Which way to go?

Hi There

I’m at the “should I buy” stage.

I have a brand new house, with (at the moment) no internet - long story! - and am currently debating what route to take. I’ve made a decision to learn to code, as I like the idea of formulating some of my own solutions, but I will need a base ‘hub’ to control it all.

I’ve a raspberry pi I never used, but that would be perhaps a lot more work to make it work in the way Smartthings does.

I know you can build your own smartapps and rules using Groovy (I’m starting with Python but as I understand they are fairly similar and you have to start somewhere) so I guess I am wondering if I should invest in Smartthings and get involved in developing smartapps (once I can code properly which could take a while learning part time) and still be able to use the generic functions as well as community led smart apps etc. or…

Take more time and build my own, similar system based on the pi. Something is telling me to go the Smartthings route and - seeing as this is a Smartthings forum - I’m sure you guys will tell me the same?

Thoughts and opinions from UK based owners specifically (although obviously all feedback welcome) as to how they have their home set up?

My goal is completely automated lights, door lock, window blinds, audio and security system plus garden features like water feature/lighting.

1 Like

[quote=“Foosh, post:1, topic:29260, full:true”]
Take more time and build my own, similar system based on the pi. Something is telling me to go the Smartthings route and - seeing as this is a Smartthings forum - I’m sure you guys will tell me the same?[/quote]

FWIW, users here have never been shy about SmartThings problems or short comings. Nor have we been against recommending other solutions if they work better for you. And ST as a company has never squashed users here from doing that either, at least as far as I know. So I believe you’ll get some pretty honest advice here. Of course we’ll lean ST but no one will try to sell you something if it doesn’t fit you best.

You’ll definitely want to get specifics from users over on that side of the pond. I don’t know if the OAuth issues in Europe have been fixed or not, but that will have some big impact some of the features you can or can’t use.

In general (from the North American perspective), here’s my advice:

On one extreme you have the: Pay someone to do everything for you model. You basically get a company to come in, charge you big bucks to install a solid, reliable system that will probably do 90-95% of what you want. You may be able to tweak some stuff personally to get that extra 5-10% or you may not be able to. But the important key words here: Dependable, Solid, Expensive.

Of course the other extreme: Do it all yourself. This will undoubtedly be cheaper, though maybe not as much as you might think. It also means you need to do it all yourself… that includes installing the hardware, learning the software, learning how to make it all work together nicely. There may or may not be a lot of helpful sites out there with sample code you can borrow so you’re not starting from scratch, but there will be a huge learning curve. Key words here: Lower Cost, Lots of Learning to do, Questionable dependability.

SmartThings sits somewhere in the middle, which for me personally, is ideal. I like to tinker with this stuff and while I’m not a professional coder AT ALL, I do have some background in it so I can tinker and build my own apps if I want. Here’s the disadvantages of SmartThings, in my humble opinion:

  • Questionable dependency --> unfortunately ST hasn’t always been solid. I think it’s been improving, but there’s still times when the backend servers go down and your automation doesn’t fire like it should. Hub V.2 helps a bit in that it has some local processing capabilities, but this is still limited.
  • Not FULL customization --> Don’t get me wrong here… there are LOTS and LOTS and LOTS of customization options, but you are tied to a Company who has their opinions on how things should work and you have to work in that frame work to some extent.

That said, there are the numerous advantages (again, my humble opinion):

  • Open hardware: SmartThings, outside of the hub doesn’t use any propitiatory hardware so whatever you’ve invested from a hardware side can almost certainly be used with a different system if you decide later to bail on ST.
  • Open software: The far majority of SmartApps that people have individually developed are freely available to use, modify for your own use, or learn from as you build your own SmartApps.
  • Great Community: I can’t tell you how much I’ve learned, both in terms of code writing as well as general HA knowledge here. Lots of really helpful people who are willing and able to guide you through everything from picking out the right hardware to giving tips on how to wire up z-wave 3-way switches to coding examples to basic information about wireless tech and protocols.

I think in some ways you and I entered the HA field with the same tool set: I wanted to get into some HA but I wasn’t great at coding (Pascal was the last language I really did extensive work in!). I had a technical background and could learn. I knew some hardware stuff, but wasn’t anything close to an engineer.

SmartThings provided, for me, that “leg up.” That boost to the first level. I didn’t have to create the whole back end as if I was doing it from scratch. But I also didn’t have to pay big bucks to have someone else to do it and be tied to their system. I can modify and expand as I need or want.

SmartThings provides a solid (though at times a little temperamental) frame that lets me accessorize as I desire. For me this is a perfect fit.

Sorry for the long post… I tend to ramble at times, but I hope this provides some basic idea of what I find really great about ST. Feel free to ask any specific questions you want of course and I (as well as others here) will be happy to answer.


I think this is the key bit for me - and also the other half, as it needs to be user friendly as well as give me the opportunity to tinker with it. So I think Samsung may have got it spot on if they can sort their reliabiity issues (I have been reading threads on here for a few weeks and that is a recurring theme!)

The overriding factor is the sheer number of different technologies I can implement. I have a few Osram lights which should connect instantly. I can wire up Fibaro dimmers into each light fitting (as opposed to the switches because they are slimline and it would be easier to install them at each fitting.) I have a Nest which can connect, and I’d like to look at geo fencing options.

Once I’ve got the hang of Python and I feel confident about learning Groovy as well it gives me more scope for customisation… just need to get the internet sorted now!

Thanks for the reply and insight!

1 Like

[quote=“Foosh, post:3, topic:29260, full:true”]
I think this is the key bit for me - and also the other half, as it needs to be user friendly as well as give me the opportunity to tinker with it. So I think Samsung may have got it spot on if they can sort their reliabiity issues (I have been reading threads on here for a few weeks and that is a recurring theme!)[/quote]
Again, take this for what it’s worth, but for me reliability has been generally pretty solid. Only very rarely have I had things go totally down. Maybe two or three times and I’ve been around since the very beginning. I was an original kickstarter backer.

I know others have experienced a lot more temperamental service and I’m not in anyway dismissing their pain or trying to indicate that they aren’t being 100% truthful. All I’m saying is that it may not be quite as bad for everyone as it, unfortunately, has been for some. Few people will take the time to write a post saying: “Oh hey, everything is been mostly great for me!” Just about everyone who is in the forums and has problems with complain (understandably so!).

Well, like I said, LOTS of people here will help you with learning groovy. I had programming experience so I knew how to think like a programmer, but I need to learn the whole syntax of groovy. When you get to the point of trying to look at groovy, take a look at some of my SmartApps. As a general rule I try to document the hell out of them and explain as much as I can about what the code is doing. Just my way of trying to pay back the community a little bit.

This is the bit I’m finding tough. I’m using a website called codecademy and following through their interactive tutorial. Its starting to make sense. I have a long way to go!

Exactly the kind of help I’m looking for and thanks!

This I understand. I’m in the retail industry so I know the phrase is - “have a bad experience you will tell 10-20 people but have a good one and maybe you will tell 1-2” Standard to only really read bad experiences. I’m willing to bet on the future and I think Samsung has a good idea and (evidently) a solid, helpful user base willing to help newbies like me!

Honest to gawds truth?

You don’t ever have to write a single line of Groovy if you can cut and paste.

For reasons that are still completely incomprehensible to me, you can post any requirements here and, within hours, a half-dozen or more people will volunteer to write a SmartApp and/or Device Handler for you, at no charge whatsoever.

Of course, hopefully a few will first point out existing features to accomplish your request. And there will be another half-dozen complaining about how the product should have this built in and what the others are coding is unlikely to be reliable…

I can’t think of an analogy to a different industry. SmartThings is an anomaly.


I’ve been reading the forums for a while and this is also something I have noticed. Not for a while have I come across such a helpful and active community.

Looking forward to getting involved.

It is a goal of mine to become proficient in code, so the additional help here will also be useful


I’m having difficulty justifying smartthings at the moment.

Lots of issues with certain things not working, stories of downtime and routines not firing…

I need a stable, working system I can use to make things easier. I don’t want to have to be reaching for my phone every two minutes or resetting the hub or re-connecting things because they have failed.

Can anyone re assure me? Or is home automation for the diy enthusiast just not there yet?

I’m willing to learn and put in some effort, but “she” isn’t. And that’s the crux. A light switch just works…

1 Like

Hue Bridge is stable and reliable. As is Amazon Echo. Lutron is very sold and reliable for lighting. A Leviton-only zwave lighting solution is also stable and reliable.

Harmony is mostly stable and mostly reliable, although when they do a big update things often break for a couple of weeks.

And there are a number of purpose-built security systems under $2,000, with excellent stability and reliability.

SmartThings is many good things, but stable and reliable, not so much. On the days when it works as intended, it’s my favorite home automation system. Those days have been few and far between since the big platform update few weeks ago. So then it comes down to whether the value that you’re getting is worth the unpredictability and the maintenance cost and time and effort. And that’s a personal calculation everyone has to make for themselves.

The Future is Coming–but it’s not here yet

My personal belief, as I’ve often said, is that by the summer of 2016 there will be several reliable, stable, plug-and-play home automation systems in the budget range of $200-$500 per room, with some voice control. I think HomeKit/Insteon will be one, works with nest will be another, and a few more. I expect SmartThings to be one of those candidates, but they have some work to do to be able to compete. We’ll just have to see. I’m not planning to make up my candidate list until May or so.

The recent mode failures are very difficult because they can throw off everything else you’re using. Just as the presence sensor failures can. But I don’t think that reflects the IOT industry in general, as the Amazon echo and Phillips hue show. Both continue to offer new improvements every month, while maintaining existing features and reliability.

So the overall answer to your question is that unpredictable reliability is not a defining feature of the current state of IOT home automation.

The real question is more specific

That doesn’t answer the more specific question of what you should be relying on SmartThings for at the present time. There are some people that it’s working great for, which is good. But it’s not clear what any common factors might be. It’s nothing as obvious as using a particular device or a particular smart app or a particular command set.

If it were, obviously support, who have access to many thousands of data points more than we do in the forums, would’ve at least found a workaround already.

So the current situation is frustrating, but I don’t know if there’s any meaningful answer beyond looking at each situation individually based on that household’s needs and preferences.

I have read pages and pages of different solutions. Different hardware, different software. I’ve researched using a Pi, using openHAB, CastleOS, Smartthings, Echo, or a combination… What I want is voice control of the lights, heating and (eventually) blinds as well as a smart video feed to the front door along with a lock. Like a portable intercom, controlled on an ipod touch, or an old android smartphone.

Ideally a set up where I can be ANYWHERE in the house or garden and say “House - xxx” and it will do what I want. Reading everything I can ALMOST get this using sharptools and tasker and a voice control app. But it seems I have to say “OK Google”. My house is not called “Google”.

I did think about using a Pi in every room, running some kind of voice control software that can then activate the individual things I want controlled. I’m happy to have microphones or bluetooth speakers - can Sonos recognise input or is it output only? - around the house.

Maybe I want too much and this sort of automation isn’t quite there yet?

Voice control can be done using various combinations of hardware and software. No plug-and-play solution for exactly what you describe but there is a community member who has a project report doing pretty much exactly that. This is a very technical approach, though, and requires a lot of set up and maintenance.

I myself am quadriparetic, so voice is very important to me for home control but at the same time I want a very low maintenance solution. So I haven’t gone this route, but if you are comfortable maintaining multiple servers, you do have some choices.

Anyway, it’s an interesting project to read about!

Ave you had a chance to take a look at the voice command solutions thread? That’s where most of the discussion is. That thread covers all kinds of approaches, some technical, some much more plug-and-play. It also discusses most of the android options.

For those who don’t want to run multiple servers and have all of those issues, the simplest equivalent would be to wear a smart watch and use voice text to IFTTT You can do that from anywhere. As to what awake word is required to launch it, it’s going to vary based on the solution using. So that method doesn’t give you all the same custom options as running a ServerNet, but it’s quick and easy to set up.

I personally use a mix of voice control options. I mostly use Echo inside the house, and it’s great. We have a fairly open floor plan, and one echo easily covers four or five rooms plus we have a second one for the master bedroom suite. Then I use text from my watch for when I’m in the yard or someplace where it would be harder for the echo to hear me.

As always, different things work for different people. Some of this is aesthetics. It doesn’t bother me to use the same awake word for home control commands that I do for other things from my phone/watch. But that’s just me. :sunglasses:

1 Like

The Jarvis option JDRoberts mentioned above is indeed a very technical approach. I have a youtube video where I’ve kind of walked through it, I intend to re-do that video to give a little more instruction, because it CAN and DOES get confusing and frustrating if things don’t work right. It does require several pieces of software, and it IS recommended this system operates on its own platform (pc, mac, whatever), because it utilizes the computer’s default microphone and speakers, so if you use that computer for personal use or anything else, it can get quite annoying. Once it’s setup though, it doesn’t require all that much maintenance, with the exception of the fact that you’ll have to set Jarvis to have whichever command you want to use. The Jarvis platform I’ve put together uses the following resources as one cohesive unit to produce an actual smart house ‘entity’ if you like:

  1. Jarvis Mark II (ai-dot.net)
  2. IFTTT (ifttt.com)
  3. Teamspeak3 (you can use Mumble or Ventrilo, any of these will work)
  4. EventGhost
  5. IFTTT Maker Channel

(you might not need eventghost if you can make the maker channel do it all for you within Jarvis).

Anyways, the idea is to have Jarvis, a program that uses the windows voice recognition software, to ask questions or make commands. Jarvis and you both communicate inside of Teamspeak. Jarvis has the ability to create eventghost events OR IFTTT Maker curls, which in turn triggers events in IFTTT to automate anything else connected to IFTTT, including smartthings.

Anyways, it is indeed a very technical method, but it’s rewarding because it’s diverse to whatever you want to (i.e. it’s extremely customization).

1 Like

Sounds very interesting. I’ll do some reading over the weekend and see if it will fit the bill. I do have a couple of spare Windows PCs lying around that I was going to make into a media server type thing…

1 Like

You could use AutoVoice to have your Android device continuously listening for whatever keyword you want and then respond to that. I personally prefer the use of OK Google as my phone is optimized to listen for that phrase even when the screen is off, so it is really battery efficient. (The AutoVoice approach can be quite a battery hog, so this would be better suited for a device which is plugged in and dedicated to home control)

I’m the developer of SharpTools and my preferred multi-room voice integration is through the Amazon Echo. Sure, I have to say “Alexa, [do XYZ]”, but it just works with no fiddling around and Alexa is really good at understanding different nuances of how I ask her to do things. I still have SharpTools + Tasker setup with voice commands and it’s a great solution for the bedroom where my phone is typically on the nightstand – instead of having to raise my voice so the Echo can hear me in the other room, I can tell my phone to turn off all the lights in the house or turn on/off a bedside lamp.

You can use multiple approaches for voice commands and while they aren’t perfect, it’s better than having no voice commands!