Is this the beginning of the end of my 5 years+ with ST?

General Discussion
1

I am approaching 6 years with SmartThings, but it is quite likely I will not get to 7. I feel let down, ignored and angry at a company that initially made many of my home automation dreams come true. My hunger for more was mostly in line with their promises, but at this point it seems that is all they are, or their timelines are simply insanely long. I am tired of waiting for the next big step (local processing, decent Android app, web browser control) from ST. This week I finally gave in and purchased a Hubitat Elevation hub and have begun analyzing the best approach to finally get rid of the slow and unreliable cloud!

Cloud:

It is slow! Way too slow! For example, many automations involving lights need to be nearly instantaneous. A second or two delay is simply unacceptable. How many times have you walked half way in a room and the motion-controlled light is still off? Even simply syncing two lights takes forever so when I turn on my entrance light, I am half way into the darkness of my hall before the lights start to turn on. The selection of local devices is too restrictive and often they only expose the bare minimum. What is the point of buying fancy automation gear if most features are unavailable if you want local performance?

While I can accept some automations not to work if the cloud, or internet connectivity, is down, my current experience is that the entire house stops working. It seems like a trivial issue until it happens and you realize how dependent and addicted to it you have become. Sometimes it is not a matter of pressing the light switch with your fingerā€¦ sometimes you rely on automation to cut off water before a freeze in a remote location, or more mundane things like lights and laser guide to park in my cramped garage fail to work when needed and I risk damaging the car or have to get out to ensure I am in the right spot. Why? We want and need LOCAL CONTROL, not the cloud BS.

App (Android in my case):

The old app was pretty badā€¦ clearly made for small numbers of devices and with an outdated look. It was very slow to load devices, and required many clicks to get stuff done. I got round this by relying on a dozen Echos around the house, and increasing the number of things that ā€˜just happenā€™ so that I would not have to open the app. Along comes the new appā€¦ and things simply got worse.

I was recently in a beta for the new app. I thought it was going to result in a 2.0 (from 1.7 or so that we are at) but to great dismay it felt like they were twiddling with insignificant features rather than addressing the big usability issues. At the end of the beta it felt like the steps forward were worth a 0.0.1 revision increase, maybe less. I get it, the most common ST user has 15 devices (rumor has it) so ST appears to only care about those users. I have over 200 objects (be they C2C, virtual or physical) and literally despise having to use the new app. The promised support for custom handlers needed to get access to all features of the ā€˜fancy automation gearā€™ has yet to materialize and usability for large numbers of devices has gotten significantly worse. Why?

I do not know whether I classify as a ā€˜power userā€™ but I definitely feel unwelcomed as most of the original power users I learnt from in the ST forums over the years. You know where I found the vast majority of them? Hubitat Community forums!

ST ā€“ why are you alienating those who can drive innovation and adoption? People who push the boundaries and develop new code they graciously share with those of us who are unable or unwilling to program? You should listen to them and keep them close!

Sorry for the lengthy soap box rant, but I am ready to throw in the sponge at ST and to move on. We will see how things go with Hubitat. It might be a full-on replacement, or maybe it will just take over all the devices/automations where I need to get rid of the cloud to benefit from speed and reliabilityā€¦ If ST ever sees the light, and offers more comprehensive LOCAL control, who knows, I might come back or stick around longer if not fully migrated yet.

10 Likes
2

Stay tuned for SDC 2019 to see what may be in store for ST :slight_smile:

5 Likes
3

Coming soon to a hub near you by summer 2050. The OP already said he is tired of waiting.

The reality is, that we, owners of 200 + devices, are purely freaks of nature that no company wants to deal with because we are too expensive to maintain. The guardrails and limits that everybody is putting in place are to stop us, and thatā€™s the direction everyone is heading to. :slight_smile: We are doomed perfectionists ā€¦ LOL

6 Likes
4

I share your theory. Hubitat appears to embrace us ā€˜freaks of natureā€™ so maybe that is where I need to be not looking back but the truth is I had high hopes in Samsung STā€¦ and really wanted it to work for me.

3 Likes
5

I wish you good luck. I am not ready to throw in the towel, just yet. It seems like everyone is struggling to keep our devices working properly, including the new kid on the block.

4 Likes
6

Unless their announcements come with well defined, short, and bullet proof timelines, they will only be more promises. I donā€™t doubt the great people at ST are ā€˜working on itā€™ but Iā€™d like to benefit from their promises before I become senile. Also, I have a super hard time believing that they will reduce their reliance on the cloud not having seem much of an effort to do otherwise. The only effort I saw was adding more DTHs to the hub so that they can run locally but they are barebones lacking even the most basic features offered by the device. I believe on some open/close sensors, the ā€˜local dthā€™ did not even report batteryā€¦ you needed a custom one to get that. Really???

7

Hey, Google is leading the herd to the ā€œlocalā€ waters. I think everyone will follow them into this ā€œnewā€ and ā€œingeniousā€ trend that no one else was able to imagine but Googlers[!], which they showcased at I/Oā€™19ā€¦

1 Like
8

@HA_fanatic - sounds promising. Hopefully in my lifetime :wink:

9

Iā€™m not quite at 200, I think Iā€™m somewhat close to 120 devices and everything I have runs locally. Ok phone presence doesnā€™t (and the number of times it has failed I could count on 2 hands over the 4 years on ST) and my Arlo cameras. While I agree that some of the early promises of being a very open and hackable system have long since been pushed aside, I dont think its totally trash just yet. Every major light in my home is automated. Every door and window sensed. Every room has a motion sensor in it. Only light switches I touch are my fans (something I just never got around to buying a switch for) and my bathroom lights (Not entirely sure I want them automated). But there has been hiccups dont get me wrong. But the closer you stay to a ā€˜stockā€™ experience the more reliable it is. Automating lights, water leaks, and SHM all work as designed and honestly Im dang near that 6+ months of MTBF that JD talks about.

Heck the other night I got in bed and hit my ā€˜good nightā€™ button and one light in the kitchen didnt turn off. That was the first time in about a year or so that I had to power cycle an osram led strip controller and boom it worked again.

The fact is this, the more you hack at the system, the less reliable it will be. In ANY automation system. Live on the fringe you will fall off the edge just a matter of fact. Sure SmartLighting and SHM has its limits, but Iā€™ve learned to live within them, or how to reverse logic into what I want. I may have a very basic setup here, but it does what i want. Manage all the lights, monitor doors, windows, and motion, and kick off camera recordings when needed.

Yes, us ā€˜power usersā€™ are not the focus of this system anymore, and havent been for easily 2+ years. But the system does what it was designed to doā€¦even if they changed the design about halfway through it.

I hope you find a system where you are happy. I went down the road of researching the alternatives and all sound 100% more work than what Iā€™m doing now. And for what? Yea having lights return to 10% after motion stop would be cool. But giving up local processing is not something I will do (ie webcore), and the reliability of the other systems seems markedly less than what Iā€™m receiving right now.

Iā€™m far from a fanboyā€¦theres times when I get frustrated when they make changesā€¦or that the classic app is going to go away and not be replaced for an equal. But this forum is just a loud speaker for all the negative voices. Thereā€™s honestly many of us with a large setup, that live within the walled garden, that are happy with the setup at least 90%ā€¦

7 Likes
10

Appleā€™s HomeKit has run locally except for voice control since its inception. :wink:

1 Like
11

Yeah, local processing was not ā€œdiscoveredā€ by the big players, not Apple nor Google, but it takes a leader to make it an industry standard.

1 Like
12

@KevinH - I may have given the impression but did not say that my ST system is trash. Iā€™ve been working on it for 5 years and have become very reliant on it. I love automation but always try to make it be complimentary to ā€˜normal useā€™ so that I am not stuck with the inability to turn something on or off due to how it was automated (ie people removing light switches and using smart bulbs only). I also do not automate stuff that would require humans to act like robots - meaning what triggers the automation must not require a change in my behavior. I am amazed you are able to run the majority of your devices locally but I suspect you either gave up accessing their advanced features or just purchased devices that do not have them. As an example, my Dimmers are Homeseer WD200+ and their special feature is to have 7 RGB LEDs that by default show dimming level (in several colors), but can also be used to provide 7 different notificationsā€¦ demo here:

With the stock DTH I can likely run the dimmer locally but I canā€™t use the LEDsā€¦ I love having those notifications. By using different LEDs, with different colors it was easy to memorize what each one means.

Even basic Aeon Door Sensors (the round one that fits inside the door) requires, or required (havenā€™t checked lately), a custom DTH just to get the battery level!!! When the battery is dead it is stuck on last status so what good is it, if you do not know whether the sensor is actually working?

I LOVE WEBCORE and given many of my devices are cloud based, I use it extensively. It appears that while it is not supported, it can be run locally on Hubitat so I get to use webCore and get local execution. See the appeal?

ST promised many times local execution - for years - but I have yet to see the fruit of their promises. I am simply tired of waiting and having to explain to family why things donā€™t always work. My 2 year old daughter, who barely said anything at the time, would notice issues (or rather us complaining about it) and say in the sweetest voice ever - ā€œno work? papaā€™ fix it!ā€.

Last - when I have lights mirror eachother, it works the vast majority of the timeā€¦ but how is the issue. If I walk into the house and turn on the entrance light with a dark hall ahead of me, I expect the system to mirror the hall lights with a very short delay, if any, so that I can walk into the hall with light. Right now it is marveling when it responds fast enough as most of the time it turns on when I am already half way through. So would you say this works? or not? My aim here is to eliminate the latency which defeats the purpose, in part, of the automation.

13

Not that this addresses the bulk of your complaints, but as another heavy android user, I side step the bulk of the app issues by barely using it. Sharptools is an amazing tool that lets you use android widgets right on your home screen, among other things (like a very very useful tasker plugin that Iā€™m heavily reliant on), most things Iā€™d need to flick on or off have a button right on my background, and tasker really makes your phone and the helpful data it has that st doesnā€™t a key part of the smart home.

3 Likes
14

@Stefan_Giudici1 - I did the same but by using the Echo to control what I need. I have about a dozen around the house so I donā€™t have to yell. I also employ automations to reduce my need for the app. I mostly need to go to the app to troubleshoot stuff, see statuses, and adjust the extra settings many of my devices have. If any of you own Fibaro dimmer products or recent Aeotec dimmer products you know what I am talking aboutā€¦ the list of adjustable parameters is long. Right now, the old app is the only practical way to adjust these parameters.

15

I will give you, I live alone therefore the WAF or anyone else isnt a concern. When I have visitors my automation is such that it doesnt bother them really, ie bathroom not automated or the guest bedroom can be in ā€˜guest modeā€™ easily. But nothing for me is a change in my behavior other than ā€˜dont touch a switchā€™ because every light is controlled via a door sensor or motion sensor and I dont have to do anything. I also dont like talking to my house, thus no voice control anywhere. Only button press I have is my ā€˜good nightā€™ cause I cant think of a better way really.

As for ā€˜accessing featuresā€™ or ā€˜buying devicesā€™. yea I was very thought out on what I boughtā€¦All Iris sensors. Seriously all my open/close sensors I got iris models while on sale at or below $15 a piece. They do open/close and temp, what more do I need. Motions, again almost all Iris, do have some ST motions that I bought near the start of my system that still work fairly well. Again motion with temp cant think of any other needs out of them. Both offer battery monitoring, also iris smart plugs offer power monitoring too. So not missing any features from those devices. I do have a couple ge toggle smart switches where smart bulbs cant go or had issues, and without changing whole house to decora style I dont want mix and match switches. Plus I have ā€˜upgradedā€™ switch plates at last count was like 500 dollars in switch plates Iā€™d have to changeā€¦I dont want mix and match nor do I want plastic covers. So changing those is not really viable.

My thoughts on the ledā€™s on the switch, do you really NEED to see the level of light? Or is that just a bit of ā€˜but i want it at 37% not 34%ā€™? Cause when I dim a light, i set it to whatever feels ok for current needs not a specific number. And honestly growing up without dimmers I rarely use a dimmer anywhere, just not what Im used to. So yea, self selecting on needs or in my case not needing it. Looking at stuff like the Inovelli switches look cool, but the morse code is just something I wouldnt use. The alert lights would be neat but again not something I cant work around, like leaving garage door open send an alert, not just light a light red.

Everything I have works as designed and full features with stock DTHā€™s and smartlighting automations. Nothing is limited at this point. Yes I didnt buy ā€˜fancyā€™ switches with tap codes and leds, cause for me, I want the lights to come on, and go off, and work via time of day or presence. Everyone is different for sure, and maybe ST isnt the system for you and your devices. But I do think thats a little bit of your selection not ST, at this point. Sure 4 years ago their promise was one thing, but they have dialed it back some. I would bet you could get your system to be very reliable, almost all locally executing, and within reason of your basic needs, if you tried. Sure you would lose some functionality, because those far ends of the fringe arenā€™t on STā€™s road map.

As you say you love webcore, thats awesome. I installed Rule Machine back in the day, just didnt need it and the complexity to turn on a light was way more than SmartLighting for my needs. Then CoRE came out and I tried it onceā€¦again more complicated and well it wasnt local. Now WebCoRE, again same thing for ME ā€¦ME. But I would warn you from reading about all the options, running WebCoRE locally on Hubitat is again going down another fringe road from what I understand. That it isnā€™t officially supported and when people have issues the first reply is ā€˜uninstall webcoreā€™. That may have changed in the last 3-6 months but last time I looked that was the stance. They lean heavily on Rule Machine since Bruce is on the team there.

We all know that we are not the normal customer. So anything beyond basic functionality really is outside of their control, design, and support. At least for the high level of reliability that one would want. So I would say at this point your wants dont line up to current ST expectations. And thats fine, but its not STā€™s fault. They have to cater to the 90th percentile. No way they can financially chase that last 10%. Yes ST has changed their philosophy on us from the start, but almost every company does.

My current biggest complaint is Hue popcorning. My kitchen motion turns on 7 lights, they popcorn one by one on. Used to not do that when I first moved to Hue, to get away from zigbee/cree bulb issues. It takes about 1.5-2 seconds for all the lights to turn on, but the first turn on is fastā€¦just takes a while for all of them to turn on. That is my biggest issue at this point, I think I must be doing something ā€˜right/wrongā€™ or abnormal since im a fairly heavy user with no issues.

16

I am too running the majority of my devices locally. It doesnā€™t require a change of behavior but rather a change in expectations. And by that I mean learning to say ā€œehā€ more often, if a light comes on at inappropriate time LOL. For those cases when is really annoying, thankfully they added the exception to run the instance only when a switch is on (or off). That addition to Smart Lighting made everything possible to run local #SmallThingsMatter LOL

1 Like
17

I am only asking them to actually follow through with their promisesā€¦ therefore my wants should be perfectly aligned. My hopes and desires may not be (ie make webCore part of ST platform now that Adrian Caramaliu works for them) and I donā€™t fault them if they never come true.

Iā€™ve read many times ā€œlocal is comingā€ and years go by. Those who formed Hubitat got fed up and got it done. Surely brilliant people, but I am pretty sure ST has similar talent.

I am also questioning why ST is alienating its power users, its influencers, in favor of the users with super basic home automation. We all started there so I am not saying they are not important but the category of people who push the boundaries, who influence others, who contribute awesome code to expand capabilities (like many of those who migrated to Hubitat) should not be ignored. I am not asking them to design their system around power users, simply not to ignore them. As an example, IMO their deafness to power users is evident in the design of the new ST app. Bit of triviaā€¦ I hid a ton of devices on the dashboard but it still takes me 23 flicks to get to the bottom of my device list. At least the old app showed me my 23 rooms (used as groups of devices & now limited to 20 in new app - GRR!) with 2 flicks, so on average I could get to a device with less arthritis inducing activity.

18

I am sure you know this, but itā€™s all about dollars and cents, the 80/20 rule, where 20% of customers suck up 80% of your profit, in our case is more like 99/1. We are the 1% that kills the business model LOL

4 Likes
19

Same. I had ST for 6 years. I switched to Hubitat. My only regret is that I waited so long to switch. I am very happy with my setup.

2 Likes
20

SmartThings hasnt made any ā€œpromisesā€ since early 2016. And management had 99% turnover since then. This isnā€™t the same company that made well-intentioned, but under-resourced promises.

SmartThings has either implemented (or implemented and cancelled) or made steady progress towards all announcements since then.

The upcoming SDC will showcase this progress.

6 Likes