Voice Command Solution - Discussion

Hey everyone,

I am sure many of you have heard of Ivee (helloivee.com). I personally don’t own one, but it did make me do some thinking. For them to achieve integration with SmartThings, it would appear they have to communicate with the SmartThings cloud. Is this correct? Have we heard of any way to do this officially as of yet?

Here is where I am going with this…: My dream would be to put 5+ Raspberry Pi’s in my house, with a good bi-directional mic, and a wifi dongle. Toss on a simple “always listening app” and basically build a voice recognition app using one of the many cloud-based voice recognition api’s.

I look at the products like Ivee, and in some ways I feel that they are too ambitious with trying to integrate into 100 solutions, they are trying to do too much with having a pretty display, solution that talks to you, or simplify it so that just anyone can buy one at Staples and have it setup in 5 min. But when you look at needing one in every room, and the fact that many of us don’t mind building a Pi, configuring Linux, dropping the software on it, providing basic account info in a config file, and placing them throughout the house—I think that would be a possible (and cheaper) solution for some of us.

I personally use to program in html4/htmlk5/.net/php, I would like to discuss this with the community and see what people think this might take, not to mention see who else might be interested in this kind of endeavor.

Thanks everyone, and lets see if this leads anywhere. :slight_smile:


Yes, you can communicate with the SmartThings cloud. The way you would do this is by creating your own SmartApp with a REST endpoint. You can then send HTTP requests to the app, which it can interpret and act on.

As for the voice recognition, I have no experience with that. So you are on your own, or might want to wait until someone else chimes in. I do recommend you take a look at Ubi, though (http://www.theubi.com/)

1 Like

You know, I forgot about Ubi, just because of how far behind schedule it is. Just like Ivee, very late in deliveries… Except Ivee is getting only complaints in its ability to detect voice, accuracy, etc. I would be fine with multiple Ubi’s in my house, once I know its compatibility to SmartThings, final cost (its not on the website), and obviously, how good its voice recognition is.

That being said, a custom built solution could bring HTML5 web control to SmartThings, etc…

If you Google for “tasker SmartThings” you’ll see that for under $100 you can pick up a cheap Android tablet and use tasker “autovoice” to connect to smartthings end points. Wall mount the tablet and you have a nice kiosk that can run SmartThings via the app, or by voice and have all the gooey Android apps too.

It might not be exactly what you’re looking for, but it’s still pretty cool.


I bought an Ivee and returned it- it was just terrible. It rarely recognized the “Hello Ivee” command, in a quiet room from right next to it, and when it did it rarely understood the next command, usually responding with something not even close to the question/command. I will keep an eye on it to see if they can improve it, and they are adding the promised integrations so ST will happen at some point, but has such a long way to go…

Even if it did work, I don’t think it is the best solution to have the mic and speakers in the same device. I like the idea of creating a whole house always on mic solution, such as Pi’s all over the place, and then using sonos, or similar, as the speakers. That way you can use your existing speakers which many already have throughout the house, and optimally place the mics and keep them minimal, perhaps just sticking out of the ceiling. That will create the aura of an actually virtual assistant/butler, and not just talking to a little box.

A wall mounted tablet or even phone is tempting as well, but doubt the mic’s would be good enough for a large room. I have to look way more into mic’s though, as my minimal experience with them is it is very hard to cleanly pick up voices from anywhere in a room.

I actually think the best place the a mic is in a watch. I have pebble and iOS environment so very limited, but has anyone messed with this with the galaxy gear? I don’t really want a speaker or even nice screen on a smart watch, but a mic would prove so so useful for home automation, especially if it auto activates when raised.

The Tasker idea was interesting. But part of an always listening (no button press) solution has my appeal. Having to pick up the phone to talk into it isnt very appealing as well. Ubi may provide an always listening solution, but so far that’s yet to be seen.

I’d be curious to see if anyone else has thoughts about this kind of thing. It would be cool to have a team of people help build a Pi-based solution. It has the power, a lean kernel (if chosen), so it would be quick start. Then its just a matter of getting a great mic and building a c app to integrate with a voice processing cloud, and send to SmartThings…

It is good to hear that there is a way to reach the SmartThings cloud though… That makes me happy.

(Note: I would be able to assist with the development, though I sadly don’t have enough time to run with this myself).

As a kickstarter backer, I’ve been collecting some of these devices. I’ve got Smartthings, Ubi, and ivee.

After setting up the example endpoint API in the smartthings cloud, I was able to configure (within Ubi’s cloud interface) custom “utterances” so I can tell the Ubi to turn my lights on/off.

ivee hasn’t released an API or smartthings integration yet, so for the moment it can’t integrate for me.

The Ubi works reasonably well though sometimes it’s phrase “Ok, Ubi” doesn’t trigger and sometimes is doesn’t understand me, there is at least a bit of a delay since it processes my voice to it’s cloud, ubi->ubi-cloud->smarthings-cloud->smartthings-hub->smarthings-outlet.

Ubi has a more mechanical voice then ivee, it sometimes feels like ivee understands me better, but it’s commandset is currently very ridged and limited.

Both devices are in very early stage development (software-wise) it seems at least. They don’t feel like mature products, I’m hoping that with new updates that this will improve.

I’m pretty sure you can use adnroid, tasker, autovoice and set an always listen mode without hitting a button. I had this working for awhile.


I’m going to try out the ivee it’s on sale for $25

I got an Ivee and it’s ok but I also like Joshua Lyon’s Sharptools app. Here’s a demo I posted on youtube: Home Automation Voice Control!: http://youtu.be/gKq0vCeoPzQ


I am curious to see if Ivee has better detection over Ubi, now that they switched to wit.ai. I’d be curious to see if it does better noise cancellation over the Ubi as well.

I am about to sell my Ubi. I’d be curious to get more Ivee updates in comparison, as Staples locally in Orlando is selling it for 29.99. Then again, that may indicate a new Ivee coming out soon, as I tend to think that’s a loss leader price.

Or it may indicate that Staples thinks some combination of Amazon’s Echo, Apple’s HomeKit, and Samsung’s smart watches is going to take over the marketspace, at least for the general consumer, and they’re waiting to see how it all shakes out. A retailer decision rather than a manufacturer one.

I was thinking about the ivee sleek it’s sold out at staples and not getting any more as far as I can tell by the follow up phone call I had with them about an hr ago. Best deal I seen was on eBay $54. I did email the company to see if there was another version coming out soon and never got a reply. They still sale them for $199 on there home page.

Ivee, from what I’ve seen, isn’t even worth what Staples is charging, let alone $200. I got one on Saturday and it’s only worked on one day since then. Even to ask it to tell you the time it needs to contact their servers and it hasn’t been able to do that reliably.

I am using https://wit.ai/ to process language for my project. The language processing is very good; if you have a decent microphone. The problem with Ubi and Ivee is not primarily the language recognition, but it’s the microphone. The best I have seen is the amazon echo or the xbox kinect. Getting good audio is the tricky part of voice control and it’s still a very unsolved problem in the places you would need it most.

1 Like

I could see voice control via Xbox One Kinect being hugely powerful w/ Cortana! Imagine having that voice control via Xbox One, Microsoft Band, or your phone (Cortana/Siri/Google). Something along the lines of what Easydom is going for: http://www.windowscentral.com/easydom-nexthome-full-home-automation-system-windows-81-and-windows-phone

Thanks for the heads-up on thie Sharptools app - its really great! It took about 4 minutes to download and get access to all of my ST devices. An hour later, I had Widgets and Tasker integration up and running.

I bought a couple of cheap 7" Prontotec tablets this week (They were on sale on Amazon for like $45 each and they work surprisingly well!). I was hoping to use them to replace 2 groups of 3 “dumb” switches that other family members just insist on using – and have been amazed at how much more control the Tasker / Autovoice / Sharptools provides (especially since I’ve been on iOS). Moreover, setting up voice controls are so easy.

If you consider the price (and the hassle) of buying and installing 6 switches to that of buying 2 tablets and $10 in apps, this is my new go-to. Plus working voice control in the bedroom (WAF!!!) and living room to boot.

The only question I have is: Why hasn’t Samsung (or another company with IoT intentions) bought Tasker?

1 Like

Hopefully, it’s because the developer is not interested in selling out. I am sure he is quite comfortable with the revenue it has generated over the years, and perhaps the project is still a labor of love.

Thank you Mr. literalinpollock!

The point I was trying to make is how amazing all that extra functionality is. Knowing that its not part of the system is like losing the piece of a puzzle. You can try to imagine you don’t see it and that it doesn’t matter – but it does…

Not everything should be part of “the system”. Acquisition, more often than not, leads to apathy. I prefer to see innovation remain in the hands of innovators.