Serverless Architecture for Internet of Things

The Internet of Things, or IoT, refers to the billions of internet-connected devices around the world that use sensors to collect real-time data remotely. As we’ll see, IoT use spans industries in a wide range of applications. Some high-profile examples include wearables for fitness tracking, smart home devices, security monitoring, farming, industrial equipment, weather tracking and connected cars—but there are so many more. Depending on the application, the data collected from these devices can be used to perform analytics, provide insights, issue alerts, drive decisions, and more.

IoT devices are everywhere and their usage continues to grow. While the number of IoT connected devices was expected to reach nearly 14 billion in 2021, that figure is projected to more than double to almost 31 billion in 2025.1 And, by 2024, the global IoT market is predicted to surpass a trillion dollars annually.2 Yes, that’s “trillion,” with a “t.”

With that kind of growth, IoT devices need a supporting IT infrastructure that can ingest and handle a continuous flow of real-time, often time-series data. Before you know it, your organization could quickly accumulate hundreds of terabytes, or even petabytes, of data. Trying to scale operations internally to keep pace, without sacrificing performance, could put a strain on your team and resources.

That’s where serverless can help. It’s a fully managed approach that shifts the responsibility for provisioning, managing, and scaling resources to a third-party cloud provider. The best serverless platforms feature elastic scalability, always-on availability, reduced costs, and security. You’ll be able to outsource the skills and resources needed to implement and maintain operations anywhere in the world, while ensuring all local regulations are met.

In this article, we’ll share key IoT requirements, along with some example applications. Then, we’ll go deeper on serverless and why it makes so much sense for IoT.

Understanding IoT applications

IoT sensors are continuously collecting real-time data in the field—often with a time-series element. And some devices may have multiple sensors collecting different types of data. All that data flows in and is integrated into the platform, where analytics are applied. Patterns emerge and insights are uncovered. This is the type of useful information that is delivered to the end user via the IoT application’s user interface. The type of that valuable information varies by each application’s purpose, but it could be a recommendation, an alert, a status update, or any of hundreds of other use cases being tackled by IoT.

To get a feel for their variety, here are a few examples of common IoT applications:

Smartphones: If smartphones were the only type of IoT device, the market would still be massive. They are ubiquitous and each one is loaded with applications that track and share data for a wide variety of purposes.
Smart homes and appliances: Devices like the Amazon Echo, Google Nest Thermostat, and Ring doorbell have become commonplace in many households. IoT in the home can take the form of automated lights, security systems, refrigerators, microwaves, washing machines, and much more.
Wearables for fitness and health: You can collect and track data about your workouts, heart rate, number of daily steps, and more with fitness wearables like the Fitbit Charge, Garmin Forerunner, Amazfit Bip, or Apple Watch. There are also plenty of wearables available to measure health and wellness metrics like ECG, blood pressure, blood glucose level, and quality of sleep. In some cases, your doctor’s office could be alerted if a measurement raises a red flag.
Smart cities: Data from IoT devices is now being used in city planning to combat problems like noise, pollution, traffic, and crime.
Connected and self-driving cars: Cars are starting to be loaded with sensors, so we can take more actions remotely, like turning on the A/C and cooling down the car before getting in, unlocking the car or popping the trunk so a neighbor can grab something, setting off the alarm, and more. Smart cars can also alert you if the air pressure in one of your tires is getting low or if you’re approaching the car in front of you too fast. And, as self-driving cars continue to evolve, they are loaded with sensors and cameras to adjust to the world around them.

And that’s just a very small sample. There are too many IoT examples to list. A few more include tracking water quality levels, inventory, crops, and even livestock.

Requirements for Internet of Things applications

Let’s review what’s required for IoT applications to run smoothly.

Scalability: While the overall IoT market is experiencing explosive growth, the same could happen with your company’s own IoT arsenal. To be ready, you need an architecture that will quickly and seamlessly scale, when your number of users, devices, or associated data volumes suddenly skyrocket. When planning for future capacity, it’s easy to get it wrong. If you allocate too little, your infrastructure will be overloaded. If you secure too much, you have idle resources, unnecessarily wasting funds.

Security: With data flowing in from IoT devices that could be located anywhere in the world, there’s a high probability some or all of it will be sensitive. It could be data that is highly valuable and proprietary to your organization, or it could be personal information your customers have trusted you to keep secure. Without outside help, it’s your responsibility to keep this influx of IoT data properly handled. A breach could severely hurt your company’s bottom line and reputation.

Control: Your company’s data could very well be its most valuable asset. IoT collection and analysis can yield insights that lead to competitive advantage. Don’t settle for any solution that doesn’t allow you to have complete control over your own data, with the ability to migrate it whenever and wherever you please.

Performance: Much of the value to be gained from IoT data depends on its timeliness. So, the effectiveness of IoT devices largely relies on the responsiveness of the system. Once data is integrated into the platform, any needed analysis needs to be performed quickly in order to get the information and insights into the hands of customers while still relevant, meaningful, and potentially actionable. A platform is needed that can not only handle millions of devices and elastically scale with any unexpected spikes, but one that can do all of that without slowing things down.

Flexibility: With IoT, it’s not uncommon for technology requirements to change. You need a platform that gives you the flexibility to quickly make adjustments when necessary. Also look for an offering that is multi-cloud, so you don’t risk being locked into one environment. Your IoT network could include all kinds of connected devices that are collecting a wide range of data. So, you’ll also want a database that has the flexibility to handle many different data models.

Availability: The success of an IoT system depends on the continuous flow and exchange of data. That being the case, going offline could be disastrous. A gap in time-series data, for example, could invalidate results. And, when it comes to smart products, customers expect access and any needed information to be available at all times. The bottom line? IoT platforms need to be continuously available, with no single point of failure.

Now, let’s see if the advantages serverless platforms provide are a good fit to handle these requirements.

Benefits of serverless for IoT applications

IoT requires real-time responsiveness and the agility to support sudden spikes in device growth and ultra-high volumes of data. If the system goes down at any time, valuable data will be lost. Serverless databases are a natural fit to meet these speed, scalability, and availability needs.

The term “serverless” isn’t literal. The servers don’t magically vanish. But, when you adopt a serverless approach, many of the headaches associated with their operational issues disappear. With serverless, you offload the responsibility of deploying, managing, and scaling servers, along with data security, to a cloud provider. That frees you and your team up to focus on building new products and services.

With serverless, you no longer need to guess and worry about scalability or having enough capacity for your IoT network. Serverless databases automatically provision resources to meet even sudden and dramatic increases in demand. Just as important, they scale back when things slow down. Your company will only be charged for what is actually used. Contrast that with operating dedicated servers at all times—even when they weren’t used– running up the bill in the process.

Serverless even features scalability at the IoT device level. Each IoT device will have different load requirements and those will continuously change on a moment-to-moment basis. Serverless provides the flexibility to seamlessly expand or reduce the resources needed for each device on an ongoing basis.

Extremely high availability is another calling card of serverless platforms. Serverless databases are distributed, with data replicated across data centers. If one server goes down, another takes its place and the IoT system will continue to operate normally. Your customers will never know there was an issue.

Having data distributed in regions across the world also boosts speed and responsiveness. The information any given customer or end user is accessing is much more likely to be geographically closer to them than if you were taking an in-house approach. Serverless also makes it much less expensive and resource intensive to globally expand your infrastructure. You simply lean on the infrastructure one of the top cloud providers already has in place.

Serverless databases also feature real-time responsiveness at any scale. But the speed advantage doesn’t stop there. A serverless architecture also enables development teams to get to market fast, deploying in minutes, using only needed computing resources.

Moving to a serverless database to support IoT initiatives, will allow you to offload all the backend work and data storage necessary to ensure constant, uninterrupted connectivity. It also alleviates the costs of physical servers and the skilled staff that would otherwise be needed to make it happen. Instead, capital expenditures can be made on increasing the number of devices in the IoT arsenal. Going serverless also opens up a whole new world of SaaS tools and cloud services to aid IoT development, connectivity, and security.

Take the next step with serverless IoT

The potential scale, availability, security, and speed required to successfully implement and run an IoT system can be overwhelming. Fortunately, serverless platforms provide much needed relief, while taking time consuming and tedious infrastructure management and maintenance off your hands.

Take the next step: Learn about DataStax’s multi-cloud, multi-model serverless database, Astra DB, and try it for free.

Also, make sure to take a look at the benefits a serverless architecture can provide e-commerce and financial services applications.

—

[1] Lionel Sujay Vailshery, “IoT and non-IoT connections worldwide 2010-2025,” Statista, Mar 8, 2021.

[2] “Internet of Things – Thematic Research”, GlobalData, May 31, 2021.