So, in this case do you connect your smart things to GA and use their available UI and commands, ie give them your details in return for a much quicker, possibly slicker UI, or do you have GA pass a dummy variable to OH and then spend a lot more time and effort making rules and sequences in perhaps a not quite so easy to use way?
Or is the reality somewhere in a combination of the two? Do you have a family that are technically minded or is your OH UI pretty slick, if so how long did it take to get there?
I totally get what OH is capable of and the massive variations of things you can do with it - that’s why I chose to come here - I’m just wondering to what level I need to reinvent the wheel with using trigger systems such as Alexa / GH and the like.
My main example is my Sonos, I have 5 speakers, and all work perfectly with Alexa at the moment, to have to recreate that in OH with the same ease of use for all the family would take me ages.
Actually, it´s the last option. But it´s probably a bit different than many others due to my IHC system, (IHC - Intelligent House Controle). In short - It´s a controller with logic which can handle 128 input and 128 outputs, plus 64 wireless in and outputs). It is already a smart home system itself. And most logic is already done in the IHC controller. The limit is, it cant communicate with any other exernal devices which is why I combine it with openhab. In that way openhab (IHC binding) becomes a gateway between external devices and IHC. Then I can use everything back and forth. Pass on values from one sensor to IHC, or the other way around.
All logic for ligthing, heating and alarm system is build into IHC. I have added voice controle, Hvac overview and controle, skyline windows controle, extra present detection (WiFi, extra motion sensors etc) by using openhab as a gateway.
Still have several things to add, such as be security (door/window sensors), eletronic locks, security cameras as well as a highly advance combination of motion detectors for our outside lighting, which turn the light sections on/off, on the move (follow movement). Because of the way our house is build, that part requires quite a lot of advance motion detectors, which I havn´t found yet, placed in strategic places and a complicated (i think) logic.
When I started this path, I had to take into account… What if one day I get tired of openhab, or for some reason have to sell the house etc… This in mind I have to keep all orginal logic inside the IHC system, and just add logic from external devices through openhab.
As for Google home it self… No devices except Philips Hue, is connected to Google Home. (Hue was just to test), cause I have chosen openhab to be the gateway, it wouldn´t make much sense adding devices to Google Home. If openhab havnt got support for a specific device, then I dont need that device
My family is far from technically minded. If fact, they dont understand much of what I´m doing, and dont share the same enthusiasme either, except for the voice controle… That part is something they seem to understand and try to make use of and getting used to. It will take quite sometime for something like voice control to become a “normal” way to controle stuff in general. I guess that goes for most people.
The UI is/have been a very slow work in progress situation. I had to test and try alot of stuff first, before putting too much time and focus on the UI. I use basicUI for testing stuff. But the last few weeks I have turned focus onto using a SVG floorplan in Habpanel, cause I really want a easy overview of every part of the part. Knowing very little (read- nothing) about vector design as well as html/css, this is very complicated for me. So atm it is very far for beeing slick in anyway. It´s quite basic actually. But I´m getting there, slowly and struggle by learning.
Bad question It feels like forever. I have actually only used openhab for aprox 1½ year, spending alot of time learning openhab, experimenting and dealing with all kinds of stuff, some working others not working. But I enjoy it, as it has become my hobby. My worse problem is to limit myself, and get one thing done before starting another. But often I get stuck on something, then leave it for a while doing something else, and then return to the deal with problem… This kind of aproach do cost quite some extra time I´ll say. The positive part is, I learn alot of different things, only need to figure out, which I dont want to use
This is as far as it has come. Not fancy looking, not slick, but it´s working. I can turn on/off layers on the floorplan, I do not wish to look at, from the not so fancy looking buttons on the down left corner.
Atm I can switch off any climate info incl the very silly looking Hvac, which I added last night, as well as all the PIR´s. PIR´s are off by default… My main plan is to have the be security and ligthing showen as default only. Outside ligthing is yet to be added.
I havn´t got much experience with the speakers yet, (neither Sonos or others). To my understanding it should be pretty easy in Google Home to define speakers and groups. I did try with my HiFi receiver (chromecast support) and my Google Home devices (speakers), which seems to work fine, beside there seem to be a problem with my HiFi receiver, I have yet to investigate. But others are mentioning that speakers are working just fine. I have no idea if it´s better or worse than using Alexa.
You might not think it’s fancy looking but that looks great, lots of info all in one place and pretty much self explanatory…
So, if you use GA and expose a yes / no type dummy variable to it so that when you say ‘GA lights on’ it sets the value to ON and so OH runs the required rules…
How many dummy variable and GA trigger phrases do you have? I already have 6 and thats just to do some basic things, I’m going to need at least 4 for each speaker, (play, pause, next, previous), so that’s another 20, (5 speakers, 4 commands), and then there is the thermometer, lights, cameras etc…
It seems like overkill to me I suppose to have so many dummy variables and Alexa routines but I guess that is what is ‘under the hood’ of Alexa anyway…
Sonos have invested heavenly in polishing there application and product range, why would you want to try and recreate that? As you’ve pointed out it would take you a good while to do so and your can find better things to do with your time. No point in reinventing the wheel unless you can make it better.
In terms of Alexa and Google Home, you need to think of them as input devices. They offer you no automation at all, remove your access to the Internet and they can’t work. Both devices are being sold as “smart” a phrase that I find annoying, because they aren’t clever on there own. They certainly can’t think for themselves.
OH doesn’t focus on having pretty UI’s, that’s just a nice to have. The main focus is on integration with as many devices as possible and to be able to make them work together.
I also have a number of Sonos speakers around the house along with a few Echo Dot’s and a whole host of devices. Using a number of rules I have created a solution that works for me and my family. Each morning when someone goes down stairs and into the Kitchen, they are greeted with light if it’s still dark outside. In addition the Sonos speaker will greet them with the present weather conditions and today’s forecast. As the next person enters the kitchen, usually myself, the greeting changes and provides the latest travel time to the train station and the status of the next four trains that can get me to work. On the weekends or when my wife has a day off (the same day each week) the Sonos switches on the local radio station.
I hope the above example gives you something to think about. Oh and I’m only just getting started with my home automation, I have far to many ideas of what to do and either don’t have the coding skills at present or the device I’d like doesn’t exist yet. So it’s a constant learning curve, for me at least.
I agree, it’s what I’m thinking but you’ve put it more elequently. My struggle currently is I’ve got so many ideas and OH offers almost complete freedom to achieve it, I just don’t want to be redoing something that has already been done by someone cleverer than me, (or with more time and money).
My problem is having too many apps / inputs to control everything and I was hoping to use OH to do it, but I realise OH is more the engineroom behind things happening. The interaction needs to be family friendly and easy to do, hence my current thoughts on using Alexa.
That then leads to the issue of things like Sonos, I can connect my sonos to the alexa skill and everything I need is already there, ‘Alexa play playlist 1 onin the living room’ etc etc etc, but this comes at a cost of giving Amazon my sonos log in details etc…and I came to OH to have a ‘local’ home automation solution.
I like what you have done regarding the sonos rules you have, I have a couple similar based on cron times, but what if you just want to play an album from spotify on them, I guess you have to get your phone and open the sonos app to control it rather than using say Alexa, so as not to invent the wheel, but then you are back to more apps controlling your home…
It’s a real catch 22 for me, the more I see potential the more I see trade offs between time, effort, money and local v cloud services…hence the orignal questions…I just wondered how others deal with this…
Thanks… It will be changed though to a much more decent design And when I find a decent 15 inch (minimum) touchscreen (monitor) with a lot better resolution than my stupid Huawei tablet, then it will be scaled to a higher resolution as well. But for now, this has to do while Im working towards the final design.
Like you say, the idea is to have all info available gathered on one screen/display only, with an option of swich on/off info I dont need to look at. This is not possible with default Habpanel or BasicUI. (I tried default Habpanel, and I hated it).
No rules required to do simple ON/OFF stuff.
Google Home (Google Assistant) passes on the voice command to openhab´s item, through the myopenhab.org Google Assistant connection. Openhab then pass on the ON/OFF (value) to the IHC system, which finally turn on/off the light.
I can open or close the garagedoor from my car (using build in Android Auto, cause it uses Google Assistant). Again the voice command is passed on via Google Assistant to openhab, which then pass on the command to IHC, which will open or close the garagedoor.
It is as simple as that.
As for the rules automation:
I have a IHC PIR outside our main entrance. If this PIR trigger when I´m not at home, IHC will send the trigger info to openhab, which from a rule will send me a pushover message to my mobilephone. If I´m at home, openhab it will send a notification (speech) to one of the Google Home devices.
More rules:
Whenever it gets too hot inside, I have a rule which opens the skyline windows automatic. I can also controle the windows from voice command, if I want to.
These rules are actually rather complicated, for me. But I´m still learning. Thanks to the community and specially Rich, at least I got something to work, sometimes
This is actually pretty simple but very powerfull stuff. It only require openhab items and tagging. I believe it´s the same for Alexa. So it´s not really dummy variables… It´s just simple items and tagging. I havn´t count, but I guess there are aprox 50-60 items with Google Assistant tags. And there is still a need of more. But it´s easy done.
Whats not easy are the rules for automatic running. Thats the complicated stuff
@MadFrankie how about you try something that may or may not help you. I’m going to guess you have One Note available to you, if not then a notepad app works just as well.
Write down your ideas and how you think they can be achieved. Once you have a list, pick the top 3 and work on them. Small steps at a time.
Most of the apps you’ll get with “smart” devices need the Internet to work, you don’t need the Internet for OH to work.
Thats two different issues not related.
OH do offer some automation from rules in some countries. But it´s limited as far as I have been told. We´re still waiting for this in Denmark. But when it gets here, I dont think Im going to use it anyway. I hope openhab will develope much faster on the Google Assistant integration, so we can get a more decent controle. (ie… atm I have to say “turn on the garagedoor” to open or close the door because openhab/Google Assistant can´t understand open/close yet.
The required cloud connection is almost the same issue for all cloud based devices. If the internet goes, the connection to the device often goes as well.
In short - Simply take advantages of whats working and leave whatever isn´t working. I dont think you´ll ever find one system beeing able to do everything. If you have an idea, you either have to try yourself, or ask if someone has already tested the idea with 100% success… Probably some have tried with some success, and others have with poor result. That doesnt necessarily mean, it doesnt work. There can be different ways to accomplish what you may need. The high prize is time, unfortunatly. I know from experience
I’ve found something that may be of interest @MadFrankie. Have you read this thread?. It’s from someone who’s created solution for their needs around Sonos. Save you reinventing the wheel.
Hi, that’s what I’ve done, I’ve done the ‘automatic / hands free’ rules, if temp is x or if it is a certain time then do that etc - the rules that happen without a user input but now I’m at a point where I’m trying to create a usable home automation system where an input is required, play something, reduce temp, bedtime routines, and so I’m trying to work out the best trigger.
I don’t want to have to open my OH sitemap every time and I know my family don’t want to and so won’t use a tablet or panel so it led me onto Alexa, which is already widely used in my house - remember any success of managing change ususally comes from subtle integrations using something that people are comfortable with already.
The issue I then descovered is that Alexa kind of already does most of what I want with existing skills but at the cost of it obviously being cloud, not local as I had set out for it to be.
I just wanted to weigh up the options from people that have had the same thought / design process.
I think that is the way I am currently thinking. I may leave everything working in the cloud so I don’t lose any abilities and then pick them off one by one and bring them into my local set up…
Thanks, I have something similar to that, (I may have ‘borrowed’ some of the logic from here ), but it’s the triggering that I am wrestling with. I know my family won’t use the sitemap so it is just coming up with a way to trigger / interact that they will…
The best trigger depends…
Bedtime routines… I guess the best trigger is time. But if normal bedtime isn´t static, another variable would have to be added as well. Ie… weekdays one static time, weekend another static time, or perhaps weekend, no bedtime routines at all. It all depends on what kind a behaviour you would like.
Reduce temp - Again, the trigger can be different things… Most often (if it´s a room thermostat) it´ll be the room temperature. But it could be present detection or a combination of both, time of year, season etc…
Untill lately I had my skyline windows triggered from the outside temperature. I changed to inside temperature insted, as it seems more natural. But next step is to add more dependencies like outside weather etc… I already got preset detection for this as well, (windows doesnt open much, if none one is at home/alarm is armed).
It can be hard to figure whats the best way to trigger something. But most often I´ll say, you probably already know the answer. You´ll just have to give it a extra thought or perhaps ask someone else what they think. Rich is very good at this.
Yes, it was the bedtime routine that started me down this whole rabbit warren…it’s a routine for my daughter when she goes to sleep, lullabies on a sonos speaker, low volume, lights down to night light etc etc. It’s not always the same time every night as she loves to mess about so I wanted something to trigger it off.
She has a Sonos One with alexa in so it seemed logical. I just set up a dummy variable in OH, allowed it through myOH and got Alexa to trigger it to ON with a ‘Alexa it’s my bedtime’ command…this then triggers a rule to do everything.
It’s been so successful that it made me realise that the easiest way for me to bring some of my OH ‘tinkering’, (as my wife calls it), into standard family use was via Alexa.
That then sent me down this rabbit warren of thoughts!!!
Things like that tend to burst my brain as well
Using Alexa/voice command for triggering a rule like that is very powerfull as well. The triggering is the simple on/off part. The rule (what to do when triggered and which circumstances) is what makes it complicated.
I’d counter with why you feel the need to recreate this. OH has Sonos integration and Alexa integration, yes. But that doesn’t mean you have to replace something that works with OH. Integrating this with OH shouldn’t break what you already have working.
Kim’s IHC example is a good one. He hasn’t replaced the IHC with OH. Instead he extends that capability by adding openHAB to the mix.
I can confirm, this is pretty easy to do in the Google Home app. I’ve one Group consisting of a Google Home Hub and a Chromecast Audio and there is no noticeable synchronization issues between the two. I’m pretty satisfied with them.
This stuff doesn’t come for free. If time and effort is a concern for you then yes, you will have to share your information with a cloud service like Alexa or Google Home. If that concerns you then you will have to spend the time and effort to set up an alternative. The services you are using are not going to make it easy for you.
So this is something where a little bit of knowledge may put your mind at ease. I don’t use Alexa nor Sonos so I can’t say for sure this is how it actually works. But typically the way this sort of integration is supposed to work is through OAuth2 (or something like this). The key reason I bring this up is if this is used then your Sonos log in information is NOT shared with Amazon. Instead there is a procedure that you go through to authorize the access between the two clouds for your specific user. This access can be revoked at any time (without needing to change your password).
Now, once the access is approved by you, Amazon will have access to all the things that you approved to to have access to (I’m sure you are familiar with that window that comes up saying something like “Alexa wants access to X Y and Z”). But they never get a copy of your password.
It’s small consolation, but it does address the specific concern that you raised.
With the visibility tag you should be able to do something like that in BasicUI. I’m not saying that it would meet your needs, but it is possible. You can easily create Switches or the like and hide or show elements of the sitemap based on the state of that/those Items. This is the standard way to show charts with a different time scales on the sitemap, for example.
You mean Google Asssitant?
It’s probably worth noting that Google is working on a local control API which will let the Google Home devices speak directly with the devices locally instead of going through the cloud. This will obviously require Google to partner with what ever company makes the devices and will be limited to WiFi and BT devices but they have at least recognized the concern and will address it.
Because it’s going to be strictly Google to Device I’m not sure how much utility it will provide in the broader home automation context, but it’s a start.
I find the best trigger is actually lots of triggers. You don’t have to settle on just one. Sometimes a physical button or switch will be most convenient. Other times a voice command will be best. Still others the phone app may be most convenient. So don’t limit yourself to just one trigger. Create several and let the user and context choose the most appropriate trigger.
The absolute best trigger is some event that doesn’t require manual human interaction. Though of course that isn’t always possible or desirable.
I cant have as much info on one screen/display using default habpanel or basicUI… I do have simular setup using BasicUI, but I have to scroll or put several devices/items into several sitemaps. the advantages with using SVG/floorplan like this is to have all info placed in one screen, no need to scroll, change dashboard etc…
Yeah, Google Assistant ofcouse
I look forward to see how they´ll address this. But I dont think I´ll be using whatever they come up with. Openhab suits me fine atm. But if it will be able to connect direct to local openhab, then it could be very interesting.
Interesting thread. My goals are, perhaps, too lofty. Privacy is among my most important goals, so no Alexa/Amazon/Google/Apple anything for me. But this is only one element of a bigger picture in my case.
I have a reasonably unique situation. My plan is, ultimately, to monitor a fully functioning farm which includes “typical” home automation functions, then soil monitoring, local weather, solar power infrastructure, water management, visitor alerts, etc. The ultimate goal is an operating farm with some intelligence behind it to assist in animal and plant management while expending minimal resources. As I get better with OH, the reach of the system is expanding. I’m not sure there’s really an end to it - there’s just building, expansion and testing.
I plan on looking at snips.ai for much of my control next as trying to manage panels spread out over many building and many acres is getting expensive and cumbersome.
Wow, that is a pretty lofty goal…I like the fact that you have an actual need for automation as I think that is vital for success, perhaps that’s part of my problem in that I am just tinkering with ‘nice to haves’…
I’d be interested to be kept up to date on the project, I’m sure lots of interesting bits will come out of it.
), but it’s the triggering that I am wrestling with. I know my family won’t use the sitemap so it is just coming up with a way to trigger / interact that they will…