IoT technologies and smart spaces are expected to start a new technological revolution, but while billions of IoT devices exist, the lack of unification across these devices continues to hold the industry back. What challenges do IoT devices face, what solutions can help, and what can engineers do?
IoT’s Worst Enemy: Fragmentation
It is truly amazing to see just how many IoT devices there currently are around the world, with current estimates putting the total count above 14 billion — this is expected to reach 20 billion by 2024. Not only is it hard to even comprehend the number of devices globally, but the amount of data being produced is also exceedingly gargantuan. Thanks to the developments made in IoT technologies and the amount of data being generated, industries such as AI and machine learning have been able to accelerate and expand massively, resulting in AI now becoming commonplace in everyday life.
And yet, despite the many IoT devices on the market and the numerous smart home options, most IoT devices used in homes are far from smart. In fact, most IoT devices used in homes are only classified as IoT simply because they can form an internet connection, and they certainly do nothing smart with the inability to share their data with other systems, nor can they intelligently respond to events in real time. Furthermore, the vast majority of homes integrate no smart technologies by default — making most homes dumb.
While there are numerous factors that have made IoT-home integration difficult, by far the biggest culprit comes down to fragmentation.
What Is IoT Fragmentation?
There is a joke in the IoT industry that perfectly describes the problems faced by fragmentation. Simply put, an engineer recognizes that there are 14 IoT standards and decides to create a standard for everyone to use — now there are 15 standards.
Unfortunately, this joke is very much how reality operates, and the hundreds (if not thousands) of IoT companies and engineers who design and manufacture IoT devices do exactly this. Not only do many of these companies create their own standards (thinking they are the best out there), but they will often create entire ecosystems of solutions that work inside their custom framework. While this may be good for those who solely use one manufacturer for their IoT solutions, it introduces numerous challenges that hold back the integration of smart spaces.
The first challenge that this presents is the difficulty in creating complex IoT solutions that combine devices across different manufacturers. For example, a smart home will require numerous smart devices ranging from a smart doorbell, smart power sockets, and smart lights to smart thermostat controls. Despite all of these devices existing on the market, trying to find one single manufacturer of all these devices or different manufacturers that support a common protocol is just about impossible.
The second challenge that IoT fragmentation introduces is the lack of software support for smart homes. Just like how each manufacturer creates their own standard, they also often create their own software platforms used to control devices and read data, but these will rarely expose an API for other manufacturers to use. Thus, even the software side of IoT devices can be fragmented and walled off from other developers.
Finally, customers who are forced to use a single manufacturer for their IoT solutions can be trapped in the brand — or worse, be beholden to any and all decisions made by that brand. For example, Hive recently announced that it would be ending its range of HomeShield products that includes security cameras and leak detectors. Once updates and services are stopped for these devices, customers will have no choice but to remove and discard perfectly usable IoT devices.
What Is Matter?
Recognizing the numerous challenges faced by IoT devices, a group of large tech companies has come together to say enough is enough and has now developed a new IoT protocol: Matter. Even though the release of Matter still follows the previous joke of creating new standards, by having many major tech companies support the platform (Google, Meta, etc.), it essentially forces manufacturers to join the standard due to the sheer size of the tech companies involved.
At its heart, Matter is a proprietary standard for home automation that provides royalty-free licenses to manufacturers with the only cost being for certification. Despite the cost of certification, Matter is entirely open source to the public, which enables users to see how the code works as well as make suggestions for edits and bug fixes. While it isn’t clear what the penalty is for using Matter without obtaining a certification, it is likely that products simply cannot bear the Matter name and logo (i.e., a product could be compatible with Matter, but wouldn’t be able to advertise it as such).
With the Matter projects starting in 2019, numerous manufacturers and companies have since announced their public support, which includes IKEA, Amazon, Google, Apple, and Zigbee Alliance. However, as the protocol was only officially released in October 2022, it is yet to become mainstream or be adopted by the wider community. Furthermore, the current version of Matter supports common household devices including lighting products, mains switches, thermostats, door locks, heating systems, ventilation, security sensors, televisions, and blinds. The next version, due to be released in March 2024, will expand the list of supported devices to include robot vacuum cleaners, CO sensors, smoke alarms, energy management, Wi-Fi, cameras, and other major appliances.
Are There Any Alternatives to Matter?
As of now, there are no IoT standards that have the same degree of backing as Matter does, but that isn’t to say that there aren’t other standards that can help engineers create well-connected devices.
By far some of the more popular methods for IoT communication are REST and HTTP, and this is because most web servers naturally support them. This means that IoT devices can send packets of data to a standard web server coded in PHP or Node.js, and the added bonus of these protocols is that it makes viewing data very trivial as databases used to power websites can also be used to store IoT data. Thus, an entire IoT solution can, in theory, be integrated into a single web server.
An alternative to HTTP and REST is the use of web sockets. PHP, HTML, and JavaScript all support web sockets, which makes them an excellent choice for IoT devices; but where HTTP and REST are single-packet messages, which closes their connection upon completion, web sockets can remain open, which makes them great for streaming large amounts of data.
Finally, MQTT is another popular protocol used by millions of devices worldwide. Initially developed for the oil and gas industry, MQTT allows for entire networks of devices to publish and subscribe to specific variables, and the lightweight nature of MQTT makes it ideal for low-energy and low-end microcontrollers. Furthermore, custom implementations of MQTT can easily be made due to the simplicity of the protocol, and there are numerous examples online that can be adopted.
What Should Engineers Do?
In both HTTP and MQTT, designers can expose their API to at least provide the option for other manufacturers to interface with devices. Furthermore, bridging software — that which can connect two different standards — can be formulated, which only helps products work with other manufacturers.
But besides this, it is difficult to say what exactly engineers should do as Matter itself is very young, the documentation for it is very in-depth and complex, and Matter may not be the best solution for all IoT products. Of course, creating new standards has clearly been proven to be of no help whatsoever, so engineers should avoid creating new standards where possible.
One possible solution is for engineers to focus on open-source designs that can be easily changed in the field. For example, creating IoT devices from Linux systems that support updates over the air can, in theory, be updated in the future to support Matter when it becomes more established as a protocol.
Another solution is for engineers to stick with mainstream solutions such as HTTP and REST, and hope that by providing their APIs to the public that Matter will be able to support their products to some degree. In fact, it may be possible to create a web interface that sits in between the IoT device and a Matter server that can translate messages between the two.
But just because Matter exists doesn’t mean it should be used. Unless it is essential for a design, simply integrating Matter for the sake of supporting it may do more harm than good — especially if unknown bugs plague Matter.
