Along this same line of thought, I've always wanted a smart switch with a motor or magnet inside that physically moves the switch when it gets toggled by software.
The UX of a physical switch whose position always accurately reflects the state of the circuit is hard to beat, and all the smart switches I've seen seem to use buttons or some other non-tactile solution that is significantly worse UX than a plain old switch.
Yep, a while back I got to use a Behringer X32 where you could control the volume faders remotely with an app on your phone and thought "Why don't all smart controls work that way?" Smart features, but you retain the advantages of physical, tactile controls.
My smart switches have a little light on them that shows whether they're on or off. The actual light they control is also a pretty good indicator of the switch's state.
Switch position isn't just a visual indicator; it's also a tactile indicator, allowing you to to feel the position of the switch when you reach to turn it on, and feel when you've successfully flipped its state. Lights aren't the only things that get hooked up to switches, so an indication of state can sometimes be helpful.
Physical switches also have other UX benefits, such as having zero learning curve, and being easy to operate without looking at them. I've built up enough muscle memory at this point that I don't even have to think about toggling the light switch when I walk into a room in my own home; it's so automatic at this point that if I'm distracted I'll sometimes even forget there's another person in the room and flip off the lights on my way out the door.
These are all minor things obviously; but I still wish I didn't have to sacrifice them to get the benefits of a smart switch. I understand the desire to avoid moving parts; but surely a couple properly engineered solenoids wouldn't be that prone to failure.
AFAIK remote controlled circuit breakers are sometimes used on large ships, and presumably in other applications with heavy loads.
When you have a 200 Amp breaker, just running a cable from a load location to a convenient control location (and back) calls for: a) extra lengths of heavy wiring; and b) a lot of crowding and layout constraints in your control panel area.
Haven't read the article yet, but perhaps something could be said for e.g. British wiring practice, where last I heard lighting was commonly all run on one circuit. I imaging that could make for an easier wired network for control. By varying the bridge to such a wired lighting-control network, new wireless communications/controls could be accomodated.
> The UX of a physical switch whose position always accurately reflects the state of the circuit is hard to beat
True, but modern houses commonly have a ton of 3-way switches with multiple switches for each light, so they rarely reflect state anyway. At least my Lutron switches and dimmers can show me status with the tiny LEDs...
There are some switches I've seen that can snap back to "neutral" when another switch in the circuit is turned off, so anywhere you go the normal "push up to turn on and down to turn off" works, but if I recall they were weirdly expensive.
That sounds nice, but I've got enough circuits with two or more switches on them, you quickly learn to just toggle the switch to toggle the circuit. Maybe if I had fancy switches, you could just press up to turn on and down to turn off and that would help the couple times where two people toggle their nearest switch around the same time.
A switch exists for this, however they’re quite expensive and had some financial issues in the past (although they seem to be under new management now)? https://dendevices.com/products/switch/
Random thought, but it makes me think that it could potentially be possible to hack into someone's light switches and burn their house down by setting the motor to leave the switch in the exact middle of on and off position so it arcs electricity.
Properly designed switches have a hysteretic snapping action on the contactor, so that there is no position of the switch that would result in the contactor not being fully open or fully closed.
The “correct” design of this concept would not have any actual current (or maybe a very small DC current) going through the switch. Like the Shelly devices in this blog post there would be a relay that actually switches the current. The switch hardware would just be for show.
This shouldn’t happen in theory in a properly engineered and built system. The switches have arc chutes inside of them already, and the contacts are spring loaded to open and close as fast as possible to minimize. On the motor end, the maximum amount of amperage a motor pulls is called the Locked Rotor Current. You can achieve LRC by stopping a running motor with a screwdriver. But a motor also hits LRC for a brief second every time it starts from a rest state. So a UL rated motor sees maximum amperage all the time in its normal lifecycle.
Alternating current helps too. That’s why many switches have no DC rating or much lower DC rating compared to AC. A common mistake of hobbyists is to undersize a switch for a DC circuit based on its AC rating.
I don't think it would start a fire since all the materials surrounding the contact area are supposed to be fire resistant, and the arch is pretty small.
I asked an electrician about this once and they told me the arcing generates heat that travels up the conductor and into the wall, damaging the insulation and sheathing over time, and eventually exposing the bare, hot (as in warm) wire in the wall. This might introduce the wire to conditions where additional arcing (and additional heat) is likely -- for one thing you probably now have multiple bare wires in close proximity -- which is a common source of electrical fires.
Ditto, been looking for exactly the same thing. I don't want to lose control of my lights just because my WiFi goes down, I want a smart light switch that becomes a dumb light switch when the smart part fails
It is a bug of light switches that they share the state of the device they command. A forgivable one, given how light switches have worked, historically.
This is fine for one switch, but it doesn’t scale to N switches. If you have more than one switch, you now have to synchronize the light state to all switches.
A better switch would have a spring loaded neutral position which you deflect towards the state you want. That way each switch is sending commands to the light, not trying to match its state.
Multi switch lights would benefit from this. That includes physical switches, as well as software ones.
I'll echo a comment I wrote downstream: If you're making your home smarter, don't buy anything other than Zigbee (or Z-wave, but I haven't used that). It's an open protocol, which means any device that supports it (and they're a lot) will work with every other device.
They also work without the internet, they can pair with each other (so they work without a hub), the protocol is simple, mains-powered devices are usually repeaters as well, a coin cell battery lasts forever, etc etc. I'm never buying a wifi-based smart home device again.
Beware: IKEA says they support Zigbee but they don't work well.
Its not all peaches in Zigbee land. I'm troubleshooting issues with an Aqara temperature sensor as we speak. Aqara has plenty of "known" issues regarding Zigbee compatibility [1]. The unfortunate bit is that their devices are the best I've seen in terms of price/form-factor.
Repeaters and a need for repeaters is a whole other can of worms for Zigbee. I've had to purchase and deliberately place some repeaters to get better coverage across the house, or else I end up with unreachable devices. And because of how the mesh works, you might have to reset devices to try and coax the mesh to rebuild if you move anything around. It's overall been a much more buggy setup than my wifi subnetwork.
Funnily enough I have ~40 smart devices in my home, around 30 IKEA and 10 Aqara, controlled by Home Asssistant on a raspberry pi with a ZZH USB stick — aside from some initial fiddliness where the aquaria devices would pick a random peer to connect to instead of a peer with good signal, it’s all working pretty great :D
In a similar theme, I can’t recommend Lutron Caseta enough. They work as regular switches/dimmers, have their own simple and rock solid wireless protocol separate from wifi, the hub can connect to the network via Ethernet, can be controlled by all the major home systems including home assistant, and are made in both neutral and non-neutral models.
I believe only the dimmer switches are non-neutral, since it actually piggybacks off dimming to provide neutral capability (i.e. it actually is always on but dimmed _verrrrrrry_ low to get the bare minimum power required to run its chips without being visible at the appliance).
Ahh, interesting thanks for the link. I bet the non-neutral aspect has something to do with both price and availability. There really are not many options for the consumer that avoid the neutral.
> and if that cloud goes down, you could struggle to turn your lights on
For the mainstream smart bulbs and switches I have (Philips Hue bulbs and Belkin Wemo switches) their app-based controls work even if the Internet connection is down.
The only thing that doesn't work is voice-assistant controls, and also the configuration of schedules, etc. And as others have noted, they all work fine as dumb physical bulbs and switches.
Seems like overkill. His assertion about smart switches is incorrect. I have a house full of smart switches that don't connect to the cloud. They are standalone, but will take commands from Home Assistant. If HA goes down, the switch never even knows, it's still just a light switch.
This is how -all- of my home automation works. The thermostat can be controlled via HA, but it's just a thermostat and works 100% normally without HA. The deadbolts? I can see their status in HA, I can tell them to lock, but they still just have a twist knob on the inside and a key (in addition to a keypad) on the outside, so they can be used just like any other deadbolt. Garage door openers, same deal.
I do agree that using something like Hue to turn lights on/off isn't a great idea.
Don't buy anything that isn't Zigbee (or, I guess, Z-wave). Wi-Fi is much too heavy and insecure for smart home stuff, Zigbee has a ton of features you'll want (much simpler protocol, automatic repeaters, etc etc).
All of my smart switches are Jasco (aka GE) Z-wave switches.
As I buy other HA toys, I've been getting Zigbee when I can. It has been less hassle to do the device adoption process and the USB dongle I use for Home Assistant supports both Z-Wave and Zigbee simultaneously.
Updated May 2022: Since writing this blog post I have setup Smart Home Shop UK, an online retailer for Smart Home Devices specialising in the Shelly device range
This submission just seems like an advertorial to me. YMMV
FWIW - if you buy a Hue bulb today (Bluetooth connected) it functions as a normal bulb. Turning it off and on again at the switch or lamp will restart it as if it were a normal bulb. If it's a color change bulb it restarts as a white light.
"This also rules out smart light switches, as these often need to connect to some kind of cloud, and if that cloud goes down, you could struggle to turn your lights on."
Hu? No it doesn't. It's a switch, you press it and the light goes on and off, it needs no cloud connection at all. You can automate it if you like, but you don't have to.
And his choice was instead to use a proprietary "Shelly" device that uses WiFi instead? With its own "Shelly cloud"?
I'm sorry, but that was a bad decision.
Get a standard Z-Wave or Zigbee switch, it functions as a real switch, because it's a switch. Then get a Hubitat, which runs entirely locally, and needs no internet unless you want to connect it to Amazon Assistant or Alexa (and you don't have to).
This is future proof, and outage proof - the light switch function as dumb switches, the Hubitat runs entirely locally, with zero internet (not even local internet). And Z-wave or Zigbee are standards and you can interoperate with tons of other devices, so no worry about some manufacturer going out of business.
Shelly devices can easily be set up to not use the manufacturer’s cloud, which the author of the article apparently didn’t disable, since he uses their mobile app.
When that cloud service is disabled, some more developer-friendly protocols are able to be used, like MQTT, which makes these Shelly devices as future-proof as the other standards you mention, and MQTT libraries and clients are easy to find.
The only bad part is that the manufacturer’s firmware for Shelly products does not support TLS; it is possible to flash other firmwares which do not have this limitation (Tasmota).
(I have a bad experience with Shelly devices, but they made me love MQTT.)
> This also rules out smart light switches, as these often need to connect to some kind of cloud, and if that cloud goes down, you could struggle to turn your lights on.
I live in an apartment, so I haven't done proper research, but this sounds odd. Why wouldn't a smart switch still work by physically pressing the button even when there's no connectivity (I'm assuming a smart switch that actually works as an electrical switch to cut off power to a normal bulb (Probably by using a relay or a really big transistor), not a button that just sends signals to a smart bulb)
"Making traditional light switches smart", the switches are not smart, they are dumb on/off switches that cut the power to the bulb.
People want this because they already have them and they come in a lot of designs that are usually used all over the home (and may not be available anymore) Running an extra wire to existing switches is often complicated.
You can buy smart bulbs that work with an app but if you switch off the power the app wont be able to switch them back on.
If you add a smart component behind the switch it can bypass the switch but then the switch cant be used to switch off the light.
You get 4 states pretty much:
The smart component would have to listen to the switch, if it changes state it should switch the light.
00) If the switch was off and the light was off it should turn the light on when you switch the switch on.
10) If the switch was on and the light was off it should invert and turn the light on when the switch is switched off.
01) If the switch was off and the light was on, it should also invert and turn the light off when the switch is turned on.
11) If the switch was on and the light was on it should turn the light off when you switch the switch off.
Powering it is also an issue as there is usually only a + and a switching wire running to the switch, no - or ground.
>This also rules out smart light switches, as these often need to connect to some kind of cloud, and if that cloud goes down, you could struggle to turn your lights on.
Baloney. None of the smart switches I have ever used/seen stop working without internet/cloud connectivity, in fact some of mine are disconnected right now as I haven't gotten around to reconnecting them to wifi after my last change. This includes switches with bridges (lutron caseta etc), and wifi switches (WeMo, LeGrand, etc).
Heck - I set up a Caseta switch in the back of my house with a pico remote without ever connecting it to my bridge (which is in the front of the house and likely won't reach that far without a repeater), yet both the remote and the switch work 100% reliably.
The author makes a huge leap that is incorrect: most smart switches do not stop functioning if they lose internet functionality. The “smart” goes away without the cloud, but interaction with the switch will still turn the lights on or off.
Before I undertook to automate my whole home (over 100 devices) I adopted two simple requirements that overlap:
1. Everything must “fail normal” - it’s ok for the smart to go away if there’s a network or device outage, but I still have to be able to use the device just like the dumb device it replaced.
2. I never want a tech support call from my family to turn on/off lights or other basic stuff, esp if I’m traveling on business 10 time zones away.
This led me to steer clear of smart bulbs and ceiling fans with built in remote controls and anything else where the switch had to remain in the on position.
So I have smart switches AND smart bulbs, because I like to change the colour temperature and colours at times. The smart switches are then configured to always power on the light bulb, and pressing the switch just toggles the state of the smart light bulb.
This works really well but it means that if the WiFi or Home Assistant goes down then it stops working. What would be great is if there was some way for the light switch to directly communicate with the light bulb even if they're not connected to a WiFi AP.
There is ESP-Now which seems to be a peer to peer communication protocol on the ESP side of things, but it would be great if something existed that could do this across Zigbee and WiFi devices. Perhaps this is something that Matter and/or Thread will help with?
If this is a thing you WANT to do, great, but even without wifi, smart switches still work just fine as normal switches. The statement that they're problematic without wifi is wrong. They work fine as normal switches, even if you never set up the smart features.
There problem with these are that if you put them in your switch boxes is that these have to be very deep. The traditional existing switch is already occupying half of the box and the rest is often not deep enough.
The author fixed this by installing new shallow switches. If he is lucky their covers matches these of all the other sockets and switches in the house. In my cases that combo did not work
The app looks like it is from 5 years ago, but a very nice project anyway. How often do users with smart home lights use the app and how often do you use the app?
In other words is it worth it or is it just a nerdy thing, which is nice but not really needed.
The UX of a physical switch whose position always accurately reflects the state of the circuit is hard to beat, and all the smart switches I've seen seem to use buttons or some other non-tactile solution that is significantly worse UX than a plain old switch.