So, last yesterday I updated my Homey to the anticipated/dreaded 2.0 firmware. In went the new – according to Athom – ”gorgeous” Android app, out went the old desktop app that I’ve used for quite some time now. How did it turn out? Well, just as I expected, I’m sorry to say. Disastrous.
OK, let’s start with an introduction. I’m Magnus Orest, I live in Stockholm, Sweden. I’m 44 years old, systems developer by trade. I’m also a trained concept developer (what others call ”innovator”, but I’ll leave out my thoughts about the common misuse of that term), designer/UX architect, communicator and solution architect.
Or to put it bluntly, with suffient time and money I could build my own home automation gateway. But since I have neither I’ve bought one off the shelf. Why? Look below.
You see, I’m also a real geek and love home automation, and have been doing it for a few years. In my Homey setup, I have roughly 50-60 devices from Philips, Sonos, Broadlink, Greenwave, Nexa, Qubino, Telldus, Fibaro, and Sensative. For each device there are a bunch of flows. A gross estimate says maybe 200-300 flows. Besides these flows I have a number of integrations through IFTTT to other services like Google Home/Assistant, Google Calendar, and probably more.
In short, buying a gateway off the shelf was the only feasible option.
That one happened to be a Homey, since it was the only option if you wanted to throw just about anything into it. Decent hardware, a well designed and innovative firmware (hardware support for just about everything, then you expand the software by installing apps that use the hardware, very flexible), as well as a decent desktop application. Yeah, the desktop app isn’t really that, it’s just a wrapper for a web browser, and you can run the web interface in Chrome if you like.
I always had a bit of a love-hate-relationship with this desktop app. It had a decent overview, but was in itself a good example of one of those apps that show up on UX- or interaction design-courses. As an example of how you should not do it. But I learned to accept the quirks. The flow editor wasn’t that bad, for example. It was divided into three columns. One for the trigger, one for the conditions, and one for the actions.
I wrote trigger. Yes, each flow can have only one trigger. That’s a weakness, and means that it feels rather stupid to have an entire column. The columns themselves were also kind of wasteful with the space at hand since a flow with a lot of cards in the action column meant a lot of scrolling. Even if you’re using an ultra wide-monitor like I do.
But nevertheless it worked. It gave a fairly decent overview. It was workable. And, as you might be able to deduce from the amount of devices and flows (but only one user, which might be obvious ;-)) I really need a good overview when I work.
I’ve built wonderful things with it. Or at least very geeky things:
In my bedroom I have an ”artificial sky” built with a few Philips Hue lightstrips. They illuminate the ceiling with different colors from four different directions (north, east, south, west). I have flows to emulate sunrise, sunset, daylight, winter sky (blue-ish daylight), forrests, thunderstorms, and so on. Since it is RGB, it has all wavelengths that you need to get a particular effect.
I have a 47″ TV in my bedroom. When not in use, it’s hidden in a hollow footboard with a small lift that raises and lowers it depending on if the TV itself is running or not. I use a fibaro wallplug to measure the power consumption for the TV, and when it crosses the tresholds set, my Homey starts a flow for either rasing or lowering the TV. This is in turn done by a relay switch from Fibaro, that’s connected to the control unit of the TV lift.
I have a few Google home mini that I use to give voice commands to my Homey through IFTTT. I can use my phone as well. Since I have an IR-unit inside the footboard I can toggle the TV on or off and set a channel with Homey, with voice commands (no, the TV doesn’t have voice control).
Smart bedroom blinds
The blinds in my bedroom are also controlled by my Homey. When it’s morning, they go up. When it’s time to sleep, the go down. When it’s very sunny and the temperature outside goes to high, they go down. And so on.
Kodi with voice and ”motion control”
I run a RPI3 as a media player in the bedroom, and have the Kodi app installed on the android TV in my living room. Both are voice controlled by Homey. The one in my bedroom also pauses when I get out of bed, because I have a motion sensor under the bed that will only trigger when something is on the floor. When there is no movement on the floor, it resumes playing. If I enter another room, Homey will know that I’m not in the bedroom anymore (another motion sensor triggered), so it will pause again.
Smart vacuum robot made smarter
Since I spend all my time on home automation (I’m joking), I have a Roomba that does the job for me. But I don’t want it to eat my snailmail when I’m at work. So whenever the mail chute opens, it sets a lock in Homey. And when it’s set, Homey won’t start the cleaner.
Everything in one gateway
This is the beauty of Homey. You can combine different protocols and platforms! When I start a movie on the Kodi I have in my bedroom, it turns off the lights and closes the blinds. When I fall asleep after a certain time, it warns me that it will turn of the TV (as well as any lights) one minute after the movie has stopped playing. Or, if it’s earlier in the evening but I want it to turn of anyway, I can just shout it to my google home mini. Or the other way around, if I don’t want it to turn everything off, I can tell it to cancel the auto-off.
A smart home
The point is that with all these possibilites, you can turn your home into a home that is actually smart. Ie, with a certain amount of logic and deduction. Light switches that are remotely controlled or run on timers aren’t smart. They won’t take action depending on certain conditions. But if I forget to close the fridge door, the temperature rises and the sensor that tells if the door is open or not says it’s open, it can send me a message on my phone if I’m not at home. If the same thing happens while the sensor says it’s closed, it can deduce that it has probably broken and that I should get home ASAP before my kitchen is flooded with melting ice.
But all this is history now. You can’t build flows that are this complex with a mobile phone! It takes forever, and often the Homey app will crash when I try.
Yes, it does work if you just need to turn lights on and off, and have no need to combine different platforms or protocols to work together. It probably works even if you’re more advanced than that.
Here’s where Athom went wrong. If you’re not a geek (or advanced user), you might not be looking for something like Homey. It is – or was – a great solution, but it’s also quite expensive. But if you are a geek that will stop at nothing short of building a robotic wife or husband to take care of everything, Homey is what you need. I see no other option than Homey. Maybe Home assistant, but that’s not an off the shelf-solution.
Another thing they might not be aware of is that it’s the geeks that actually sells Homey. We build the things you can’t build with other gateways, and that makes others choose Homey. You don’t buy Homey because someone tells you that you can turn your lights off when you’re not at home. You buy Homey because someone shows you that you can turn your lights off, close the blinds and start a movie on Kodi just by getting into bed.
And here’s what I think is the problem with Athom – they focus too much on innovation and totally disregards stability and maintainability.
Essentially, it’s a bad corporate culture. They behave like Homey is some kind of open source or freeware-solution. But it is not. It is, as I’ve said, quite expensive. You can’t disregard your customers and say ”well, the mobile app will work for most users”. It will, but not for those users who actually sell Homey for them.
Personally, I don’t know what to do if they stick to this bad decision of theirs.
Right now Homey isn’t a workable solution, but I don’t know if there will be any alternative solution on the market shortly (I’m sure others are working on one since Athom dropped the ball, though – they have proven the concept, now it’s just for someone who knows how to treat customers to pick up the ball).
Anyway, I can only conclude that Athom would do well to hire people who understand the concept of customer care better. The entire home automation business would, as a matter of fact. As long as it is run by geeks who thinks ”it usually works, usually for most people too”, it will never grow beyond what it is today. We need responsibility and accountability to turn these solutions into solutions for all people, not just geeks.