With the high demand for IoT in the industrial sector and beyond, there has been an exponential growth in the number of IoT platforms. So what types of IoT platforms can you find on the market right now and how can you decide which is right for you?
You may need an IoT platform that only takes care of the connectivity or the management of IoT devices within your network, with or without advanced analytics.
But you may also require app development capabilities and the option to deploy your machine learning models right at the IoT edge.
The graphic below provides an overview of the IoT architecture, giving you an idea of the architecture layers that IoT platforms cover:
What are the key components of an IoT platform?
IoT platforms are multi-layered solutions. They can greatly simplify the consolidation of many different IoT devices and come with a variety of services. The building blocks of an IoT platform typically include the following IoT architecture layers:
Connectivity is the one absolutely necessary component of any IoT development effort. And, on a very basic level, an IoT platform can be as simple as a software infrastructure enabling and overseeing the connectivity between objects in the physical world.
Within today’s complex connectivity environment, solid connectivity management will allow the secure and effortless incorporation of diverse IoT devices as well as the administration of multi-protocol & multi-network connections.
Other than that, every IoT platform has a connectivity layer that enables a variety of services. These include but are not limited to data collection, monitoring & control, as well as the management of assets.
IoT data management
Data management is the practice of developing, executing, and overseeing architectures and software that increase the value and availability of data. This includes but is not limited to logging, storage, processing, and transaction management. IoT data management takes care of data coming from many different locations and many different data types. It covers data extraction, storage, processing, and analysis.
Cloud computing services
Handling large amounts of data in commercial IoT settings usually involves a cloud backend that reliably and securely streamlines data collection, storage, data processing, and analytics while performing service orchestration on top.
IoT device management
IoT device management may cover the entire lifecycle of an IoT device. It involves device registration, device identification, updates, authentication, logging, remote administration, troubleshooting, rebooting, and device diagnostics. This is how you oversee all of your connected physical assets from within one platform.
A recent thing in the world of IoT platforms is the ability to code directly in a cloud IDE fully integrated into the platform environment. Platforms such as Portainer.io, Particle.io, and Record Evolution make it possible to write apps and deploy them in seconds while getting live feedback from the device and observing app performance. Over-the-air development (OTA) is currently offered by the fewest IoT platforms and is especially well received in the developer community.
IoT application enablement
Essentially, this means providing action management to define rules and triggers for automated jobs and services, including the provision of application logic to coordinate the flows of data and execute certain actions. Thanks to this capability, developers can build, integrate, and manage custom applications.
Security by design
Most IoT platforms on the market right now come with built-in security features. These guarantee high standards of privacy and identity management across networks, devices, and cloud services. Data integrity, the extra protection of certain proprietary data types, authorization mechanisms, safeguarding privacy, authentication, identification, and encryption are all part of the security layer of the IoT platform and can operate on a variety of levels.
What types of IoT platforms exist on the market right now?
As different markets have different middleware needs, IoT platforms can be classified based on the services provided to end-users and their place within the IoT value chain. Platforms that meet the needs of consumers, for example, will have capabilities different from those of platforms made for industrial IoT applications or those tailored to the automotive, manufacturing, finance, agriculture, or energy sectors.
We explain the difference between generalized IoT platforms and industrial IoT platforms in the article “An IIoT Platform, an IoT Platform, or MES: What’s the Difference and How to Pick One?”
Today, it has become standard for an IoT platform to provide real-time intelligence, M2M communication, interoperability, and end-to-end security by design. Even more so, IoT platforms are now also expected to have AI capabilities and machine learning.
IoT platforms fall into the following categories:
IoT platforms focusing on connectivity management
Again, this is one of the most basic yet highly specialized and widespread types of IoT platforms. It involves connectivity orchestration, connection management, plus various communication services for IoT devices. Such platforms also take care of invoicing with the network providers.
Connectivity management platforms have the following capabilities:
- Connection management,
- Invoicing/billing management,
- Connectivity between sensors and servers,
- Management of data rates.
IoT platforms focusing on device management
Remaining on top of growing fleets of diverse assets is one of the most common concerns in the world of IoT. Therefore, IoT device management platforms have gradually established themselves as standalone products over the past years.
Some of the capabilities of IoT device management platforms include the following:
- device provisioning,
- device authentication,
- remote device monitoring and control,
- administration over the air,
- software updates,
- security patching,
- troubleshooting, and more.
IoT platforms with a strong focus on end-to-end analytics
Collecting and making sense of IoT data calls for a robust IoT analytics infrastructure that not only takes care of the harvesting and processing of massive data volumes, data mining, and building ML solutions, but also involves data storage capacities for structured and unstructured data.
The key capabilities of IoT analytics platforms involve:
- IoT data collection via compatible communication protocols,
- data processing, including transformation, modeling, and the creation of long-term data histories,
- scalable data storage clusters to accommodate both structured and unstructured data,
- advanced data analytics / Big Data analytics capabilities to identify patterns and relations within the IoT data,
- flexible reporting and data visualization tools that can be easily customized.
Application enablement platforms
Application enablement platforms specialize in the development of custom IoT solutions with a focus on data processing and insight generation. These platforms usually have many built-in applications and also come with interfaces to external applications or to the user’s own software.
Some of the functionalities of an application enablement platform include:
- rule engines,
- highly customizable data visualization tools,
- the abstraction of the underlying infrastructure.
The IoT orchestration hub
Instead of a simple connectivity management platform or a device management platform, the majority of IoT platforms on the market right now are a combination of multiple services around IoT. These can be comprehensive at varying degrees. At times, they cover only a portion of the value chain and sometimes provide an end-to-end solution in just one offering.
The IoT orchestration hub is where all services around the IoT solution come together. They ensure connectivity, seamless device management from a single location, plus transparent data management and data processing. As hubs usually come as cloud services, their scalability also plays a role. Orchestration hubs are usually built as open solutions and can integrate with a variety of IoT services on top.
The key capabilities of IoT orchestration hubs include:
- communication management,
- forwarding capabilities,
- IoT device management,
- integration with analytics services.
The IoT development platform
Some IoT platforms strive to improve the developer experience and come with a variety of services that make it especially easy for developers to collaborate and develop applications.
Platforms such as Siemens Mindsphere and Record Evolution additionally provide developers with a marketplace, or “a platform within the platform”, where they can offer their own apps and app templates.
Development-centric platforms also place an emphasis on community-building and IoT collaboration.
Some of the capabilities of developer-focused IoT platforms include:
- OTA application development in a cloud IDE,
- live development and deployment of IoT apps,
- integrations with external repositories such as GitHub & GitLab,
- a marketplace for IoT apps and app templates,
- collaborative & community features.
The end-to-end IoT platform
This is a full-stack platform that consolidates and performs all services across the IoT value chain. Ideally, the end-to-end IoT platform is suitable for a variety of use cases and transcends industry borders. It should be robust enough to accommodate large-scale industrial IoT solutions (think Thingworx). But it should also remain flexible and open to allow for third-party integrations and anticipate future developments.
The end-to-end IoT platform takes care of the entire IoT value chain. It may cover all of the elements listed below:
- security by design,
- protocol and data structure interoperability,
- IoT device management,
- configuration & update management,
- data storage clusters & scalability,
- advanced data analytics and ML,
- reports with data visualization and dashboard customization options,
- user management with a highly granular authorization structure to enable IoT collaboration,
- a cloud app development studio,
- ideally, an IoT app store with ready-made IoT apps and app templates,
- services that allow for IoT app reuse and duplication,
- open APIs to allow for further integrations.
So we see that the IoT platform landscape is complex and multifarious. While we speak of various types of IoT platforms, the boundaries between these are not necessarily clear-cut. You can have an end-to-end platform that is also a development suite all the way down to IoT deployment, an IoT hub that includes an analytics platform, an enabler platform allowing you to build your own IoT product, or simply a connectivity platform with some device management capabilities. The possibilities are varied and the one thing to guide you within this jungle are the necessities posed by your specific use case.
About Record Evolution
We are a Data Science & IoT team based in Frankfurt am Main, committed to helping companies of all sizes to innovate at scale. So we’ve built an easy-to-use development platform enabling everyone to benefit from the powers of IoT and AI.