Internet of Things. Chances are big that you have a “thing” in your house, your office building or your factory that is connected to the Internet. But what are they really? A commonly accepted definition for Internet of Things (IoT) is:
“The Internet of things (IoT) describes the network of physical objects—a.k.a. "things"—that are embedded with sensors, software, and other technologies for the purpose of connecting and exchanging data with other devices and systems over the Internet”.
(source: https://en.wikipedia.org/wiki/Internet_of_things)
But that doesn’t really paint a picture. We’re talking physical devices, sensors and transmitting data. IoT is best described by a few examples. Think of the following:
- Activity trackers you use for sports;
- Home Security;
- Temperature sensors (thermostat).
These are usually devices you wear or install in and around your house and you can usually identify them as they are often called “Smart” devices. These are just common known uses for IoT but the amount of use cases are enormous. And to be fair.. Smart devices or systems you install at home are just the tip of the ice berg. Expectations are that by 2025 there will be more than 64 Billions IoT devices (source: 47 Internet Of Things Statistics to Show How Big It Is in 2021).
The use cases
We can talk about how smart cars are and that they are always connected to the internet, or how security gates at the mall can register how many people are inside and that telemetry is send somewhere for processing, or how factories optimize and automate processes using IoT but it’s best to start with simple examples.
If we go back to the definition, it states “a network of physical objects” and that’s where it starts. These physical devices, are connected to the Internet and more often a public cloud like Microsoft Azure specifically. These devices have the capability to connect to a service (IoT Hub in case of Azure) and allow themselves to be either managed by that service or just send data (telemetry) to that service.
What makes it interesting is if you start connecting sensors to that device. Let’s take a temperature sensor for example. We can configure that device to check the temperature sensor every 5 seconds and send the actual temperature to IoT Hub. We can then configure IoT Hub to store the data somewhere and build a report that displays the temperature of that room (and create a nice graph while you’re at it).
And that is just a very basic example. In the meanwhile there are loads of initiatives that focus on building so called smart cities. Cities with logic powered by IoT devices.
The power of these devices is that they are generally small, cheap and do exactly what they need to do and just that. In fact, you can learn how to set up your own IoT device and connected it to Azure within hours (Introduction to Azure IoT - Learn | Microsoft Docs).
But let’s take it a step further. Once you have gathered all that data, what will you do with it? Let’s say you have an office building with 150 offices. Each office has one temperature sensor. If you connect them all, you are now gathering telemetry from 150 temperature sensors. That’s a lot of data. But just having data doesn’t mean a thing unless you do something with it. Maybe you want to create a dashboard for the building manager to show the temperature in different offices, maybe you want to create an alert when the temperature drops below a specific value or maybe you just want to identify that one office where a person leaves the window open during the day causing a drop in temperature for the surrounding area. And why stop at temperature, maybe you want to enrich your data set with telemetry from other sensors like motion detectors or weather stations.
There are unlimited examples when it comes to IoT. Perhaps you’re in the agriculture business and you have sensors that measure the moist of the ground. What if we connect that to a watering system? Sounds like technology and IoT specifically can make a lot of peoples life easier.
How do you get started?
Well that’s the thing. You can get started really easily by just connecting a device to Azure IoT Hub. But a bunch of devices connected to the cloud isn’t really going to run your business now is it?
You need to plan and decide on the type of architecture you need and there are a lot of questions to be answered before you can start doing that, going all the way back to the device you want to use and the environment. You need to think about the location, the type of data, how much of that data you need, when you need it, how fast does it need to be processed? And.. What if someone connects a new device? Do we want to send an engineer over to configure it? Probably not!
Answering these questions can lead to architecture choices. Maybe you are connecting a camera and you just need to know if an event occurred. You don’t need the image or video stream itself just a simple “yes or no”. That means you might want to use a machine learning model that you can run on your IoT device, have it detect said event and only send a message every minute to the cloud saying “event happened yes or no”. This saves a lot of bandwidth and if you have hundreds of devices, you probably don’t want hundreds of video streams being processed by your IoT solution in the cloud.
Long story short. The answer to getting started is not simple. In the upcoming weeks we will publish a series of blog posts where you go through different scenario’s and what that solution would look like both on the device said as well as Microsoft Azure. We will discuss architectural choices both on the software and infrastructure level. Sign up for our Intercept Insights here and receive the next IoT articles in your mail. Stay tuned and feel free to drop us an e-mail with questions if you are looking for a specific subject!