What Is an IoT Framework vs an IoT Platform
We were collectively using around 42 billion Internet of Things (IoT) “things” in 2022. We’re expecting to use almost double that by 2025. That includes everything from smart fridges to wearables for real-time monitoring to apple-picking drones. That’s why McKinsey predicts the total economic impact of IoT that year could be north of 11 trillion dollars.
It’s no surprise then everyone wants to connect their disconnected devices and develop IoT applications — but that’s not easy. We’ve written before about the failure rate of IoT initiatives (it’s as high as 75%), so you probably don’t need me to tell you how important it is you pick the right IoT framework… or do you need an IoT platform?
Which IoT solution is the right one (for you)?
What exactly counts as an IoT framework vs an IoT platform is debatable.
As of July 2023, the top result on Google for “what is an IoT framework” identifies a wide range of them, including (the deprecated and soon-to-be discontinued) IBM Watson IoT.
Yeah, well, IBM might disagree with that, since it repeatedly refers to Watson IoT as a platform on its own website.
A framework or a platform? In 2024, it won’t matter anyway!
Still, knowing that there are some generally agreed-upon differences can help inform your decision-making and ensure you make the right choice for your next IoT project.
What is an IoT framework?
One of the common ways people describe IoT frameworks is that they’re a lot like a “blueprint,” a kind of design plan for engineers to follow.
When we think of it like that, it’s easy to see how communication protocols could be considered the archetypal example of a framework in an IoT context. Communication protocols are the guidelines IoT ecosystems use to transmit device data between different nodes on the network (and, eventually, into the cloud). These nodes could be the devices themselves or IoT applications like admin gateways and user-facing mobile apps. Choosing the communication protocol you’ll use is like choosing whether you want to mail a package (data) with FedEx or USPS (MQTT, AMQP, and others).
But there are other ways to conceptualize IoT frameworks that make room for more, let’s call them, robust solutions. These might be things like cloud-based storage or message brokers. If communication protocols are like different shipping services, message brokers are like letter carriers and mail trucks.
It’s probably safe to think of anything calling itself an “IoT framework” as a component part of your complete IoT ecosystem and as strictly technical infrastructure. They help you connect to cloud services and facilitate communication between devices. Generally speaking, if you’re “building an IoT solution from scratch,” you’ll still rely on and use frameworks — they’re that foundational. Here are some specific examples:
EMQX describes itself as the most scalable open-source MQTT broker. MQTT is one of the most popular communication protocols for IoT because of its flexibility and strong feature set.
Node-RED is a “flow-based” program editor specifically designed for IoT use cases. According to its website, its browser-based GUI enables engineers to connect “hardware devices, APIs, and online services in new and interesting ways.”
RIOT is an operating system for low-power IoT devices. It’s particularly well-known for its passionate developer community (RIOTers) and annual RIOT summit.
What is an IoT platform?
If a framework is a piece, then we might think of an IoT platform as more of a whole. It’s an abstraction layer that builds upon underlying frameworks and includes multiple (including UI) components.
Case in point: Onymos IoT consists of software and services that support data collection, device management, monitoring, and the option to license all of the actual source code.
IoT platforms can differ extensively from one another. Some include actual hardware devices that enable IoT connectivity. Others focus on machine learning-powered software services for real-time data processing and analytics.
An open-source IoT platform might be more cost-effective but require significantly more expertise to actually use. Low-code/no-code ones prioritize for their non-technical users’ ease of use but sacrifice innovation and flexibility. Pro-code ones like Onymos IoT put a greater emphasis on data security, ownership, and developer enablement.
Ultimately, how (e.g., using a platform or one or more frameworks) you bring your connected devices to market should depend on your unique business requirements and your team’s expertise.