The reason I refuse to hand over my home automation to a cloud service. And my automation is relatively simple.
I have no experience of these. I guess they were mostly on the US market(?) but based on a quick glimpse of the article, at least the DYI users aren't left with expensive paper weights:
With the Insteon Hub app no longer working, Reddit users recommend alternative apps such as Home Assistant and OpenHAB that work with Insteon gear, at their own risk.
While I prefer local only stuff, I don't have too much issues with things that can be made to work local only. Especially things that don't require WiFi and therefore additional work to prevent them calling home. 
This isn't good news and the way the founders have wiped and run is quite shocking.
It's a risk we all run I use LoRaWAN which is open standard with a good sized eco system but on proprietary RF from a single chip manufacture Semtech who have a wide portfolio of products.
It's widely used both by Home Automation folk and large publicly operated networks for commercial use.
But still I'm hoping I don't get an Insteon moment, there are not the first HA manf. to pull the plug. I've also experience this with two well funded home medial devices. The consumer space is hard and it's going to be the global brands that win out even Tesla say they are considering entering the HVAC market in the USA.
Didn't the Things Network also change the protocol in an incompatible way? But of course this shouldn't matter that much as I assume one would anyway use just LoRa without the WAN if wanting to focus on local only...
Although I haven't got any Insteon devices myself, the reason why they pulled the plug seems like a lot to worry about and not a lot to blame for the business:
In 2019, the onset of the global pandemic brought unforeseen disruption to the market, but the company continued to move forward. However, the subsequent (and enduring) disruption to the supply chain caused by the pandemic proved incredibly difficult and the company engaged in a sales process in November, 2021. The goal was to find a parent for the company and continue to invest in new products and the technology. The process resulted in several interested parties and a sale was expected to be realized in the March timeframe. Unfortunately, that sale did not materialize. Consequently, the company was assigned to a financial services firm in March to optimize the assets of the company
Source: https://www.insteon.com/
I would say the best bet (for wireless devices) at the moment would be Zigbee together with WiFi as both have a long history and aren't tied to a single company. Matter seems promising in this sense (although sadly also a proprietary protcol).
Hi ..Things Network also change the protocol in an incompatible way
No just upgraded their stack and all users have been migrated.
Also they have a large enthusiastic user base as their Community edition is free.
And there is a crypto earning network called Helium that uses LoRaWAN.
But there are many large commercial LoRaWAN deployments for meter reading and street lighting 10,000s of devices and LoRa/WAN is used in Amazon sidewalk.
I've been deploying wireless automation devices in buildings for over 15 years.
Any 2.4GHz based wireless will have range issues.
However LoRa now operates in 2.4GHz what we call LoRaLAN and with a new cheap chip to support it. We expect to see a lot more consumer based products and could support running Matter.
So I feel safe for now but ....
What you described sounds more on the business/organization level rather than individual people. I have also heard on several occasion LoRaWAN is used for reading meter data from residential buildings in the Helsinki area (where I live). Now that you reminded me of this, I would believe there to be very little to worry for a tech so widely used by commercial businesses to just disappear without warning.
I wonder if Nextion was used by commercial entities (in a big way)?
What I was trying to say that given it's wide use commercially I don't see going away.
But the The Things Stack has over 160K of average joes with 20K of community gateways in 151 countries. And Node-RED is quite a popular way for them to interact with the devices at home.
Ok sorry! I misunderstood your comment on it being a risk LoRa chips are proprietary to Semtech as the only manufacturer.
I can for sure understand the benefit of TTS but to me it's somewhat similar than relying on IoT providers services. Of course I don't see it nearly as big of a risk to disappearing out of the blue but I assume the routing backend also runs on servers provided by a single entity(?). I haven't dug into it myself so I'm happy to be corrected!
I want to also add that I would definitely be using it already if I just had a suitable use case, although the coverage in Finland is pretty abysmal as shown below. I've been contemplempating on building a gateway somewhere around the area I've marked on the map just to fill my curiosity. It's such a remote place I'd be very surprised if it ever got any random users but (unsurprisngly) the Helsinki center looks to be pretty well covered so our cottage would seem more appropriate.
You don't have to use TTN you can run a open source LoRaWAN stack like ChirpStack and a concentrator card on a RPI and use Node-RED as many people do. Some manufactures include it all of this in one Gateway we call it a private network. Many of my customers use this approach for on-perm solutions. BTW a TTN LoRaWAN gateway is only 80 EUR would cover your home up to 2k range in urban environment.
My home use case it lighting/heating/plug-top loads
Thanks for the ChirpStack tip! Somehow I've missed this even though I've been following Andreas Spiess since his YouTube channel started. I have bought a pair of ESP32 LoRa development boards some years back and indeed the range is mind boggling. I've so far only done some experiments at our cottage premises and I even with these simple boards I could easily get hundreds of meters of range (with dense forest and buildings in between).
Recently I bought at bunch of different dev boards with the intention to play around with Meshtastic during the summer but perhaps later use them for projects. My plan for sensor nodes was to use one of the boards as a receiver and use something like RadioHead or RadioLib for communication. However I've now got a mixture of of 433MHz and 886MHz boards so the plan does not make as much sense anymore. A two band router would indeed solve this issue. 
This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.
