Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

EV Car/My Tesla - An Alexa skill for Amazon Echo / Tesla integration

This site may earn commission on affiliate links.
Great, this skill is jinxed! After a few hours of debugging trying to isolate the problem, I was going to post that the problem seemed to be on Tesla's end. I should have just checked the API thread first. Model S REST API

Tesla is apparently blocking all incoming traffic from Amazon/AWS. This of-course had to happen on the very day this skill got reinstated :mad:

Great and I just enabled the skill but haven't had a chance to try it since i did it away from home.

So what's the plan now? Can we run a proxy on our own pcs at home so that tesla doesn't know it's AWS?
 
OK, I've created a proxy to bounce the tesla APIs off of another server I own. I've used it a bunch today and it seems to be working reliable albeit with a little more latency. You can go to the Alexa app and go through the account linking now. If you previously enabled the skill, it should just work.
 
Update: Looks like its been working fine since yesterday.

For anyone interested, here are a list of changes that have gone in since the first release (release notes):

2017-03-05 Ver-1-0
==================
Added a proxy as a workaround for AWS blacklisting
Bug Fixes

2017-02-15 Ver-0-9
==================
Added support for "what is the charge limit"
Added support for "set charge limit to <85>"
Added support for "is my car locked"
Removed lock/unlock intents (as required by Amazon)
Location data is more specific and reliable.
Handle location-not-found gracefully

2017-02-02 Ver-0-8
==================
Removed Intent names from card title (request by certification team)
Identify "charge completed" scenario (instead of just saying "not charging")

2017-01-30 Ver-0-7
==================
Added variations for honk ("honk the horn").
Spoken range numbers are now more specific.
Added additional support for climate start/stop:
- start the A.C.
- stop the heater

2017-01-26 Ver-0-6
==================
Added support for charging start/stop
In some instances, conditioner start/stop would return errors.
Graceful handling of some error conditions.
Fixed "Ask EV Car for location"
 
  • Informative
Reactions: stevo1100
And finally, our other Alexa Skill called "My Tesla" has been approved by Amazon. This is for people who were not very thrilled with having to use the "EV Car" invocation name. With the new skill, the invocation name is "my car". So the interaction would be like:

Alexa, ask my car to turn the heater on
Alexa ask my car how far can I drive

To enable this skill, look for the "My Tesla (Unofficial)" skill. Here is a direct link:
https://www.amazon.com/dp/B01N9Y4I1E

For now both "EV Car" and "My Tesla" have identical functionality and you have both enabled at the same time if you want. Going forward, EV Car would be expanded to add support for other vehicles.
 
And finally, our other Alexa Skill called "My Tesla" has been approved by Amazon. This is for people who were not very thrilled with having to use the "EV Car" invocation name. With the new skill, the invocation name is "my car". So the interaction would be like:

Alexa, ask my car to turn the heater on
Alexa ask my car how far can I drive

To enable this skill, look for the "My Tesla (Unofficial)" skill. Here is a direct link:
https://www.amazon.com/dp/B01N9Y4I1E

For now both "EV Car" and "My Tesla" have identical functionality and you have both enabled at the same time if you want. Going forward, EV Car would be expanded to add support for other vehicles.


No problem if too soon to ask: excited about multiple car support when available (as we have a Model S and Model X and really want to use this on the Model X).
 
I am getting the following error message the first time I ask Alexa for car updates:

"There was an error connectivity to car: Code 408"

If I ask for updates soon after the first error is works ok.

I am using Alexa in FireTV so I don't know if this is causing the issue. I just purchased an Alexa Echo and it will arrive on Saturday so I will test it with the Alexa Echo.
 
I am getting the following error message the first time I ask Alexa for car updates:

"There was an error connectivity to car: Code 408"

If I ask for updates soon after the first error is works ok.

I am using Alexa in FireTV so I don't know if this is causing the issue. I just purchased an Alexa Echo and it will arrive on Saturday so I will test it with the Alexa Echo.

I don't think it is related to using the Fire TV. I just tried using a Fire TV and it seems to be fine. Looking at the logs, there were a few "408" and "500" errors about 9 hours ago. Since then a few hundred hits have all been fine. I suspect some intermittent issues with Tesla servers. If you consistently see issue please post again.
 
No problem if too soon to ask: excited about multiple car support when available (as we have a Model S and Model X and really want to use this on the Model X).

That is the next feature planned. The implementation is a little complicated though. One of the design goals is to not store any user information (token/email/password etc). That makes it a little difficult to store persistent user specific information. Some interaction will also have to to be built to be able to choose and switch between multiple cars.

I do have a few idea though and hope to make some headway this weekend.
 
I don't think it is related to using the Fire TV. I just tried using a Fire TV and it seems to be fine. Looking at the logs, there were a few "408" and "500" errors about 9 hours ago. Since then a few hundred hits have all been fine. I suspect some intermittent issues with Tesla servers. If you consistently see issue please post again.

I got an Alexa Echo on Saturday and ask for the car status and got another Error code 408 so it is definitely not specific to the Fire TV.

it is still a GREAT skill. Thanks for developing it for us.
 
Looks like Tesla is doing some kind of throttling on its API endpoints. Other API users have reported the same problem (see below). Thats not something we control so cant really do much about it.
Model S REST API

Part of the problem I believe is that there a number of apps that run in the background constantly checking for how "you and your friends are driving". They constantly hammer the Tesla servers irrespective of whether you use the app or not. That is just irresponsible design and will end up with Tesla throttling aggressively or just turning off public API access. the solution for Tesla is to make a formal program for giving out unique developer keys and then control or throttle based on individual developers.
 
Looks like Tesla is doing some kind of throttling on its API endpoints. Other API users have reported the same problem (see below). Thats not something we control so cant really do much about it.
Model S REST API

Part of the problem I believe is that there a number of apps that run in the background constantly checking for how "you and your friends are driving". They constantly hammer the Tesla servers irrespective of whether you use the app or not. That is just irresponsible design and will end up with Tesla throttling aggressively or just turning off public API access. the solution for Tesla is to make a formal program for giving out unique developer keys and then control or throttle based on individual developers.

i assume you mean Teslab. I have to say that it's one of the most PIA apps developed for the car.
 
i assume you mean Teslab. I have to say that it's one of the most PIA apps developed for the car.

Teslab does not ping the car unless you are in motion. It knows that from the phone's location information, not the cars. Teslab uses most of the location information from the phone, not from Tesla's servers so it puts a great deal LESS load on Tesla's servers than most other sites.
 
  • Informative
Reactions: davidc18