Integration with Amazon Echo

Thanks did you upload the files as a zip when you created the lambda function ? What blueprint did you use to build the lambda ?

Thanks!!


Sent from my iPhone using Tapatalk

I used the Alexa skills kit color expert. You need the app ID to edit the index.js before you make the zip. I uploaded the zip when creating the lambada function. I think the arn you get is the same for all your lambada functions except for the name which is at the end of the arn line. Once you create a skill, that app ID will be the same no matter how you edit it. After I made the skill, I deleted the lambada function I made and then recreated one with the same name and uploaded the zip file to it. Now that I think about it, you could make the skill first because the app ID won't change. Use the ID to edit the index.js and zip. Create the lambada function and get the arn line which you can then copy and paste by editing your skill.

Thanks dproctor. All of this is explained by Thrag in the homeseer link I included in my very first post:
http://board.homeseer.com/showthread.php?p=1189626
I am working on some screen shot tuts for the github repo. Stay tuned.

Matt

Thanks Matt, I did not do a good job of reading the thread. I will follow the link and instructions and see how it goes.

Thanks again for the work on this.

Joe

Fantastic Matt
Screenshot tutorials, thats what I need to get this up and running. If you have a go fund me or something I would be happy to contribute to help offset the demand on your time.
Thanks, Matt

I've updated my Github project's README.md file with screen shots. It still needs a little love, but it is better than what was there before. #progress

https://github.com/msreynolds/askIndigo ... /README.md

If you have a go fund me or something I would be happy to contribute to help offset the demand on your time.

I've never really considered it, but the gratitude is appreciated. Mostly I do this for my own enjoyment, and for my own purposes. However, I do want to make the world a better place and I believe giving this technology away can do that. Imagine the possibilities this opens up for elderly and/or disabled people, for example.

[rant]
In addition, its 2015 people!! We have jetpacks, we have hover boards, we have speeder bikes... Every home should be automated, with natural language interfaces, every home should be self-sufficient with solar/wind/geothermal, etc.

Indeed, we had this very same capability 3 years ago via SiriProxy, but Apple quickly responded by improving security on the Siri services and breaking the SiriProxy project in the process. Now we're just starting to see the emergence of the same thing from Apple via HomeKit. Point being... granted these large companies are most often the creators of great technology (Siri), they also lock this technology up for their own needs and profits. Amazon seems to understand the importance of this technology (natural language interfaces) well beyond their own need for profit as a company. By giving the technology away (and still making money from their AWS tiers that make use of it) they are being good stewards of humanity with the power they've created.
[/rant]

I am considering making a nicer version of this script to publish in the store. Some features I am considering would include:

1) Prompt for all configuration info, store in Amazon DynamoDB (or where ever).

2) Prefetch devices.xml, actions.xml, et al, to get all names up front, store them in DynamoDB. Create a map of variances for different naming schemes. This would allow greater accuracy and less configuration when asking Indigo to perform. You would know that the request only gets made if the device exists and has the requested capability. Prevents a request being made to dim an undimmable device, etc. You could also get feedback from Alexa that the requested action would not be executed because of these things. Provides a better experience for natural language interfaces.

If I make this version a reality, would you be interested in purchasing a copy for a nominal fee?

Matt

Matt, I would be interested in the additional features you mentioned for a nominal fee. Hard work deserves to be rewarded.

I've read a bunch of the documentation on the Amazon developer portal and I haven't come across some of the concerns/wishes that I had. I ordered an Echo the other day and I'm eagerly awaiting it's arrival. Perhaps you might be able to help me find the answers?

1. I can't find any information on how long Amazon will let me run my application on their cloud. It seems that at some point they could shut it all down.

2. I am looking for a solution where I can control my home from all over the house. For me that would mean 5 Echoes placed throughout the house. Everything seems to be geared toward just one in the household. I would want them to be location aware. e.g. when I'm in the master bedroom and say turn on the lights, it would turn on the lights in that room. I would still want to be able to say a command that would completely identify an object in any part of the household, i.e. "turn off the living room lights" when I just got in bed and I remembered that I left the lights on before retiring.

3. Did you see anything in the documentation that would suggest that the Echo could identify one user from another? It would be nice if my wife said "play my music" it would play a default playlist that she likes and when I said the same thing, it would know who uttered the command and play my playlist.

That's all for now. Can't wait for my Echo to arrive and start playing with it.

Jack

I would be willing to fund purchase a good working solution for sure. In fact if I can get this working I would contribute. I am planning on working on this tonight and tomorrow and hoping I can have it working.

Have others gotten this working?

Joe

Made great progress tonight. Have it working as far as when I build everything it is successful. and when I ask Alexa to do something she said "ok" but nothing happens

Is there a way to look at what alexa or the skill is executing?

thanks for the work on this - my errors so far have been:

1) Needed file access modifications to build the zip file
2) the text editor I was using was putting bad characters in the files. ended up using ms word

Can someone let me know how to check on what gets executed when alexa says "ok"?

thanks
Joe

Matt,
I'd be willing to pay for that also. Thanks for your efforts. What I would like to see most would be full integration with indigo so that we didn't have to use "Alexa, ask indigo" first before giving the command.

Joe,
I had the text editor problem also. As far as testing, I checked the indigo event log and the settings/history in the echo app to make sure Alexa correctly heard what I said. I had problems with my device names in indigo having capital letters but Alexa translating lower case letters thereby causing no device found errors in the event log. When I changed the caps to lower case, all worked fine.

Dave

Ran into some more snags last night on configuring the test json. I kept getting syntax errors until I stripped out every thing but just one test. it worked fine so decided it was good enough as it was working and I could see stuff in my indigo log and the device responding. (thanks for the tip on no upper case letters - I have to rename all my devices as I have capitalized the first letter on almost everything)

Next will be to work on the utternces to make sure they make sense and are working the way I want.

for example I kept saying something like " alexa, ask indigo to turn master light off" this would not work. but "alexa, ask indigo to turn off master light" works

thanks Matt for doing all the heavy lifting and getting this going.

again I would be happy to pay something for this and continued features.

Joe

Hi Matt
I absolutely would pay for a solution to integrate echo and Indigo. The average Z-Wave device is $50, I think 3x the cost of a device is very reasonable to integrate natural language control.

Hi Matt,

I finally worked up the nerve to try this out. A little confusing for a non-developer newbie, but I did succeed in getting Hello World going and put together a test Indigo action.

However, the test case is throwing an error that I'm having trouble analyzing. The URL in the log is correct (I can execute it directly in the browser) but the process aborts before actually issuing it and Alexa says something about having trouble accessing the skill if I issue the command verbally. I get the same log report whether I issue the command verbally or through the test function.

Would you mind taking a look at this log to see if anything jumps out at you? (of course, I blocked out the server name) Thanks!

--Dave

Run Action test
2015-08-19T01:59:50.187Z f9a0072a-4615-11e5-8a2a-b135d3bcfbeb Making request: https://xxx.xxx.xxx/indigo/actions/test?_method=execute
/var/task/node_modules/request-digest/request-digest.js:39
var challenge = this._parseDigestResponse(res.caseless.dict['www-authentic
^
TypeError: Cannot read property 'caseless' of undefined
at handleResponse [as _handleResponse] (/var/task/node_modules/request-digest/request-digest.js:39:50)
at Request._callback (/var/task/node_modules/request-digest/request-digest.js:29:12)
at self.callback (/var/task/node_modules/request-digest/node_modules/request/request.js:354:22)
at Request.emit (events.js:95:17)
at Request.onRequestError (/var/task/node_modules/request-digest/node_modules/request/request.js:1011:8)
at ClientRequest.emit (events.js:95:17)
at CleartextStream.socketErrorListener (http.js:1552:9)
at CleartextStream.emit (events.js:95:17)
at SecurePair.<anonymous> (tls.js:1400:19)
at SecurePair.emit (events.js:92:17)
END RequestId: f9a0072a-4615-11e5-8a2a-b135d3bcfbeb
REPORT RequestId: f9a0072a-4615-11e5-8a2a-b135d3bcfbeb Duration: 3615.77 ms Billed Duration: 3700 ms Memory Size: 128 MB Max Memory Used: 18 MB
Process exited before completing request

Hi again, Matt.

I just had a breakthrough! On a hunch, I changed https to http and, sure enough, the request was processed and the URL hit my server. Unfortunately though it won't work since I have set up indigo to process https. I'll have to work on this but do you happen to know why this is not working and if there is any way to process https URLs? Thanks again.

--Dave

Is your https endpoint using a self-signed cert? AmazonWS might not like that...

Matt