Alexa Announcements

ano

Senior Member
I have the voice board for my OmniPro II but I never got around to connecting them up to my house speakers. The problem is, I'm not sure I want to run a big Denon amp 24 hours a day.
 
So its plan B.  I have many Amazon Echo's in my house, so why not use them, but how?  I know Amazon has skills to announce things and stream things, but I'm not sure how to make use of them. I'm pretty open on how this will work. Maybe feed the text into Alexa for it to read it, or announce it using the OmniPro II voice card somehow. I just want it to announce on ALL the Amazon Echos in the house.
 
I have one idea.  The OmniPro II has custom phrases, so what if I recorded one that said "Alexa Announce" I'd follow this with the message to announce, like "Front Door Open" and play this into an Echo Dot.  The Echo would repeat the announcement through the house.
 
CONS are: Its pretty slow, and takes Alexa a few seconds to repeat it. Also, since I'm speaking with a speaker connected to the OmniPro II sound card and the Echo Dot listens through a mic, audio quality is not the best. Also, if you have two messages in succession, the second one will likely be missed.
 
PROS are: It should work, and the Echo even adds some sound effects to your message based on content.
 
Does anyone have any other suggestions? 
 
People on the Homeseer forums have been trying to figure this out as well.  The best I've seen so far (just glancing over that forum's new posts every now and then) is someone who put a speaker in a box with an Echo dot (speaker used for announcements from HomeSeer). Not a very eloquent solution but I guess it works.
 
BraveSirRobbin said:
People on the Homeseer forums have been trying to figure this out as well.  The best I've seen so far (just glancing over that forum's new posts every now and then) is someone who put a speaker in a box with an Echo dot (speaker used for announcements from HomeSeer). Not a very eloquent solution but I guess it works.
Yeah, I could use the OmniPro II to do that, but slow and certainly not elegant.
 
Not sure what is not elegant about it. Acoustic coupling is a valid interface with high reliability and good future proofing since it will always work regardless of feature changes within the Echo platform. It is easily expandable to other platforms like Google home and modifying events is as simple as editing the text string that is spoken. I'm not sure I would use a different mechanism even if there were others available. This is simple and reliable and for me that is the definition of elegant.
 
I am currently exploring Alexa home automation integration... unfortunately the clarity of information I've found so far has been lacking. It is my understanding that Alexa does not allow/support TTS to be sent to Alexa to be spoken. As previously mentioned, the current solution to trigger Alexa automatically requires audio detection by Alexa. I am using HomeSeer (HS) and have an event that performs several tasks involving Alexa. For example, I have a morning routine that has Alexa report the weather then news and then traffic conditions. I created various TTS scripts that HS reads out loud that trigger Alexa; I incorporated a delay between the execution of each TTS script in the HS event to allow Alexa time to reply. All I do to activate the event is say, "Alexa, start morning routine.". One thing I discovered was that in the TTS scripts, if I had something like "Alexa, get the news." The Alexa device would not respond; but if I changed it to "A lexa, get the news." it would work.
 
Just discovered this....
https://forums.homeseer.com/forum/homeseer-products-services/cloud-service-integrations/alexa-smart-home-skill/1260770-alexa-tts-that-works-well-same-solution-as-for-home-assistant
 
upstatemike said:
Not sure what is not elegant about it. Acoustic coupling is a valid interface with high reliability and good future proofing since it will always work regardless of feature changes within the Echo platform. It is easily expandable to other platforms like Google home and modifying events is as simple as editing the text string that is spoken. I'm not sure I would use a different mechanism even if there were others available. This is simple and reliable and for me that is the definition of elegant.
The biggest problem is the time lag. There is slight lag in the Omni, then it has to say "Alexa announce front door open."  Alexa has to record the full phrase, then tell all the speaks to repeat it. This takes some time.  The second problem is when a second announcement occurs during this time. The Omni may say "Alexa announce back door open" but if Alexa is speaking the first phrase, or anything else for that matter, the phrase is missed.
 
One of the guys that figured out the python / cookies thing with Amazon Echo has posted an alexa.api that I run in docker for more functions for Home Assistant.
 
Still tinkering as I do not like the cookie thing.  It is what they are playing with relating to Homeseer. 

I have not noticed much of lag using the python script for Alexa TTS.
 
There is an official API to do announcements, but it was limited and they aren't currently taking any more participants, AFAIK. I would hope that's because they think they have it figured out and are getting ready for the actual official release of it.
 
Alexa app Routines can do prompted vocal scripts through Alexa appliances.

The ISY portal supports this by telling alexa its ISY994 devices are motion sensors or other Insteon mechanical position detectors.
These devices can then trigger Alexa Routines to say whatever you program. ISY can interpret any variable value into whatever device you want to trigger those vocal responses.

Sent from my SM-G930W8 using Tapatalk
 
OK I have got it to work. At least for me.
 
First I should note I'm using SmartThings along with the add-on that allows SmartThings to mirror all your Omni zones and units, OmniLinkBridge.
This software is discussed here: http://cocoontech.com/forums/topic/31314-omnilinkbridge-to-integrate-home-assistant-smartthings-node-red/
 
So once your SmartThings knows the state of your doors and windows, you need to add two other things; Echo Speaks, and BigTalker2, both free smartapps for SmartThings.
 
Echo Speaks can be got here:
https://community.smartthings.com/t/release-echo-speaks/142252
It only was released about a month ago and its a still bit buggy but works.  Echo Speaks allows you to control Alexa speaking. 
 
The second SmartApp is one called BigTalker2 which allows you to trigger a voice response when a contact or smoke alarm or flood sensor is tripped.
BigTalker2 can be found on Github here: https://github.com/rayzurbock/BigTalker2
 
I am having a problem controlling the volume of announcements and multi-device groups don't seem to work yet, but for the most part, it works.
 
Let us know if it works for you.
 
If you installed Echo Speaks before Jan 15th, the author just did a major update. It seems to work much better.
 
The SmartApp has the ability to make a multi-room device, if you have multi-room audio devices in Alexa, but this is the one function I haven't been able to work.  Instead just speak to as many Alexa devices at once as you like. They are not always in perfect sync., but pretty close.
 
The latest versions I have are:  App: 2.2.0, Device: 2.1.2, Server: 2.2.0

Update: Device handler has been updated to 2.2.0 This work MUCH better than the previous on.
 
Here it has helped using an Oracle VB Android (latest version) build to do this stuff on my desktop while concurrently logged in to the Samsung Smartthings IDE.
 
Found a nice tool to install many applications called Community Installer. 
 
It does everything for you via the Android GUI and it is easy when the Android gui is computer monitor sized verses tablet or phone size.
 
It is majorly awkward to use the Android application and would be beneficial if Samsung created something along the lines of the IDE to do this sort of stuff.
 
communityinstaller.jpg
 
 
 
Back
Top