KNX IoT and Matter: “There Are Advantages to Not Being First to Market”

An Interview with Dipl.-Ing. Karsten Gräf, Product Manager for Building Automation at Insta GmbH. The company, jointly owned by German building technology specialists Gira and JUNG, not only develops products for its shareholders but also designs and manufactures OEM solutions for other companies – including products based on Matter and KNX IoT (link).

Dies ist die Übersetzung eines deutschen Interviews. 
Zum Original bitte hier entlang
.

Mr. Gräf, alongside Matter, KNX IoT is currently a hot topic in the building automation industry as the newest extension of the KNX standard. What exactly is KNX IoT, and where did it come from?

Karsten Gräf: The idea goes back to an initiative launched by the KNX Executive Board in 2014. This highest decision-making body of the KNX Association, where Insta has been represented since the association’s founding, recognized that the standard needed to move closer to the IP world and embrace IT and web technologies. At the same time, it was important to preserve KNX’s strengths – interoperability, reliability, and vendor independence. The first specifications emerged in 2018 and 2019, followed by the initial version of KNX IoT in 2023. The first products appeared in 2024, primarily from Siemens. To date, however, the market remains tiny. Aside from Siemens, there are only a few other small manufacturers.

2014 was twelve years ago. Matter products have been available for quite some time. Why is it taking so long with KNX IoT?

Gräf: KNX IoT was not developed in isolation. At the same time, the existing KNX standard continued to evolve and expand. The KNX organization has limited human resources – perhaps only a handful of people are working on specifications. In addition, the IT landscape did not stand still during that period. It continued to evolve as well. Thread, for example, was not even on the radar in 2014, and Wi-Fi did not seem like the ideal solution for wireless devices because of its power consumption. So the industry was searching for an alternative, and then Thread emerged in 2019.

“The KNX organization has only limited human resources.”

KNX already had an IP-based version of the standard with KNX IP. Why wasn’t that used for IoT?

Gräf: That’s true. KNX IP – or KNXnet/IP, which is its official name – was introduced in 2004. It helped overcome the limited bandwidth of KNX Twisted Pair (TP). The green two-wire cable can transmit data at a maximum speed of only 9,600 baud. IP connections served as the backbone for large KNX installations and, for the first time, enabled programming and diagnostics over LAN and Wi-Fi.

However, KNXnet/IP was never intended to be connected directly to the internet. It was designed to operate behind additional security measures such as VPN connections. As a result, it was launched without encryption. As connectivity increased and cybersecurity requirements became more stringent, this vulnerability had to be addressed. This led to the initial development of KNX IP Secure and Data Secure. KNX IoT, by contrast, was designed from the outset with native encryption. It is based on IPv6, whereas KNXnet/IP still uses IPv4, and it supports web technologies such as REST (Representational State Transfer) and CoAP (Constrained Application Protocol). That makes it much closer to today’s IT and IoT architectures.

Is IPv6 also the reason why Thread was chosen as the wireless protocol for KNX IoT? After all, KNX already had a wireless solution: KNX RF.

Gräf: KNX RF has always faced challenges in the market. Until 2014, wireless communication was not even supported by ETS, the KNX engineering software. Like Twisted Pair, it also lacked encryption for a long time, and because it operates around 868 MHz, it could not be deployed globally. Different frequencies had to be used in markets such as the United States and Australia. KNX RF Multi introduced improvements such as longer battery life and bidirectional communication. Even so, KNX RF products still account for less than ten percent of the overall KNX market.

With Thread, KNX IoT operates at 2.4 GHz, enabling global deployment. It also offers better suitability for battery-powered devices and more modern communication capabilities. Thread is viewed as a state-of-the-art wireless solution for retrofitting in buildings.

“Thread is viewed as a state-of-the-art wireless solution for retrofitting in buildings.”

With its new hardware platform, Insta plans to offer both Matter and KNX IoT products. Image: Manufacturer

But KNX IoT doesn’t have to be wireless, does it?

Gräf: No. KNX IoT can use any IP-capable medium – Thread, Wi-Fi, or Ethernet. Other technologies are conceivable as well. One example is Single Pair Ethernet (SPE), which requires only one wire pair instead of the eight conductors commonly found in today’s LAN cables. The physical transport medium may change, but the system itself remains the same.

Will KNX IoT eventually replace traditional KNX?

Gräf: I would say it will complement it rather than replace it. Manufacturers will continue to promote KNX Classic over the next ten years, and it certainly won’t disappear. In the longer term, things could change if IP-based technologies such as Single Pair Ethernet become as cost-effective as today’s twisted-pair wiring. For now, however, traditional KNX still has a clear economic advantage. Every cent matters when it comes to node costs, connection technology, and infrastructure. Connecting every KNX device in a control cabinet through Ethernet switches would simply be too expensive with today’s technology. The situation is different in the retrofit market, where wireless solutions play a larger role.

That puts KNX IoT in competition with Matter. Both standards rely on Thread and IPv6.

Gräf: You can look at it that way. From my perspective, however, KNX IoT and Matter address different markets. KNX IoT targets the professional building automation sector, while Matter is focused more on consumer and smart home applications. In the professional market, stability, longevity, and planning reliability are more important than rapid market penetration.

Still, there is a growing number of professional-grade Matter products from companies such as Eltako, Hager, Warema, and Wago. When can we expect a similar range of KNX IoT devices?

Gräf: Many KNX manufacturers are currently focused on KNX Secure because the EU Cyber Resilience Act makes that necessary. As a result, KNX IoT will lag behind somewhat. I believe we will see more products beginning in 2028, but broader portfolios are unlikely before 2030. At Light + Building, Insta presented a modular hardware platform that supports both Matter and KNX IoT. That gives OEM customers investment security.

The professional market operates on much longer timelines than the consumer market. Systems such as KNX and DALI do not disappear overnight; they evolve and improve over time. In areas such as energy management and smart lighting, KNX Classic and KNX IoT are developing in parallel. Devices like heat pumps and EV chargers, which often already have IP connectivity, are particularly good candidates for KNX IoT integration.

“We’ll see more KNX IoT products starting in 2028.”

So far, manufacturers of heat pumps and solar systems have shown little interest in making their products Matter-compatible. Why should KNX IoT be any different? After all, gateways already exist.

Gräf: Gateway solutions are always possible, but they are complex, expensive, and cumbersome to commission. That is not an attractive approach for electrical contractors or HVAC installers. The goal has to be making integration faster and more affordable. Finally, it comes down to economics.

What does an installer need to expand a traditional KNX installation with KNX IoT?

Gräf: To connect KNX IoT with KNX Classic, you need a KNX IoT router that translates between the two standards – essentially between IPv4 and IPv6. This is primarily a software function. It does not necessarily require dedicated hardware and could be integrated into existing KNX IP routers or server platforms. During the transition period, there may be standalone devices, but in the long run, KNX IoT should become a software component within existing products to keep infrastructure costs low. For Thread-based installations, you also need a Border Router. That is standard hardware that already exists today and, in most cases, wouldn’t be located in a control cabinet anyway.

If KNX IoT devices won’t reach significant volumes until 2028 or 2030, there probably won’t be many KNX IoT routers before then either, right?

Gräf: That is probably true. At the moment, I expect newcomers to be the first to offer such products because they want to gain a foothold in the market. Established KNX manufacturers are heavily focused on KNX Secure and currently need to adapt their portfolios. As Insta, we have made an offering and received positive feedback, but many customers are quite open about saying: We see the potential, but we are waiting to see how things develop. Launching too early can also be problematic, as the discussions around Thread within the Matter ecosystem have shown.

You support both technologies. Which requires more development and certification effort: Matter or KNX IoT?

Gräf: The KNX certification process itself is unchanged. It works the same way as before and is comparatively affordable. Initial certification costs typically range from around 4,000 to 7,000 euros. After that, OEM variants generally incur only administrative fees of less than 100 euros. With Matter, it can cost €20,000 or €30,000 to bring a new device or an OEM variant to market. One reason KNX is more affordable is that the KNX Association is heavily funded by sales of the ETS software. That said, it is clear that Matter certification processes continue to evolve. One example is the move toward group certification for product families.

“One reason KNX is more affordable is that the KNX Association is heavily funded by sales of the ETS software.”

Matter covers a wide range of applications. Which product categories are likely to appear first with KNX IoT?

Gräf: Most likely sensors, especially for retrofit applications: push-button sensors, climate sensors, presence detectors, and motion sensors, typically based on Thread. On the actuator side, we will probably see decentralized solutions such as in-wall modules first. KNX has always followed a decentralized architecture, and that remains the right approach for retrofit projects. A KNX IoT sensor communicates directly with a KNX IoT actuator without requiring a hub or internet connection.

Mixed installations combining KNX Classic and KNX IoT are entirely possible. Actuators can remain in the KNX Classic domain while sensors are added to rooms via KNX IoT. This hybrid approach will likely become the standard solution in many installations initially.

And after that, will the ecosystem broaden the way Matter’s has?

Gräf: Over the past 25 years, KNX has standardized a wide range of applications – such as lighting, shading, and heating, ventilation, and air conditioning. These functionalities remain the same, regardless of whether they are implemented in KNX Classic or KNX IoT. New areas such as energy management are developing independently of specific platforms. Interoperability is ensured as usual through certification. That is the difference from Matter. KNX provides a clear promise of functionality. What is stated in the product description works with every installation – regardless of which ecosystem I use. With Matter, the platform decides what I can do with it.

“That is the difference from Matter. KNX provides a clear promise of functionality.”

At the same time, there will clearly be differences. Many traditional consumer-oriented products are likely to remain on the Matter side of the market – the often-cited intelligent refrigerator being one example. As a result, gateways and bridges between Matter and KNX or KNX IoT will continue to exist. That is not a problem because the two technologies serve different segments.

There is a sense that competition from Matter is influencing KNX development as well. Areas such as commissioning seem to be getting simpler. Is that impression accurate?

Gräf: Smart Linking – the concept of linking channels and functions rather than individual group objects – was actually planned within KNX as early as 2016, well before Matter existed. The topic was temporarily deprioritized in favor of KNX IoT, but it is now gaining momentum again thanks to the addition of semantic data. Starting next year, both KNX Classic and KNX IoT are expected to support Smart Linking. ETS will include a Smart Linking Explorer to accelerate engineering and commissioning. Insta also demonstrated a simplified parameterization approach for standard applications (Xpress configuration), at the trade show.

KNX Classic and KNX IoT will continue to evolve in parallel, benefiting from synergies with each other and with Matter. That’s why I’m optimistic about this development. There are also advantages to not being the first on the market – because then you can learn from the initial problems others encounter.

Mr. Gräf, thank you very much for this interview.

Share this information:

Sponsor / Advertising

Sponsor / Advertising

Scroll to Top