one of the biggest complaints about the new Connect ZWA-2 controller is that it does not have power over ethernet (PoE). Considering that PoE has been pretty popular in the smart home space for quite some time, I would like to hear an official response as to why the Nabu Casa/Home Assistant team decided to go with USB. Is there any technical or financial reason behind it or was it simply a bad decision?
Aug 13, 2025ConnectZWA-2 was built on our second-generation platform, which prioritizes high performance and openness in its design. It is a big leap from our first-generation Connect ZBT-1
Bad decision? Hardly. It is an upgrade from an existing product, dramatically reducing the development time. Moving it to an Ethernet platform would likely mean rebuilding from the ground up. A blank page design. Maybe in the future?
Also, I don’t think POE is as popular as you think. Maybe you think it is because you prioritize POE in your searches?
I probably shouldn’t have outright called it a bad decision as there might be certain advantages to USB I’m not aware of (that’s why an official response would clear things up) but questions about a PoE version/complaints about the lack of PoE were the top replies under every ZWA-2 announcement post I’ve seen so far so there clearly is a demand for it.
I didn’t say there is no demand. I was suggesting probably why they upgraded an existing design instead of starting anew with an Ethernet device. Who knows? Maybe an Ethernet POE device is in the works now?
One reason is because the TCP-serial protocol everyone is using is just not robust enough. With the ESP the adapter is capable of using a more reliable network protocol, one that might even work over WiFi.
My approach to smart home technology is shaped by a preference for robust, familiar standards. I’ve standardized my home on Ethernet and WiFi because they are open protocols with a massive ecosystem of support and diagnostic tools. With a house fully wired for Power over Ethernet (PoE), expanding my network or fixing a dead spot is as simple as adding a new access point.
Why I’m Hesitant About Z-Wave/Zigbee
My reluctance to adopt protocols like Z-Wave or Zigbee comes from negative experiences with technologies like X10 and Insteon. The primary concern is long-term viability; I’m wary of investing in a closed ecosystem that could be deprecated, forcing a costly and time-consuming migration. For me, the long-term stability of Ethernet and WiFi is a major advantage.
I’ve read reports from users needing to migrate from older Z-Wave versions to newer ones, and this sounds exactly like the problems I had to solve with Insteon. In the world of Ethernet and WiFi, backward compatibility is rarely a concern.
What It Would Take for Me to Adopt Them
I recognize that Z-Wave and Zigbee offer access to devices that may not have WiFi or Ethernet equivalents. I am open to trying them, but the bar is high. The experience would need to be as simple and reliable as my current PoE-based network.
Ultimately, this is intended as constructive feedback, not a complaint. If a product emerges that integrates seamlessly into a robust network infrastructure and promises long-term stability, I would be eager to try it.
Yes, at the chip level.
Z-Wave certification requires that you use approved chips and software from Silicon Labs or Trident. But everything you buy that is Z-Wave compatible have one of those chips inside and the Z-Wave JS integration works pretty well.
If the NabuCasa team wants to make a new Z-Wave device, they would have to be a member of the Z-Wave alliance ($15,000 annual dues) and submit the product for testing. I really doubt there is enough market for a POE Z-wave adapter to cover just the Alliance membership. The smarter move would be to approach existing makers of a similar device to join the Works With Home Assistant program.
Homeseer released their first Ethernet based Z-wave coordinator in 2015 and have been selling versions of it since then. It isn’t POE but can be easily converted to POE.
I also agree that it’s likely less than 10% maybe even less than 1% of Home Assistant users have POE in their homes. It makes no sense for them to target that low of an audience.
Compare the number of sales to the number of people who posted about the lack of POE support and see if you can continue to make an argument for an official response.
I personally would like to see a POE-based ZWave coordinator from Nabu Casa. BUT I absolutely understand the product decision to do USB (first?). Possibly as an additional product?
However, if ZWA-2 is as good as the reports indicate (esp. using 800 LR), maybe positioning is no longer as big a deal?
I don’t know how many people use PoE however your guess is as good as mine unless you have data in your hand. I would argue that PoE is more future-proof though.
There are people who would prefer PoE but don’t complain about it so comparing it with sales numbers is meaningless. If someone (ideally someone from Nabu Casa) presents actual data showing PoE did not make sense I will be fine with that, I would simply like to know the reason behind this decision considering more and more newly released sensors support PoE.
I NEVER said POE doesn’t make sense. Absent data showing that POE would sell more units, then what manufacturer would add the engineering and product cost?