Amazon Echo to HA Controllers

It's still WAY too limited. No pre-fab 'skill/api' is going to address the broad level of functionality that a customized home automation solution is going to need. What they need to do is to allow the user to configure the Echo to a specific skill so that it doesn't require any extra verbiage. Then products like CQC could provide extensive control using the 'native' syntax. You wouldn't need the other stuff, since the automation system can provide that functionality. It would be completely optional of course, so you only set that up if you are using in conjunction with an automation system.
 
I posted a long rant on Reddit about this. Not to give love to competing sites, but here it is if you want to read the arguments for this approach:
 
https://www.reddit.com/r/homeautomation/comments/4dwx83/amazon_get_serious_about_supporting_automation/
 
Nice post Dean.
 
It is limited as you mention.
 
That said to those whom have never used VR/TTS it is their automation introduction (even though it is only a remote control) and newbies to automation know nothing else.
 
 
 
Here have lost track where I put the Amazon Echo (it is behind some piece of furniture). 
 
Looking forward to testing the Securifi Almond + Amazon Echo stuff.  Not so much for use but rather to tinker with.
 
I read some place that Google will be pursuing their competition to the Amazon Echo with the Nest stuff that they own.
 
I wonder if the Nest thermostat will just be a blinking eye that talks and listens?
 
Dean Roddey said:
It's still WAY too limited. No pre-fab 'skill/api' is going to address the broad level of functionality that a customized home automation solution is going to need. What they need to do is to allow the user to configure the Echo to a specific skill so that it doesn't require any extra verbiage. Then products like CQC could provide extensive control using the 'native' syntax. You wouldn't need the other stuff, since the automation system can provide that functionality. It would be completely optional of course, so you only set that up if you are using in conjunction with an automation system.
 
I posted a long rant on Reddit about this. Not to give love to competing sites, but here it is if you want to read the arguments for this approach:
 
https://www.reddit.com/r/homeautomation/comments/4dwx83/amazon_get_serious_about_supporting_automation/
 
Agree that they need much more native control syntax.  Amazon has stated that they're adding more, so hopefully that will come soon.
 
In the mean time, I've created virtual devices that have simple on/off control for all my AV components.  I can say, "Alexa, turn on Tivo" or "Alexa, turn on the outside amp".  I've also mapped volume into it by using the (virtual) dimming options of the devices.  I can say "Alexa, set Tivo to 60 percent" which in turn sets the volume to 60db on my Denon amp.  Yeah, it's a kludge, but until Amazon expands on the available syntax, it works.
 
They will never have enough syntax. If you look at what you can do with our Echo support, they will never match that, ever, because ours is completely open ended and let the user define the commands and how they are reacted to. But, our stuff remains saddled with the less natural syntax because of how the Echo works. 
 
The tech graveyard is littered with claims of "they will never"...

Really, if you've got a better mousetrap then pitch that.  
 
It's not a claim, it's just a fact. Consider how many possible commands or queries could be used by any given person. It's not infinite but it's enormous. There's just no way they are going to pre-build in that many possibilities because it would mean that they would have to compare every incoming spoken phrase against all of them for every single incoming command. And of course it would result in enormous ambiguity as well. So, even if they could, they wouldn't want to for practical purposes. 
 
And I've already pitched how to get around this issue, which was the whole point of the article I posted the link to above. All they have to do is provide the option to make a specific skill the only one active, and therefore it doesn't require the extra syntax. Then, if you have an automation system that supports the Echo, you can make its skill the active one. You would use it to do anything you want to do via the Echo. Or, alternatively, they could allow you to configure an invocation name for other stuff, so that those could still be used but require the extra syntax.
 
Some of the intent lists for skills have hundreds of entries.  It IS possible to be quite varied in the requests.  Yes, those are specific to the skill and that's behind the trigger phrase.  At some point it may well become possible to have more access to the first tier command structure.  Meanwhile, the task of forming an effective and robust set of intent commands is a task to undertake.  Especially when you start to factor in other languages...
 
Tangentially, it's fascinating how systems like Mathematica/Wolfram approach some of this.
 
But the thing is, a voice control system shouldn't have any fixed set of commands. It's really just a remote control, and it should be able to send any command or make any query. If you have an automation system. The automation system can be configured to respond to anything you want, in any syntax you want. In a fully customized system, that's the optimum scenario. Someone is going to come along with such a product, and it'll be so much more desirable for use with an automation system (as apposed to the Echo's trying to pretend like they are the automation system) that the Echo will likely lose out within the automation market, though of course it'll still do well within the market where it's just a fun gadget.
 
Given the ever growing DSP power available in small devices, whoever comes up with a small box, that has good recognition, works totally locally, and has a very easy integration capability, that will be what wins in the automation world. And Amazon, by dint of what they've done, has now made that a viable target, where before it probably wasn't something folks where thinking too much about investing in. Of course the Kinect is sort of that, but it's not easy to integrate with, being an extremely general purpose product.
 
Is Alexa not a case of "good enough"?
 
I agree it's not optimal to have the consumer keep track of what's a native skill and what's not. Should I say "Alexa, open the pod-bay door." or "Alexa, tell HAL to open the pod-day door."???
 
On the other hand, canned grammar ("Alexa, turn on the lights") seems "good enough" for wide-scale market acceptance and isn't hamstrung by the lack of custom grammar ("Alexa, the lights, switch'em on.")
 
I can see custom grammar being desirable (my house, it unnerstands me good) but (lack of it is) not a deal-breaker. I'd say Echo's appeal and acceptance indicates it's "good enough" for most people.
 
It'll be interesting to see if Google Home proves to be a worthy competitor.
http://www.theverge.com/2016/5/18/11688376/google-home-speaker-announced-virtual-assistant-io-2016

BTW, the end of the video posted on The Verge (timemark=5:25) shows a family interacting with Google Home (kind of like the video Amazon made for Echo). I don't know if that's "real" or "staged" but there's some very impressive natural-language abilities being demonstrated. It seems to go well beyond "basic grammar" it appears to understand context (listen carefully how one phrase refers to the prior one without mentioning the subject beyond using "it"). Real or marketing?
 
Interestingly, Google Home will not have an API to start, they want to control the experience. 

Funny, we started that way and are now about to open an plugin API for others to create native voice commands with the Kinect. Now that stick PCs are inexpensive and used Kinects are $30, and considering it all runs locally, we hope it will be a nice alternative for devs looking for local far field voice control. 
 
If you look at how some of our customers are using it, it demonstrates what is possible when there's not a pre-defined grammar, particularly in the area of queries. The issue isn't how many ways can you say turn the kitchen lights on. It's things way beyond lights and thermostats. There are so many commands or queries you might want to use. If you treat the Echo as what it is, a voice driven remote control, and let the smarts be in the automation system, then it's completely open ended. And they wouldn't have to give up the current scheme to do that, since it would just require allowing the user the option to say skill x is the default, and that one doesn't require the extra verbiage. That's allow it would take.
 
Then you can do anything you want, such as:
 
Alexa, is today's high temp greater than X
Alexa, how long till sunset
Alexa, what is the ground saturation percentage
Alexa, how many cars are in the garage
Alexa, what are the currently violated zones
Alexa, set the living room to party mode
Alexa,  Bob is currently in vacation mode
Alexa, which rooms have active motion
Alexa, how many inches of rain fell in the last x days
Alexa, have music follow first floor rooms if active
Alexa, have the sprinklers run today
Alexa, set all bedrooms to music playlist x
 
and so on and so on, whatever you can come up with that might want to do. All of those things can be done with CQC's Echo support, though none of them are built in. You can just define what you want. Though of course now they require the extra 'tell CQC to' or 'ask CQC if' and whatnot. Some of those you could probably kludge by pretending that something is a light, but you shouldn't need that sort of kludge.
 
And of course the above leaves aside lots of potential specialized applications beyond basic home automation, i.e. commercial applications where quite specialized commands and queries could be needed.
 
ano said:
Wow from the company that tracks my web browsing and reads my email, I can't wait to put one of these in my house so it can monitor my conversations!!
 
And has a habit of absorbing 3rd party products and abandoning development for them and leaving the users stranded.  No thanks, as if their being an advertising company was despicable enough.
 
At least with Amazon I know they're dedicated toward running a VERY large infrastructure of computational services and a product distribution company.  As yet they've managed to avoid being jerks in those capacities.  More than can be said for other players.
 
Back
Top