Basics of wireless smart home solutions

If you are sure that you actually want a smart home, I would always have an electrical installation suitable for it done. Run everything to the switch box and lay KNX cables, use electrical cables as 5-core. Then you can operate everything completely normally with switches (push buttons) and still, if you want, make everything smart. However, you will then also need a large box and/or sub-distributions. Radio frequency is always problematic because of the range, some smart radio parts fail to reach the room below from the first floor in my case. Concrete ceilings and tightly laid underfloor heating interfere significantly.

As already emphasized several times, KNX is not off the table, I am interested in basic information.

Your input factors include, among others: * Thick wallets * Willingness of your construction partner If you can and want to pay the additional cost and your construction partner can (or allows) to implement this, then I would already implement it as suggests. Then you just have to think a lot "from the end" and consider all the possibilities in the planning. But forgetting something and then retrofitting it via radio would again create a "hybrid," possibly even with its own control unit. I have since become familiar with ioBroker. Currently, I only use it to turn a night light on/off and retrieve sensor data. In a visualization, I display the room temperatures and relative humidities. My Enocean actuators are still waiting to be integrated because the enocean adapter from ioBroker is not running smoothly yet. But Zigbee is working. I am still struggling with integrating Alexa... I still need to come up with a solution to retrofit presence sensors.

That was a fundamental piece of information :).

You don’t have to use KNX even if the cable is there, but by centralizing the wiring in the distribution box, you have everything within reach.

The goal is to have a modern electrical installation; if you plan the old conventional "we've always done it this way" installation, everything that is smart home will always seem tacked on and you will always encounter problems during implementation.

For example, if you let the wiring of the light outlets lead to the distribution box and the wiring of the switches as well, then you can:
- Use a normal switch
- Use a push button
- Use a sensor
- Use an automation system

Without having to tear up anything in the walls, re-lay wiring, or have to build anything into the switch boxes.

You can then do exactly what you plan, start with some things first, and see what comes next.

Whether you then use radio, KNX, other actuators, or none of that at all is irrelevant.

If you use standard wiring from the switch to the lamp, you have to
- Use light bulbs with radio
- Put actuators in the light outlet and/or the switch
- Be fixed that a certain switch only controls a certain outlet
- Make the outlets and boxes larger to accommodate the actuators
- Cannot arbitrarily decide what the switch should be for

but there is also nothing against having the light switches next to the door simply turn the room light on/off. And that reliably for the next 50 years. As a minimum, deep [UP Dosen] should definitely be installed so that a [Funkaktor] can be installed if necessary.

If switching is done via conventional wiring, everything speaks against it, as already mentioned above.

You can do it that way, but then it just sucks.(c)

More or less everyone who has worked in construction has probably made a statement like this at some point. Be it as a plumber, electrician, or otherwise.


About the radio systems:

[*]eNet - boasts the following claim about the facts of the system on its own website: "Over 30 components". Um, like 30 components? I have more on one floor??? And a typical single-family house nowadays often has more than 30 lighting groups. So here we would already be hitting limits.
[*]You see Somfy and everything from one source as an advantage? You probably did not experience the debacle with various pseudo-smart homes from large electricity providers, etc. It should be said that anything proprietary is and remains a dead end.
[*]EnOcean yes good especially for retrofits. In new buildings only if nothing else is possible. Because here too the selection of components is limited, the haptics require some getting used to, and there are further restrictions.
[*]Zigbee interesting but only quasi open, only partially compatible, and little interference with the average neighbor's WLAN. So yes, it could work but rather only for light controls. Entrusting it with more important tasks is rather suboptimal.
[*]Only Z-Wave remains. As the only truly open system with a similar range of manufacturer support as KNX and a comparable selection of components. However, as always with radio, it is uncertain how it will ultimately function and concrete ceilings often prove to be a knockout criterion for such solutions.

The other small and large systems proprietary, open, or semi-open are just unfiltered exhaust fumes. Especially everything that uses WLAN and/or BT as a communication channel. That is sufficient to switch garden lighting remotely but for more interesting tasks rather not.

but there is nothing wrong if the light switches next to the door simply turn the room light on/off. And do so reliably for the next 50 years. As a minimum, however, deep flush-mounted boxes should definitely be installed so that a radio actuator can be installed if necessary.

It is always the same thing to read on this subject. A thousand ways are always suggested to somehow botch something instead of doing it right once. Radio actuators are and remain crutches. A light switch at the door is a relic.