i am not sure if this is the proper way to suggest something, but since a long time i am not being able to control my vera devices through Alexa. it used to work before but not anymore (device not responding after a long time)
can i suggest having two layers of connection just like the Vera app, so when you issue a command, Alexa will try the internal IP first and communicate with it, rather than going to the cloud and start searching for the Vera server. that will be much faster and more efficient. if the internal IP is not responding then Alexa will try the cloud.
before Vera came up with the plug in, we were using the hub installed on a PC and it was so fast, now its not working anymore.
The Alexa hardware, is basically a complex microphone with DSP built in so it can recognize the trigger word. The command itself must be first sent to Amazons Servers to interpret what you want to be carried out, then sent back to your local network for execution by a local device (Vera, etc.).
I think you’re misinterpreting the question. Yes, the voice stuff has to take place in the cloud, but what the current skill is doing is talking to the servers at GetVera and then sending the command down to your local device over the persistent connection your Vera has with the Vera servers. With the HA Bridge (which mimics the Hue Bridge), the Echo processes the voice stuff in the cloud and then sends the command over the LAN to the Vera thus eliminating the Vera servers and speeding up the connection.
I turned on the Alexa skill yesterday and at the end of the day turned it off and reverted back to the HA Bridge due to the sluggishness.
I also, switched back to HA bridge. I think I’m going to invest in a raspberry pi to take the bridge of my PC. Was just running it on the PC til the skill developed, but it has been too inconsistent. Sometimes Alexa responds she can’t communicate with that device , then boom it comes on…
Exactly, that is my point… inconsistent, when i checked with the support they said your net is not very good although i can run vera remotly with no problem at all. Why do they have to go through vera servers…it is just slowing things… at least check locally first then go through the cloud.
As for the HA bridge, can it support thermostat and other things or still lights and scenes like before?
This clearly needs to happen. Alexa works seamlessly (and quickly) with other smarthome devices (such as the Tado theremostats) but with Vera it is sluggish, and I invariably get the timeout message (even though it actually works).
Presumably this relates to the (still rather inexplicable) timing for the request and confirmation to be routed out via the vera cloud, to the vera device, and all the way back. A possible kludge could be for the Vera cloud service to immediately provide an ‘ack’ to Alexa, without waiting for the confirmation from the Vera device (I can see if my light turned on or not, I don’t need Alexa to tell me even if the command does fail). Obviously better still communicate locally.
It isn’t clear that Vera acknowledge this is actually a critical issue with the implementation, however.