So this is more of a theoretical topic right now, but I plan to build a new house within the next couple of years and would like to make it completely Home Assistant enabled from the very beginning.
In my current home I have retrofitted everything with HA already and am asolutely in love with it - all my lights are smart (Zigbee & ESPhome), all the wallswitches were replaced with wireless Aqara wallswitches, there’s temperature sensors in every room which feed HA with the necessary information to control the heaters (which are horribly inefficient electrical heaters right now, which makes them easy to control, but expensive to run), same goes for the ACs, wich are ESPhome-enabled, there are CO2 sensors which (right now only) tell me when it’s good to open a window in the relatively small office for example, even my bedside lamp runs on ESPhome and controlls the alarm system (which itself is in HA), I get notifications when the washing machine, dryer, or dishwasher are finished, and so on
So, as long as its not too deep of a financial cut, I HA-enable everything … and since last year or so, it started to just work (before, I faced quite a few bugs regularly)
So now the only concernes with building a house around Home Assisnant, that are left are:
Will Home Assistant, Zigbee, Esphome, Node Red, etc stay around for als long as the house will? So 50+ years at least?! or at least, will there likely be devices that can seamlessly replace them (yes, I already asked my own crystall ball - the answer was non-conclusive)
Basically I want a system that’s as long-living as E27 Sockets - I don’t care if I will screw in LED, halogen or traditional bulbs - as long as I can just screw them in… the same should aply to my smarthome
Will Home Assistant really stay an open source community project forever? There’s nothing more terrifying to me than building a house and suddenly Apple, Microsoft, Google, Amazon, or - the horror - Meta knowing when I go to the toilet! These companies are banned from my smarthome and as soon as there is even the slightest sign of HA being compromised by them, I’m out!
now more technical questions:
Right now I have Aqara WXKG02LM wallswitches - they work great, but I feel like the delay between pressing them and the light turning on is, with around 1s, a bit long (it’s definitely them, because if I turn on the light in HA, it works instantly) - do the newer versions react faster? Are there maybe alternatives from other brands (buying Chinese products in the current situation feels wrong anyway)
Sould I even go down that route I am currently pretty much already on (and happy with) and completely forgo traditional light switches?! I would not add wiring to the locations of the lightswitches, but wire the lights directly (of course there would be something like a small breaker-box in every room for maintance). Also I would add a couple of regular lightbulbs and switches throughout the house, so in case of HA going down (which fortunately hasn’t happened since I have upgraded from a Pi 3B to the Blue) there are still emergency lights everywhere (maybe I could even find a way to automate them )
wired vs. wireless … generally I would always prefer wired solutions - but KNX for example is really expensive and complicated to integrate (hard- and software)… but at least for door/window contacts I plan to mostly go wired - how sould I do that?! the simplest solution would probably be to put a pair of wires from every window/door to the serverroom - but it’s going to be quite a long distance in some cases, which means I would have to up the voltage they operate on … is there any good documentation on how to realise that?! Should I use a LAN-enabled ESP32 to collect that data and forward it to HA? Whats the least error prone solution?
This is not about being extremely high security (I live in a rural region with little to no crime), but about consistency (even though I mentioned that my wireless lightswitches work great, every now and then, when I press them, nothing happens - not sure if it’s a Z2M issue or a general issue with Zigbee, but routers tend to go offline for exactly one minute every now and then, even though they have perfect linkquality…)
wired vs. wireless 2 … another aspect to consider is, that I would like to build in the style of a modern Canadian loghome, so EVERY wire I need needs to be considered beforehand and adding wires afterwards it pretty much impossible - also adding more conduit makes the price go up quite significantly afaik … BUT thick wood is not so beneficial for wireless signals either - that’s no problem for WIFI because I just plan APs accordingly. However for Zigbee this could become a real problem - does anyone have experience with that?
Are there wired Zigbee Routers, so I can extend the Network to a location where there’s no wireless connection?!
not a question, but a consideration: everything has to be as open source as posible and local. I do not accept any cloud based devices in my smarthome (right now I only have a Harmony Hub - which is prohibited from accessing the internet because I have not found an acceptable solution for to replace, as of yet)
Obviously there’s going to be CAT7+ cables to at least two locations in each room
so these are the things I can think of, that need consideration before even actually starting to plan a house
Of course there will be details that also need to be adressed some time (like solar panels, batteries, heat-pumps, etc), but right now I only think about the very general aspects of building a house around Home Asisstant … did I forget anything?!
That’s also what people say, who still run Windows 7 … it is a huge security risk not to update - and yes I said no cloud, but I use NabuCasa (which is not “cloud” in any way, but means HA needs an internet connection … same with the Telegram bot)
that’s true - but we all know, even though the companies try to make us believe LED lights will last hundreds of thousands of hours, this is not true… I had to replace a couple of HUE lamps already within 3 years, because they just broke … so if they are no longer available for purchase, I am forced to update the system … but yeah relastically, upgrades to the hardware are probably going to happen anyway
that’s the problem … on the one hand I say NO WAY - always go wired … on the other, I pretty much exclusively use wireless wallswitches since over 2 years and they just work like I’d expect them to
HA should never go down unexpectedly … I’d have bigger problems than no lights, if it does - back when I used the Raspberry, it got stuck regularly, which (besides other probelms) resulted in me coming home to a freezing cold or blazing hot house!!! It was the main reason I upgraded the hardware … it just has to work, no ifs or buts
Sadly though nobody can tell the future with 100% confidence, which is what you’re asking.
However, if the worst case were to happen, you’d have time to switch to something else, such as an open source fork. If you’ve been sensible and based your hardware around widely supported standards then the impact would be minimal.
Yes, but no. If you’re not using their hub then you have a lot more freedom.
That is a good question, but imho 10% is a huge threshold.
If a new house costs about 500k€ (Which is probably the lower end nowadays) that would result in 50k€ - which quite much, should suffice for a full featured, professionally built, Knx installation.
1 second delay is your expected behaviour I’d tear these things off the wall on day one.
Where will the house be built? There are also some regional concerns - like I’d recommend Knx for Europe every day, but probably wouldn’t for Americas (because I don’t know if there are as many devices for their different standards).
I don’t think 10% is that much, but it depends on what you’re aiming for.
If you need to put extra conduit in walls, extra cabling, different switches, different heating system, different garage port, etc. it will cost extra.
Well the budget is kind of a tough question … the house will probably be worth over 500k€, but I plan to do the vast majority of work myself (except for the basement probably) with the help of neighbours and friends, so the actual costs will have to stay well under 300k€ (which is the also the kind of credit I can comfortably pay off) - not considering the favours I owe then xD … however a close friend is a carpenter who “has to pay off debt” and looks forward to helping me - which is a huge plus!
For home automation I don’t want to spend more than 10000€, which is realistic, considering right now I’m actually under 1000€ - I only buy stuff, if it’s really good value - for a new (and my last house in life), I’m definitely going down a more professional route, but will still watch the costs…
Well Zigbee and ESPhome are the standards I use - so it’s a two-edged sword … WIFI will likely stay around, but who knows when todays standards get discontinued, and the old devices won’t be able to connect… Zigbee will likely be easier to replace or rather will just continue to work, because it’s a closed system… WIFI APs will definitely get replaced two or three times over the lifespan of a house - zigbee could just stay as long as it works.
Well for Zigbee for example I use Zigbee 2 Mqtt - but who says that zigbee itself will be a thing in 20 years?! If a lightbulb breaks, ok - replace it with something else - but if my coordinator breaks in the future, I’m out of luck and will have to replace two thirds of my houses electronics… (obviously I would have a spare one to hand, like I do right now, but what after that?!)
Central Europe - but KNX is too much of a hassle and expensive … I’ve already looked into it and doubt it’s possible for me to implement - it’s obvioulsy not designed for endusers to install and configure - this is also a plus, because there will probably still be replacement parts available in a couple of decades.
I put HA in as a retrofit to bring together existing lighting, heating, security etc systems and have thought a bit about what I would do again if I had another big house renovation. I currently have HA as the integration and automation hub plus supplementary Zigbee sensors / controls. These could fail and the house would still work.
My heart would say go full HA and install lots of Zigbee devices but my head would be worried about ongoing support and complete reliance on HA for basic functionality. Plus - the hidden cost of tinkering time.
Another option is to consider if some aspects can be delegated to specialist sub-systems which can be integrated into HA but are supported by stable vendors and would continue to work stand-alone. For example, if I were starting from scratch again, then I would (again) implement a professional alarm system such as Texecom and hook PIRs, door/window, fire alarm sensors etc. up to this (wired). This sensor input is then available to HA via MQTT. Security panels and accessories are cheap and reliable, good at what they do and have long, supported, lifetimes.
It’s a harder choice with lighting. First time around I went with Rako which became a sub-system of the overall HA setup. But - this was before relatively cheap relays, dimmers and LED controls from the likes of Shelly. Today, I could easily replace the Rako functionality with HA and Zigbee devices at lower cost than Rako. I would still look for an option which kept basic lighting functionality if HA failed.
Similar discussion with heating - I have a thermal store, multiple heat inputs, thermostats in each room, manifold-fed UFH and radiators. I could migrate the control of this to HA or, I could buy a couple of specialist solar / heating control boxes and link this into HA. In this instance I think I could design/build a simpler system now in HA but I would still scan the market in case there are off-the-shelf systems which are HA friendly but already work out-of-the-box.
As other have suggested, a lot of this is down to personal choice and how happy you are keeping the HA ecosystem up-and-running. If the balance in my current home was more skewed towards HA then I would also be more careful about keeping separate development and production environments to protect the working system from the inevitable failures which are inherent in a platform which is evolving so fast.