One of the best things SmartThings ever did was the integration with Amazon echo. It’s fantastic. @tyler and I have talked about it a lot.
I also have the harmony home control and with the combination of Echo, smartthings, IFTTT, Harmony, and the extensive use of virtual switches in SmartThings, I now have granular voice control of my home entertainment system, something I never had before.
So I can say “Alexa, turn on Netflix,” or “Alexa, turn on ESPN” or “Alexa, rewind Roku” and everything works great. I’ve used a lot of voice technology since the wheelchair thing, including at big medical centers, and I haven’t seen anything like this.
SmartThings makes echo 100 times better because of the availability of virtual switches and the IFTTT channel. I use that every day, all day.
But in terms of unusual situations, the echo is dependent on the Internet for voice processing. That’s true of pretty much all good voice processing these days.
So if the Internet is out, I have to assume voice is not available either. Which is why I need the toggle switch for my Plan B.
It’s got the processing power to do it. We will get there.
Side note, I hope y’all appreciate the incredible focus we’ve put into security and privacy (ie with the new camera functionality). In some cases, those security implications and our dedication to being the most secure platform are making us have to be steady in our rollout big new features such as local processing when it comes to developer submitted apps and DTHs.
That would be great! I’m up for the video challenge. My daughter and I did our house on a hub to hub migration and it took 8+ hours doing only about 50 devices. The longest were excluding the zwave locks, even with the master codes, it didn’t work too well.
I would love to see how fast you are at this… Let me know and I can bring the cameras out and time you
Thanks for the reply @alex , and your engagement in the community. I have a lot of motion sensors that don’t need to turn lights on, but instead turn lights off after a period of time of no motion. That is missing, and really needed. Perhaps that’s coming and what you’re referring to, so I look forward to more from ST.
At our house, we ended up using a two device approach for presence, which is working really well. We ultimately added in ibeacons, which gave us customizable distance ranges. That was really great, because I could literally set the presence detection area to just the area right in front of the front door, for example.
But different things will work for different people. One of the things about Geofencing is it turns out that there’s a lot more variation in use cases than we tend to think at first.
You quote 15 seconds to exclude a Z-Wave device and 15 seconds to include, which is technically correct. However, switching the hub from exclude to include mode and back using mobile app takes another 15 seconds and another 15 seconds to enter device name (so you don’t end up with a bunch of identically named devices). So, its 60 seconds per device in the ideal case. In practice however, it takes much longer because:
Contact and motion sensors have to be taken apart and the battery has to be removed for at least 5 seconds to exclude and again to include.
To include/exclude some devices I had to drag the hub around the two story house.
Nothing ever goes smoothly. A few of my wall mounted switches just would not exclude and and I had to turn off breakers to disconnect them.
Even if everything goes smoothly, 200 devices in 3 hours (i.e. 54 seconds per device) sounds a bit unrealistic to me.
Wow! So the “local processing” is basically no more than a “theoretical possibility” at this point. In practice, everything will still run in the cloud, unless you have a trivial setup and don’t use any custom devices. Good to know at least.
Where have you been, this has been the case forever? Cloud performance comes and goes, and in an automated home it is painfully obvious when it’s not there as it should be, as you advertise it to be, as @Ben says it’s expected to be.
Do you know what happens when my wife walks into her closet, which has motion controlled lighting, and the lights don’t come on at all? Or, while she’s in there the lights go off when they shouldn’t. These are routine occurrences. I had hoped that Hub V2 would fix these types of problems with local processing, but it won’t if we can’t run custom device types and custom apps locally. The point is, the stock built-in apps don’t do the job for me, never have, and probably never will. So I have spent a lot of effort automating my home with custom device types and custom apps so that it will function the way I need it to. But, your cloud performance ruins the party. Now I’m just stuck.
3 Likes
tgauchat
(ActionTiles.com co-founder Terry @ActionTiles; GitHub: @cosmicpuppy)
154
I have little doubt of that, though it’s an easy statement to make when there is no time frame commitment included. As we have observed in several Hub V2 Topics of hundreds upon hundreds of posts, there are a significant number of Community members with zero patience and will leap down throats if any hinted dates are missed, while begging for the next release date…
SmartThings has about 100,000 current Customers and only a few thousand ever visit here. The next million or two Customers will have increasingly better experiences with the Product and Company.
Sadly, the most active and loyal here will have a long memory of the challenges we’ve been through and of the less than perfect record of SmartThings responding to ongoing problems and specific feature feedback.
In that light, I can only state what you already know: even the most harshly worded criticisms and frustrations voiced here are due to our love of the product and desire for it to optimally lead the market, with no personal slight intended.
It is unavoidable, however, that when patience is stretched too thin, bad experiences pile on, and upgrades fail to meet expectations, it becomes increasingly difficult to have faith in your company.
Respectfully, this erosion of trust is exemplified here by your incredible migration experience story; with several very skilled prolific Community members sharing credible contrary experiences and much more realistic and tedious expectations. It comes across as lack of empathy for the effort we’ve spent learning and discussing and tailoring our configurations to maximize value of the product to fit our families.
Well said. It also comes across as being completely misleading. Why @Alex chose to put out that email with this misleading claim is beyond me. Leaves the impression that he lives in a fantasy land where SmartThings works the way he wishes it did, while ignoring the reality that we all live with. @Alex, you should hang out with this community more often if you want to understand what’s actually going on with your customers and your company.
Now I really don’t have any reason to migrate to V2. I need local app execution of my apps, which use my device types. I can’t do what I do using only built-in ST apps. All of my lighting is on my own app, and my biggest automation failures involve lighting. Virtually all of those failures are cloud related, and I had hoped that by running locally, I’d get reliable performance.
I can’t think of any other feature of hub V2 that would be enough to make me go through a complete tear-down and rebuild. @alex was cutesy to say that he liked the fresh view and redesign of his apps and setup that tearing everything up almost forces on you. But in my case, I’ve done that already a number of times, and I have the design and implementation that I like. So for me, it would just be the drudgery of doing this migration for no real gain.
120 devices, 20 smart apps, 3 integrations, arduinos… I’m at about 40 hours currently, hue and harmony to go with a device or two left. This was with a minimote for zwave exclusions. Most of the automation configurations have been completed. Several hours were spent extracting app and device configurations.
Obviously I must be a challenged individual for a migration to take this long right?
So, since the deviceTypes have to be included in the firmware, can someone from ST @tyler@ben@slagle, etc. etc.
Provide a list of devicetypes that are allowed to run local? This would be a great thing to know as we all get our hub v2s in a matter of days, weeks, etc.
For example, will my Cree bulbs be supported?
How can we tell if a device will or won’t work locally?
2 Likes
tgauchat
(ActionTiles.com co-founder Terry @ActionTiles; GitHub: @cosmicpuppy)
160
Only time will tell, as more Community members share their migration stories.
At the moment, perhaps we’ve just seen the extremes of a statistically invalid sample…
All of us early adopters, tech types, who actually understand things, are probably the only ones who would be ambitious enough to undertake this project. And then shows up @Alex, being basically a Smart Alec, boasting about 200 devices in 3 hours, as if…
New name, new hub, new market, new customers. It’s interesting though that they are showing some compassion for existing customers, even though they are saying suck it up and move on.
