Automation choices for 2019

[wkearney99]

Home automation software has always been a series of compromises.  One system has a decent UI, another has a particular driver you need, other have better programming environments, etc.  Lots of things have gotten us to where we are now.  

These days there's quite a lot of coverage for devices.  There's perhaps more flexibility now on just being able to integrate more, rather than having to use something because it was the only thing that supported your devices.  
 
It's probably far to say that a great many home automation efforts are little more than fancy remote controls for lights.  I know, in this audience there's a lot of exceptions to that generalization.  But of the ones that have risen slightly above basic remote control, how many of us haven't taken it to greater levels because of their automation framework?  How many are willing to entertain making changes if there's 'something better' available?

I pose this question because while I've had very reliable results using a particular solution, I'm always less than thrilled with the somewhat unique way it approaches programming.  Every time I start digging in deeper I find myself asking "should I just look at something else instead?"
 
Note, I'm not specifically 'naming names' here.  There's enough defects in almost all of the existing solutions to find blame.  It's not my intention here to bash any one particular implementation.  Rather to invite a conversation about what's on the market now and if anyone's made a change to one of them from a previous solution.
 
Anyone else made a big transition from one kind of automation system to something new?
 

 

[lleo]

Bill, you articulated well the process I am also going through. Previously I advocated that automation should not be visible to the eye, and all I cared for was the integration between various systems, but lately I am longing for something that would more please my eyes. Whether on the wall or my phone, I am looking for more eye-candy.
 
Recently, I installed one of the open-source freely available systems, wanted to see how much they evolved. It impressed me by finding most of my devices, except one which is EOL anyway and configured it, ready to be controlled from phone or browser. But it did not have any of the logic or automation I accumulated in my previous system, and to transfer all that sounded daunting, so stayed with the old.
 
Home automation and the smart-in-the-name of things became even less defined over the years and means just about anything for folks out there. And many are not about automation at all. They are about data and data collection. Of all big players, not even Lutron (just to pop 1 name in here) could resist the urge to start collecting data.
 
For me, it remains to live with my electrified entry doors and automated garage, automated shades and blinds, RA2 lighting with software processor providing conditional logic, security and secondary fire alarm, whole house audio and 2 TV/multimedia locations that just work as expected. None of this requires the internet or cloud and they work as long as there is electricity. Measuring detailed electricity use, water use, water leak, temperature/humidity/pressure in 16 locations, and several other environmental attributes, weather including rain and snow outside are probably superfluous and not needed. If I ever move, not sure I would do it again, but I shall never say ever. I did not adopt technologies like voice control, or anything that is cloud based. Importantly none of the above are very visible, so any guest can get around and use it without causing trouble.
I do not know how many hours of tinkering this took, but probably this is key compared to lot of folks who would never spend the hundreds if not thousands of hours to bring a home to this level, and most many would not have a budget to even invest this much in it.
I do not work or have a background in electronics/computers, although I know a lot about it. I interestingly watch where it will evolve from where we are now.

 

[jeubanks]

I also emphasize with the assortment of systems and possibilities.  I feel as though I'm on a never ending quest at times.  I've ran almost every "common" system I can think of whether it be commercial or open source and with each I find something lacking.  Either the UI is horrible and makes it a very daunting task to get things setup or the automation logic is ackward and sometimes just backwards which makes it difficult to really get things setup correctly. 
 
Traditionally I've also been all about the automation sight unseen but now I'm wanting to put up some tablets for dashboards and get the "wow" factor out of the wife and kids.  I think for that I'll have to look at other systems again. 
 
As I type I'm in the process of doing a conversion of lighting from z-wave switches to Insteon.  Why?  I had a major z-wave "crash" so I decided why not?  I never used Insteon before so I took the leap to explore that territory.  Going to Insteon also opens the new possibilities of exploring other systems that I had not tried before.

 

[LarrylLix]

Sounds like ISY994 with a zwave board andan Insteon/x10 PLM would fit your bill. Not many ever change again after that.

Sent from my SM-G930W8 using Tapatalk

 

[BraveSirRobbin]

Seeing home automation develop/progress for many, many years I've decided the only way to get a system that "I" want is to find a hub/software/system that can support as many of the eclectic and disparate systems I want to use.  There will never be 'one' ultimate system that will satisfy the majority of serious home automation enthusiasts, especially with recent examples of (what I call) dumb downed systems made to do one simple task (without any options) so people don't have to do much thinking to get that one system to work.
 
The reason (in my opinion) is there isn't any money to develop such a complex system that would warrant the development (and maintenance) time needed to produce it. 
 
This is the main reason I use HomeSeer.  It isn't perfect, but it does offer the capability to interface with a large variety of automation and security products out there while bringing in the human interface features wanted.  Of course one has to live with the possible instabilities of dealing with a variety of developers that produce the plug-in interfaces to this product (as well as other problems).
 
There are competitors such as CQC which seem as capable but I have no experience with them.
 
Of course there are some hardware/PLC vendors who attempt to interface with a variety of products, but I haven't found any that can do everything I (or most serious users) want.
 
I also believe most people out there don't want a full blown automation system and are perfectly content with turning on their bedroom lamp or looking at their front door camera from their phone.  Maybe they go out 'full blown' and get a canned security system that can be controlled via their phone, but that's about as complex as it gets for the general public.

 

[wuench]

I agree, complexity is the primary issue.   There are  just too many disparate devices, protocols, etc out there and a constantly shifting landscape.    This leaves you with some hard personal choices.    It also seems important for a product to compete to  also provide some sort of out of box/limited solution just to get new users onboarded.     For me,  I have  made it a point to select  solutions that  I can customize to suit my needs.    I went from Homeseer to CQC to HASS.   In CQC   I was able to write a complete SNMP driver at the binary level and lots of other stuff.  And you get what  you pay for, which is support and stability and it has a positive supportive community.
 
For HASS, being open sourced, I can write modules or even modify the source code to fix issues as the crop up.   It being in written in python aligns with my day job.   But I have  yet to use an open source product either at home or at  work, where I didn't have to tweak the source now and then.   That's the downside  of open source  projects, they are generally poorly managed  from a quality control perspective.   You shouldn't assume they are secure, or tested.   There is no one to call when it breaks, you log your issues and wait or fix it yourself.  I have also found it important to have a good rollback strategy.   So I have moved  to docker/git.  Hell, I might even have to deploy jenkins at home at some point.   But for out of box I don't think you'll get  too far,  unless  you are willing to limit yourself. The most common threads on their forum are about people trying to deploy HASSIO, their out-of-box RPi solution and make it do things outside of  its limits.  And their forum is full of a lot of people, both positive and negative.  More Reddit than CT

 

[wkearney99]

It's really fascinating, in the 'watching a train wreck' sort of way, how home automation has lurched along.  We all keep slowing down to gawk at the accident but then realize we have better places to be and move on.

Automation products are often like this.  We stop to wonder/gawk/rubberneck at a particular product, maybe even dive-in to help a bit but then discover, it's a tar pit of time wasting pointlessness and stop bothering.  Rinse, repeat.

As for "outside of its limits", yeah, I've learned over the years the MOST important thing is to know what sucks about a product. 
 
Not what's great about it, but what kind of hidden minefield of surprises exists, ready to blow up in your face.  Or, more importantly, drive the wife crazy and unnecessarily complicate your marital relationship. 
 
It's hard to ask these questions, because there's a ton of fanatics that build up what's practically a cult around any given product and insist it's equivalent to the Second Coming of Christ.  That everything else is trash.  And any criticism of it is BLASPHEMY.    This is not a new phenomenon.

Cue eye-roll and my exit...

I think what has dampened my interesting in HASSIO has been the whole "make it easy to install" approach which then relies on either underpowered/unreliable hardware or throws you into a sea of complexity with containers.  Containers are great, once you have a firm definition of the things to be containerized.  But if not you've then got a whole other layer of things to 'work around' and barely any fucking documentation on doing it.   

Which circles back around to my perspective of "show me where it sucks" and I'll make an educated decision on how much of my time I want to risk investing in it.

 

[jeubanks]

Sounds like ISY994 with a zwave board andan Insteon/x10 PLM would fit your bill. Not many ever change again after that. Sent from my SM-G930W8 using Tapatalk

 
Currently in the test bed.  I'm on UDI forums as well - simplextech

 

[LarrylLix]

Then I didn't need to tell you what might "fit your bill". You seem to have a good handle on most things already.

 

[jeubanks]

Then I didn't need to tell you what might "fit your bill". You seem to have a good handle on most things already.

 
Thanks!  Playing with everything helps Simplex Technology improve where we can and help with integrating whatever we're ask to.  

 

[bbrendon]

That's the downside  of open source  projects, they are generally poorly managed  from a quality control perspective.   You shouldn't assume they are secure, or tested.   There is no one to call when it breaks, you log your issues and wait or fix it yourself. 

The first two sentences apply equally to commercial software and is specifically why I left a commercial automation product and went to HASS.  The last sentence partially applies. If you're running commercial software and your bug isn't "big" enough, you get put off indefinitely and even if you could fix it yourself you can't.

These downsides you speak of sound almost like upsides