IoT platform decoding: three types you should know

We have all heard that the IoT platform is the key to the IoT ecosystem, but what exactly is an IoT platform? If you search the term online, you can easily find millions of results, but in the end you may still be confused. The fact is that not all IoT platforms are the same, and the difference between them is not just in the scope of use cases they are aimed at.

IoT platforms can be classified according to their core functions. Each platform type has different uses and provides different functions and services. Usually, different platform types can interoperate in the Internet of Things system to provide the full set of functions required by the end user. Understanding the nuances of the IoT platform is essential to determine which options are right for your business. Here are the three main platform types you should know.

1. Network and device management platform

As the name implies, such platforms provide users with a tool to easily set up their edge networks, as well as connect and manage devices on a large scale. The key functions of the network and device management platform are not refined data analysis and visualization, but device security, network monitoring and diagnostics, and automatic software / firmware updates on the device. At the same time, it provides an overview of all input traffic, as well as key hardware indicators (such as device ID, sensor type, gateway CPU and memory load, etc.), so you can always understand the network operating status and device status.

Depending on the provider and the type of network connection, such platforms can run in the network operator's proprietary cloud, in the public cloud, or in the internal data center. Therefore, it is necessary to study which deployment options are available and whether they are consistent with your IoT strategy. For example, cloud deployment brings high resilience and scalability, and the flexibility to manage the network from anywhere, and when strict data privacy and ownership requirements are required, local deployment is ideal.

MYTHINGS Central is an example of a network and device management platform.

2. Cloud Platform / Infrastructure as a Service (IaaS)

Cloud platforms or IaaS solutions provide inexpensive and scalable computing infrastructure, such as servers and storage repositories, so users can bypass the cost of installing and maintaining their own data centers. Although cloud platforms provide the ability to receive and store massive data streams, there is no tailor-made IoT application to analyze this raw data and convert it into user-friendly and actionable insights.

In addition to the back-end infrastructure, many cloud vendors have expanded their product range to include basic rule engines and tools that enable third-party developers to build and host their applications in cloud environments. Nevertheless, there is still much in-depth development work to be done. Therefore, end users must either implement a third-party IoT application platform or write their own enterprise applications on top of the cloud infrastructure to truly understand the data.

Microsoft Azure, Amazon Web Services or IBM Watson are typical examples of IoT cloud platforms.

3. Internet of Things Application Platform

The Internet of Things application platform is equipped with built-in algorithms that can be used for various services from basic data interpretation to complex machine learning and predictive analysis. With an intuitive visual dashboard with 2D or 3D graphics, users can easily identify trends and patterns to take action on their data. Alerts based on events and / or predictions can also be set up to notify abnormal behavior and facilitate task execution.

Some IoT application platforms are completely out of the box and customized for a specific set of IoT use cases. Offered as cloud-based software or "software as a service", they are usually bundled with a supported device in a product package. Users do not have to do any top-level programming or development, but there are some limitations in the interoperability of devices and network connections. Some even lack the ability to integrate with external systems for further data transmission. (Source Internet of Things) Although not all, many of these platforms are used in smart city and consumer IoT environments. In these application environments, availability and readiness are important. An example is the Samsung SmartThings platform for home automation.

On the other hand, IoT application platforms for large-scale industrial and commercial deployments are mostly designed to be independent of devices and industries, and can support data in different formats. These are called Application Support Platforms (AEP). Generally, powerful application templates can be used for instant data analysis and visualization in key IoT use cases. At the same time, these platforms provide tools and services that enable developers to build and manage multiple customized applications with little cost and complexity-many times, they don't even need to write code. The deployment model is usually flexible, and it is easy to integrate the platform with existing business systems and other IoT platforms.

to sum up

As you can see, each of the IoT platforms discussed above is specific to the domain, and in most cases, it ’s not just one of them. Instead, they should interoperate with each other through open interfaces and messaging protocols to deliver all the key building blocks of the IoT platform architecture. Specifically, the application support platform can interface with one or more network and device management platforms to aggregate data from different sources for analysis and visualization in various use cases. Both can be hosted in a cloud infrastructure, thereby taking advantage of massive data processing and scalability at a very low cost. Through seamless integration, end users can easily manage all aspects of their network from a central interface to simplify operational tasks.