Jump to content

Amazon Echo and ISY


madmartian

Recommended Posts

My apologies if this has been asked and answered before. I just ordered echo to use with ISY994. I was reading instructions posted in the first post now. I am supposed to use g1of4's configuration app to get devices. But it seems that this is executable file. Can I use this tool from Mac? Thanks!

Link to comment

I have a Mac as well and asked the same question on page 15 of this post. Here is a Mac version of the configuration tool. It works perfectly.

 

https://github.com/k...dgeConfigurator

 

You will HAVE TO use the Chrome browser and open with security settings turned off via: 'open -a Google\ Chrome --args --disable-web-security'

 

With those two things done and running the jar file via terminal everything runs smoothly.

Link to comment

Also, I have failed to get this to run on my Windows 7 64bit laptop.  Got it to run on a pc at my brother's house, but not at my home on my PC.  I run the jar program, it shows success in running, but when I try to run mapper, I always get a response "Exception occured - send g1of4  System.Net.WebException:The remote server returned an error: (404) Not Found"

 

Any thoughts what is wrong?  I confirmed the addresses of my ISY and my PC that is running the JAR.  

Link to comment

I used these instructions for rpi

 

http://www.airedalez.net/?p=263

 

I never did get wired to work on the rpi  wireless works great

 

Did you use the version 2.1 jar file from your link, or did you use the older 1.3 from this thread?  

 

In either case, I am able to navigate to the /configurator.html page which indicates my jar program is running fine.  And I can talk to my isy at its IP address. But the mapper program just keeps returning the "Exception occured - send g1of4  System.Net.WebException:The remote server returned an error: (404) Not Found" error.  I am stumped.  Tried running mapper from several PCs with same result.  

 

I have attached the full error message.  It appears it might be an issue when getting programs for the ISY?

 

 

post-2740-0-42883200-1439855107_thumb.jpg

Link to comment

Well, I reverted to version 7 of mapper, and it now works.  Apparently there is some limitation to number, name, or type of programs that causes versions above 7 to crash on my system.  I have LOTS of programs, perhaps too many for the mapper to handle(?)

Link to comment

In trying to added nested program folder support I ended up introducing a bug where I end up trying to connect to a program URI that doesn't exist.  I will see if I can get an update of the mapper out shortly with a fix.  Sorry you've had trouble.

 

C

Link to comment

There was this ISY/Echo update from Michel in another thread that I thought should also be copied here, FYI.

 

 

Hello everyone,

 

Just a quick update on Amazon Echo:

We completed our first milestone of mapping Amazon Account to ISY Portal Account. This was a major hurdle. The rest is about converting voice-intents to REST calls on ISY and caching spoken words (if any).

 

Again, the only drawback is that everything has to go to the portal and, as you all may have noticed, we are not that big on portal and cloud computation as it simply adds an unnecessary point of failure especially since everything is already on the local area network. Alas - and at the moment - we have no choice is local Echo APIs are non-existent.

 

With kind regards,

Michel

Link to comment

In trying to added nested program folder support I ended up introducing a bug where I end up trying to connect to a program URI that doesn't exist.  I will see if I can get an update of the mapper out shortly with a fix.  Sorry you've had trouble.

 

C

 

I am SO grateful for your work on this program.  It really has sped up my ability to work with the Echo and ISY.  Thanks for your time and energy.  Great work!

Link to comment

I In what I am doing I do not use the Amazon web services. The only contact to amazon is from between Alexia and the Amazon system. All intents go to my "Proxy server" on my LAN The only issue I had was I didn't want to handle SSL but the stunnel solution took care of that very nicely eliminating the need for the proxy to understand anything about SSL. Stunnel submits certificates, I am using a self signed one as that is allowed if test mode is enabled when the skill is registered. 

 

I started playing with the Hue emulator and got that up and running in a couple of hours on a Win 7 systems. The emulator is currently handling 26 devices each with two URL's (one for on and one for off).  I am concerned if the "29" device limit that I heard about is a hard limit as I have many more devices I would like to add. This week I will move the emulator to an RPi model 2. I use tiny VNC with the RPi so I can sit comfortably in my office and deal with the Pi.  In this mode the Pi is running headless, no keyboard, no monitor just power and an Ethernet connection.

 

I have a question that perhaps someone can answer. Can multiple Echo's talk to a single Hue Emulator?  I have already put a Request into Amazon to pass some means of identifying the Echo e.g. serial number or MAC address, with each Intent that is passed to a skill.

Link to comment

 

I have a question that perhaps someone can answer. Can multiple Echo's talk to a single Hue Emulator?  I have already put a Request into Amazon to pass some means of identifying the Echo e.g. serial number or MAC address, with each Intent that is passed to a skill.

 

I have two Echos talking to the Hue Emulator.  It works fine.

Link to comment

...The emulator is currently handling 26 devices each with two URL's (one for on and one for off).  I am concerned if the "29" device limit that I heard about is a hard limit as I have many more devices I would like to add...

Some of us suspect this is caused by an Amazon-imposed 20-second timeout. If it isn't finished discovering all the devices on your emulator in 20 seconds, it fails and nothing is discovered. I contacted Echo support and they said they would note it as a feature request to be able to set the timeout in settings. It is not currently configurable. Discovery using an emulator runs much slower than with an actual Hue or WeMo bridge.

Link to comment

I started playing with the Hue emulator and got that up and running in a couple of hours on a Win 7 systems. The emulator is currently handling 26 devices each with two URL's (one for on and one for off).  I am concerned if the "29" device limit that I heard about is a hard limit as I have many more devices I would like to add. This week I will move the emulator to an RPi model 2. I use tiny VNC with the RPi so I can sit comfortably in my office and deal with the Pi.  In this mode the Pi is running headless, no keyboard, no monitor just power and an Ethernet connection.

 

 

There is indeed a hard limit of "around" 28 devices with the Hue emulator, but sometimes only 26 or 27 will be discovered by Echo. As madmartian noted, it appears related to the 20 second discovery time used by the Echo.

 

To get past the limit (I have over 50 devices), I first tried adding the WeMo emulator to my RPi2 and got up to around 32-42 devices total, but the WeMo emulator (running in Python) is even slower and discovery was pretty hit or miss.

 

Presently I have two RPi's running the Hue emulator (each with a different list of devices) and the Echo is consistently finding 46 devices so multiple Hue emulators do appear to be additive. I can get up to a maximum of 50-56 during discovery, but NOT reliably so I'm just sticking with the 46 for now.

Link to comment

I thought I would add this tidbit of information from the Echo support staff (for what it's worth):

 

 
Hello Barry, 
 
I hope you're enjoying the Echo! 
 
At this time, the Echo does not have a way to configure the amount of time that it discovers devices. You do not have to discover all of your devices during this 20-second window. You can set a few up then start the discovery process again and set up more. However, making the time to discover longer is a great idea for improvement! 
 
Thanks for your suggestions to make the amount of time the Echo spends discovering devices configurable, to use different wake words, and to send a unique ID from the Echo to a Skill that is created so the Skill will know which Echo is originating the request! Those features sound like great ideas! I particularly like the idea about adding ID's to the Skills; I think it would widen the abilities that can be created. I'll share your suggestions with our Amazon Echo development team for consideration. 
 
I forgot to add in my previous email that our Echo development team is planning to add more wake words in the future. 
 
If you have additional questions about Amazon Echo, visit our help pages at: 
 
 
You can also reach us by phone directly and toll-free at 1-877-375-9365 or by e-mail at echo-support@amazon.com. We're available from 3 a.m. to 10 p.m Pacific time, seven days a week. 
Link to comment

And now for my status.

 

I moved the Hue emulator to a Raspberry Pi model 2. The Pi sits on the wall in my "Server/electronics room" It has no keyboard and no monitor, ergo, it is headless. I run tiny VNC server on the Pi and tiny VNC viewer on my development desktop system. Tiny VNC like all of the VNC variants gives me full control over the computer running the server. It is as if the Pi was running with a monitor and keyboard on a long cable. I am going to renter my devices today. I thought I had started Samba on the Pi so I could just copy stuff to and from it but it is easy enough to reenter. I am not a Linux expert but I can generally figure things out.  I followed the instruction at the site posted in this thread, and if i could accurately type I would have been done in under 30 minutes.

 

Now if I talk directly to Alexia, as in "Alexia, turn off the office lights", the intent goes to the Echo Bridge Hub Emulator on the Pi. If I tell Alexia to tell Sarah to do something, as in Alexia, tell Sarah to start the SPA" the intent goes to my endpoint which is under development. This endpoint which I call the "ISY Proxy server" is an app written in VB6 running on a windows machine it listens on port 9099 (my choice) for traffic forwarded from the sTunnel app. The sTunnel app handles all SSL issues since AWS requires all local (LAN) endpoints to use https and not http.  The ISY proxy gets intents from Alexia based as what I designed as my intents and utterances.   It will be more flexible than the HUE emulator.

 

At the current time interface between AWS and the ISY proxy is working. I get IntentRequests, LaunchRequests and SessionEndedRequests. I can respond to these with replies to be spoken by Alexia as I want. I can also issue 'Cards" for the Echo app.  I have finished all the code to read the configuration (nodes, programs and scenes) into the proxy. For nodes I only add the Node to the proxy if it has a "Note" associated with it. This will allow me (hopefully) to keep all echo related data on the ISY994. I plan to make the description portion of the node's Note entry have additional information such as reply phrases for various conditions. The spoken word portion of the note will probably the speech fragment such as "Office Lights". If the spoken name is blank it will use the name stored in the node. The proxy subscribes to the ISY and gets all status feedback. I handle and monitor the heartbeat from the ISY and will resubscribe if the heartbeat is missing for 1 minute.   With the status feedback coming in all the time there is no polling required and I will be able to cause action and Alexia replies to be state dependent.   

 

So much code, so little time.

Link to comment

I have hit the 29 device limit with The HUE emulator.  I operate fine with 28 devices, but as soon as I put in the 29th and ask the Echo to discover devices, it discovers no devices.  So it is back to my ISY proxy.  I have been in discussion with Amazon Echo customer service who are escalating this, however I do not think it will be resolved quickly

Link to comment

I am in the process of moving my emulator from my Windows 7 pc over to a RPi2.  Can I just copy over the data folder to the RPi to avoid reconfiguring?  

Yep. I did the same thing when I switched to my RPi2. (Now using two Rpi2s to get past the 28 device limit, with different data files of course.) 

Link to comment

I just installed HUE emulator to control my ISY and it works fantastic. Thank so much to everybody who made it possible!

 

I have only one issue - how to make echo to forget the device. I use Echo iPhone App and I tried (1) "Forget" individual device; (2) "Forget all connected home settings". In both cases it seems I  successfully remove device(s) and I do not see them anymore in the App. I then restart Mac configurator and select new set of devices from ISY. Strangely, Echo reports discovering BOTH new devices and old devices that were selected previously.

 

How can I wipe it clean and remove everything I selected (and Echo discovered) before? Should I stop/start java script? Reboot my computer? Restart chrome running configurator?  Thanks much!

Link to comment

Use two RPi's to get past 28 devices.  Clever!  I need to do that.  Any idea how many "Hue's" the echo will accept?

 

Does anyone know of somewhere that states all the verbal utterances the echo will accept for the HUE. obviously "Turn off" and "Turn On". It seems to try and do something for "Shut". Any clues on what to say for dim/brightness control or is that not there yet?

Link to comment

Use two RPi's to get past 28 devices.  Clever!  I need to do that.  Any idea how many "Hue's" the echo will accept?

 

Does anyone know of somewhere that states all the verbal utterances the echo will accept for the HUE. obviously "Turn off" and "Turn On". It seems to try and do something for "Shut". Any clues on what to say for dim/brightness control or is that not there yet?

 

So far I've only tried two emulators at a time (Hue + Hue on 2 RPi's and Hue + WeMo on one RPi) but the echo consistently discovers both, though each has a limit: 26-28 devices for the Hue emulator and 10-14 for the WeMo emulator.

 

Barry, your setup sounds very impressive. I've been into DIY home automation since 1985 (custom stuff) but the last 3-5 years have really opened up the possibilities. I just finished replacing EVERY bulb in the house with LEDs which was a huge undertaking since we have lots of light fixtures with every imaginable bulb type.

 

Voice control is the most fun though. I started integrating voice about 2 years ago using Tasker + Autovoice + Google TTS and now all the lights, media rooms, and HVAC units are voice activated. The Echo finally adds in hands-free voice control. Not sure what's left to do 8)

Link to comment

Archived

This topic is now archived and is closed to further replies.


×
×
  • Create New...