keep

up with rapidly changing technologies, protocols, or features. Flexible
platforms are often those that are built on open standards and that commit
to keeping pace with evolving IoT protocols, standards, and technologies,
as well as offering third-party integrations and robust APIs (APIS are
covered in the next chapter).

It’s also important that the platform is network agnostic. This means that it
can integrate and work with all major tech systems out there, rather than be
locked into one vendor.

3. The Past Work of the Platform Provider

As mentioned above, IoT applications can vary greatly. If the platform
provider has done previous work that is similar to your application, that’s a
good indicator that they can meet your specific needs.

However, note that it need not be an exact match. If you’re building a smart
agriculture application, for example, you might look for a use case with
similar characteristics. That would be an application that also involves
hundreds or thousands of sensors/devices generating data, a similar
connectivity (such as LPWAN), and applied data analytics to create useful
insights.

4. The Pricing Model and Your Business Case

Make sure the platform provider is transparent in their pricing; some will
show an introductory rate and then hike that up significantly when you
actually go to sign up.

Also, how are you going to be selling? If you’re doing a subscription model,
then it makes sense to pay a subscription for the IoT platform service, since
you can wrap the costs into the pricing. However, if you’re selling hardware,
it might make more sense to pursue a platform option with an upfront
license so you can wrap that into the development costs of the hardware
product.

5. How Does the Platform Provider Handle

Security?
Security is absolutely critical to any IoT system and an IoT platform must
have security built into every layer.

When you ask about security, be on the lookout for: device-to-cloud

network security, user app-to-wireless network security, cloud security,
device security (including authentication and up-to-date certificates),
application authentication, data encryption, data protection (at rest, in
transit, and in the cloud), secure session initiation, and concrete plans for
updating security, including via over-the-air (OTA) communications.

6. Time to Market
In the previous chapter, we saw that one of the biggest advantages of
using an IoT platform is that it speeds up the time to market. Ask for a
realistic estimate of how long it will take to get to market and how the
platform provider intends to support you during that journey.

This is a big focus of ours at Leverege, which is why we introduced

the Jumpstart Package to rapidly speed up development and ultimately get
you to market with a better product/solution.

7. Data Analytics and Data Ownership

The value of the Internet of Things is in the data. Data can provide
actionable insights into operations or simple day-to-day activities to reduce
inefficiencies or improve experiences. You should look for basic descriptive
analytics, visualization, diagnostics, predictive analytics, and perhaps even
machine learning tools. We’ll cover analytics and machine learning in the
coming chapters.
Also, make sure to ask who owns the data. If the answer isn’t a simple,
“you own the data generated by your products”, this is a big red flag
because, again, the value of the Internet of Things is in the data.

8. Does the IoT platform Provider Care About

You?
In addition to all of these questions, you should be asking, also take note of
the questions that they ask you. Do they ask about your budget, timeline,
expectations, use cases, etc.? Do they seem like genuinely nice, caring
people?

This is one of the most critical considerations. A platform provider that

cares about you and your success will go the extra mile and make up for
any areas in which their platform might be lacking.

APIs
If there’s one thing you’ve learned so far, it should be that a complete IoT
system requires many different components all working closely together.
We’ve explored the hardware that collects data, the connectivity that sends
that data, and now the cloud and IoT platforms that ingest that data to
make it useful.
However, even at just the cloud level there is a need for systems to
communicate and work together and that’s what Application Program
Interfaces (APIs) make possible. This is especially important for programs
because they can be written in different languages, so APIs provide a
means for different programs to overcome the “language barrier”.

In addition, APIs mean that users of your system don’t need to leave your
system to use another organization’s application. For example, by using
weather.com’s API, you can request current weather data and display it on
your site or app for users. That way users can get weather information
without having to leave your site or app and go to weather.com.

Also, APIs reduce complexity. When you use an API to request something
from an application, many complex processes occur behind the scenes that
you don’t have to worry about. You just get whatever it is that you
requested in return.

A great example of this is the Alexa Voice Service API. Individual

developers can’t build Natural Language Processing like Alexa, but instead
they can use Alexa’s API to make tools based on it. So rather than needing
to figure out how to take speech and understand the meaning (which is
really hard), developers can focus on cool new applications that involve
voice control.
Key Takeaway
APIs are hugely important to the Internet of Things. APIs allow companies
to focus on their own expertise, plugging in the tools and programs of other
companies as needed to create an IoT product/service that’s greater than
the sum of its parts.

And this also means that you can build a business by creating an API that’s
extremely valuable for other organizations to use. If you’re a city, you may
have data on foot traffic or vehicle traffic that could help local businesses or
create new services to provide value to residents.

Data Analytics vs. Machine

Learning
With all the hype around machine learning, many organizations are asking
if there should be machine learning applications in their business
somehow.

In the vast majority of cases, the answer is a resounding no.

As you learned a few chapters ago, one of the major benefits of the cloud is
that it enables you to leverage virtually infinite storage and processing
power to gain critical insights from the data your sensors/devices will be
collecting. Both data analytics and machine learning can be powerful tools
in doing so, but there’s often confusion on what they actually mean and
when is best to use one or the other.

Later we’ll explore the value of machine learning in greater depth, but at a
high level, machine learning takes large amounts of data and generates
useful insights that help the organization. That could mean improving
processes, cutting costs, creating a better experience for the customer, or
opening up new business models.

However, most organizations can get many of these benefits from

traditional data analytics, without the need for more complicated machine
learning applications.

Traditional data analysis is great at explaining data. You can generate

reports or models of what happened in the past or of what’s happening
today, drawing useful insights to apply to the organization.

Data analytics can help quantify and track goals, enable smarter decision
making, and then provide the means for measuring success over time.
So When Is Machine Learning
Valuable?
The data models that are typical of traditional data analytics are often static
and of limited use in addressing fast-changing and unstructured data.
When it comes to IoT, it’s often necessary to identify correlations between
dozens of sensor inputs and external factors that are rapidly producing
millions of data points.

While traditional data analysis would need a model built on past data and
expert opinion to establish a relationship between the variables, machine
learning starts with the outcome variables (e.g. saving energy) and then
automatically looks for predictor variables and their interactions.

In general, machine learning is valuable when you know what you

want but you don’t know the important input variables to make that
decision. So you give the machine learning algorithm the goal(s) and then
it “learns” from the data which factors are important in achieving that goal.

A great example is Google’s application of machine learning to its data

centers last year. Data centers need to remain cool, so they require vast
amounts of energy for their cooling systems to function properly. This
represents a significant cost to Google, so the goal was to increase
efficiency with machine learning.

With 120 variables affecting the cooling system (i.e. fans, pumps, speeds,
windows, etc.), building a model with classic approaches would be a huge
undertaking. Instead, Google applied machine learning and cut its overall
energy consumption by 15%. That represents hundreds of millions of
dollars in savings for Google in the coming years.

In addition, machine learning is also valuable for accurately predicting

future events. Whereas the data models built using traditional data
analytics are static, machine learning algorithms constantly improve over
time as more data is captured and assimilated. This means that the
machine learning algorithm can make predictions, see what actually
happens, compare against its predictions, then adjust to become more
accurate.

The predictive analytics made possible by machine learning are hugely

valuable for many IoT applications. Let’s take a look at a few concrete
examples:
Machine Learning Applications in
IoT
Cost Savings in Industrial Applications
Predictive capabilities are extremely useful in an industrial setting. By
drawing data from multiple sensors in or on machines, machine learning
algorithms can “learn” what’s typical for the machine and then detect when
something abnormal begins to occur.

A company called Augury does exactly this with vibration and ultrasonic
sensors installed on equipment:

“The collected data is sent to our servers, where it is compared with

previous data collected from that machine, as well as data collected from
similar machines. Our platform can detect the slightest changes and warn
you of developing malfunctions. This analysis is done in real-time and the
results are displayed on the technician’s smartphone within seconds.”

Predicting when a machine needs maintenance is incredibly valuable,

translating into millions of dollars in saved costs. A great example is
Goldcorp, a mining company that uses immense vehicles to haul away
materials.
When these hauling vehicles break down, it costs Goldcorp $2 million per
day in lost productivity. Goldcorp is now using machine learning to predict
with over 90% accuracy when machines will need maintenance, meaning
huge cost savings.

Shaping Experiences to Individuals

We’re actually all familiar with machine learning applications in our
everyday lives. Both Amazon and Netflix use machine learning to learn our
preferences and provide a better experience for the user. That could mean
suggesting products that you might like or providing relevant
recommendations for movies and TV shows.

Similarly, in IoT machine learning can be extremely valuable in shaping our

environment to our personal preferences.

The Nest Thermostat is a great example: it uses machine learning to learn

your preferences for heating and cooling, making sure that the house is the
right temperature when you get home from work or when you wake up in
the morning.

Key Takeaways
The use cases described above are just a few of the virtually infinite
possibilities, but they’re important because they’re useful applications of
machine learning in IoT that are happening right now.

However, to reiterate, traditional data analytics are usually good enough for
most IoT applications. Don’t be fooled by an IoT platform selling you on its
machine learning capabilities when you’re just trying to look at trends over
time to measure and improve your efficiency.

To make one final, critical point: with both traditional data analytics and
machine learning, you need data. Gaining and maintaining large sets of
clean, relevant data is an essential prerequisite to unlocking all the value
that both data analytics and machine learning have to offer.

Chapter 5: UI & UX

Introduction to UIs & UX for

IoT
So far we’ve covered the sensors/devices that are out in the world
collecting data and performing actions. We’ve covered the connectivity that
enables those sensors/devices to send data to and receive data from the
cloud. And in the previous section we saw how that data is ingested and
transformed to provide valuable insights and automate processes.

But for any given IoT system, there needs to be a way of interacting with it.
And just as we saw with sensors/devices, connectivity standards, and IoT
platforms, there are many options you can pursue depending on your
specific application and business needs.

“We live in a time full of opportunity for imaginative individuals. In our

lifetime, we will witness the emergence of more and varied forms of
human-computer interaction than ever before.” - Learning and Thinking
with Things by O’Reilly Media

User Interface
Users need a way to view and understand the data captured by IoT. That’s
where the user interface comes in. In the simplest terms, a user interface
(UI for short) is the means by which a user and a computer system interact.
Many think of UIs as just software or apps on phones and computers, but a
user interface could be anything from a smartwatch to voice-controlled
Amazon Echo to the buttons on a smart tractor dashboard.
Apple pioneered the first graphical user interface (GUI) in 1983 with the
introduction of Lisa. A graphical user interface is a visual way of interacting
with a computer using items like buttons, windows, and icons. This meant
that people didn’t have to learn complex command languages to interact
with computers and thus made the computer more accessible to everyday
users. When they made the leap to touch interfaces and smartphone
technology in 2009, Apple helped to further open up the door for new types
of interfaces.

When it comes to mobile interfaces, there are a few important distinctions

that you should be aware of:

Native Apps
Native apps are what most people think of when they think of mobile UIs.
Native apps are applications that you download directly onto your phone.
The advantage of native apps is that you have greater access to the
phone’s capabilities and can create a better overall user experience (we’ll
talk more about user experience below). The disadvantage is that they can
take more time and resources to build, particularly because you need to
build for both iOS and Android (iOS is the operating system created by
Apple for iPhones and Android is an open-source operating system from
Google) since they’re not compatible with each other.
Web Apps
Like a website, a web app is accessed by going to a certain url
(e.g. https://examplewebapp.com). However, while websites are largely
informational (like Wikipedia), web apps are built to have certain
functionalities (like controlling a device remotely). The advantage of web
apps is that they can work on both iOS and Android because you’re just
using a web browser instead of actually downloading something. Also,
because you don’t need to download anything, it can make it easier to get
into the hands of users (just send them the url link). The disadvantage is
that you have somewhat limited access to the phone’s full capabilities (like
the inability to send push notifications) and less control over the overall
user experience.

Hybrid Apps
As the name implies, hybrid apps are between native apps and web apps.
You still download something, like a web app, but when you open the app it
is essentially opening a web page meaning that it can act like a web app.
This can be a good option if you know you’ll be creating native apps
eventually, but you want to get a minimum viable product into the hands of
users early, and can therefore benefit from the speedier development
offered by web apps.
Beyond Mobile Apps
The above distinctions are for mobile apps, but as we mentioned above
there are many different types of user interfaces beyond just mobile.

Chapter 6: Future of IoT

The Future of IoT

Congratulations! You’re nearly finished with this Introduction to IoT ebook.
At this point you should have a solid foundation in IoT, what it means, and
how to pursue building and deploying real IoT solutions for your
organization.

You might also be left feeling like you now have more questions than when
you began, and that’s a good thing. It’s the pursuit of these questions that
will lead you to build something that’s never been built before and to create
a better future for you organization and for all of us.

The future isn’t set. The future is the culmination of all the choices we are
collectively making today. While it’s inevitable that we’ll continue to connect
our things, people, and environments to make them more intelligent,
efficient, and user friendly, it’s not inevitable which problems we focus on
solving and who benefits from those solutions.
At Leverege, we’re committed to amplifying human potential and we see
IoT and related technologies as essential tools to make that happen. We
strive to make those tools more accessible to everyone and to share the
knowledge we’ve gained through our decades of experience, so that all
businesses are empowered to unlock human value in and out of their
organization.

Thank you for reading! If you still have a burning question, if you think
there’s an opportunity to work together, or if want to just say hi, please
don’t hesitate to reach out.