MagicMirror² v2.8.0 is available! For more information about this release, check out this topic.

New Project! I think!



  • Hi guys!

    I have this project in mind that I hope you can help me to realize!
    I am a beginner and I am slowly trying to understand how the MM works.
    I would like to command with the voice (Google Assistant) a light (or door closure, temperature, thermostat temperature, etc …).
    I would also like to display the status in the OpenHAB dashboard always on a page of the Magic Mirror, as in the following picture:

    0_1550050398629_dashboard.PNG

    Below a small drawing of the project:

    0_1550049525518_Cattura.PNG

    I believe this project can be implemented this way:

    1. With my voice, I ask Google Assistant (MMM-AssistantMK2) to turn on the kitchen light.
    2. The request is sent to the OpenHAB Server always installed on the same Raspberry where the MM is installed.
    3. OpenHAB via the WiFi network commands a Wireless SWITCH (Sonoff, Shelly, etc …) to turn on the light.
    4. The OpenHAB Dashboard on MM (MMM-Openhab) shows the state of light.

    Questions:
    Is it possible to do this?
    Am I leaving out some aspect?
    The modules I would like to use are correct?
    Has anyone already tried to create something like this?

    Thank you all.


  • Project Sponsor

    Well there is an MMM-OpenHab module already.
    You might find a way to use it or expand it.

    Regarding speech commands: It might be possible to use one of the local speech recognition modules here (Lucy?) to send commands to the Openhab module or even to your local Openhab server. You wouldn’t have to use google then…
    Just a thought and not very much thought through…



  • @lavolp3 said in New Project! I think!:

    Well there is an MMM-OpenHab module already.
    You might find a way to use it or expand it.

    Regarding speech commands: It might be possible to use one of the local speech recognition modules here (Lucy?) to send commands to the Openhab module or even to your local Openhab server. You wouldn’t have to use google then…
    Just a thought and not very much thought through…

    Hola @lavolp3!
    Thanks for the reply.

    Why google assistant is not good for launching commands? Is it difficult to implement the commands?

    What is “Lucy”? If I’m not wrong, it’s an “extension” of the MMM-Voice module…right?

    If I understand correctly, I can send the commands only to the OpenHAB server and not to the module. The MMM-OpenHAB module can only show me the status … right? On the MMM-OpenHAB module page there is written:
    The module displays all values from your Openhab sitemap. You can also change all values and switches (like lites, temperature and s.o.)



  • No other advice?
    Has anyone ever created something like this?

    Thanks a lot!



  • is it possible… certainly the command and execute side… I do not know about the data availability for the display side…

    I would be interested in something like that…

    re-reading your original post… you said display the openhab dashboard… does it look like that display you played out? if so, you could paint that in an iframe, without much work… See MMM-IFrame or MMM-IFrame-Ping modules.


  • Project Sponsor Module Developer

    Unless you already bought the light hardware etc … I would sugest what I4m currently using on my mirror:
    A philips Hue system , with light connected anywhere you want. You need a bridge connected to your network and any light you want. Once you have setup your philips hue system . You can add the MMM-HUE module to your mirror and control it with Mk2 assistant ( I use in addition mmm-hotword)

    That work perfectly to monitor your light system and give all possible philps hue command to execute to the assistant . I do prefer Google serve for speech recognition because it’s more universal and running on the cloud , so your speech is very well recognized . It’s very rare when the assistant don’t get what i’m saying , even with a porr mic and some background noise.

    Since I think you already have open hab system at home what has been suggested higher should work with the MMM-openhab modules and iframe module to render you actual “home status” .

    Ejay



  • @sdetweil said in New Project! I think!:

    is it possible… certainly the command and execute side… I do not know about the data availability for the display side…

    I would be interested in something like that…

    re-reading your original post… you said display the openhab dashboard… does it look like that display you played out? if so, you could paint that in an iframe, without much work… See MMM-IFrame or MMM-IFrame-Ping modules.

    I hope I understand your question…
    I seem to have understood that the “OpenHAB dashboard” can be reached through a simple link and shows the status of the various sensors / wireless switches installed at home (correct me if I’m wrong …).
    So it seemed useless to have to “build” a page in MM for viewing these sensors!
    In practice I do not need to install the MMM-Openhab module but to choose between MMM-IFrame or MMM-IFrame-Ping modules … right?
    What is the difference between MMM-IFrame and MMM-IFrame-Ping modules?

    Thanks

    @ejay-ibm said in New Project! I think!:

    I would sugest what I4m currently using on my mirror

    What is “…I4m…”?

    @ejay-ibm said in New Project! I think!:

    I do prefer Google serve for speech recognition because it’s more universal and running on the cloud , so your speech is very well recognized .

    Is it possible to send commands from “Google Assistant” to “OpenHAB Server”?
    GA -> OpenHAB Server -> OpenHAB Dashboard

    @ejay-ibm said in New Project! I think!:

    It’s very rare when the assistant don’t get what i’m saying , even with a porr mic and some background noise.

    Which microphone do you use?
    Because I also use the MMM-AssistantMK2 module, but I have problems with the recognition of the commands and I do not understand if it is the sensitivity of the microphone or the quality!



  • @egnos I use a Logitech webcam… using google speech reco it is very accurate… using the local voice reco, pocketSphinx) it is not very accurate.



  • @sdetweil said in New Project! I think!:

    … using google speech reco it is very accurate…

    Which module do you use?



  • @egnos that is on a different mirror project (github smart-mirror, modules here cannot be used there)…

    I don’t know of a general purpose voice module here (like MMM-Voice) that uses the google speech api