DEPRECATED!

This module is not supported anymore. The legacy dependencies are too old to maintain.Thank you so far.

MagicMirror Module: MMM-Assistant

A Voice Commander and Google-Assistant for MagicMirror

STATUS:	Version	Date	Maintained?
Working	`1.0.3`	2018-03-30	YES

What is this module doing?

MMM-Assistant is a MagicMirror (MM) module for using your voice to:

control your other MM modules
provide Google Assistant functionality
provide Alexa functionality [WIP]

Screenshots

There is not much to see since this is a voice based module. But there is a small microphone icon in the modules position. The icon changes depending on its status, so if you use the wake workds, it will change and also indicate this change by playing a "ding". The 2 available wakewords are: smart-mirror (for MM control) and snowboy (for GA interaction). The following icons will appear, depending on the operating mode it's in:

Normal `Listening`	Magic Mirror `Control`	`Google Assistant`	`Amazon Alexa`	`Error`

Installation

There are 5 parts to the installation:

Make sure your RPi OS is not too old...
Update nodejs and npm
Install the dependencies
Install the Assistant module
Setup your GA API account
Post Installation
~~Configuration~~

(0) To Update your OS

You should always use the full (not Lite) version of the Raspbian OS, if you are running on a Raspberry Pi. The current Raspbian OS version is: Stretch (9.4). You can find your version with:

uname -a
sudo lsb_release -a

If you need to update, use:

sudo apt-get update 
sudo apt-get upgrade

(1) To Update nodejs & npm

Before anything, make sure you are using the latest stable release of nodejs and npm. They are currently:

# Check the nodejs and npm versions
node -v && npm -v

v9.10.1
5.8.0

If they are very different from this, then update using these instructions.

(2) To Install the Depdendencies

sudo apt-get install libmagic-dev libatlas-base-dev libasound2-dev sox libsox-fmt-all libttspico-utils

(3) To Install the Assistant Module

cd ~/MagicMirror/modules
git clone https://github.com/eouia/MMM-Assistant.git
cd MMM-Assistant
date; time npm install         # ~8 min, ~705 packages

(4) Setting up the API Accounts

WIP

🔴 Now go to the Wiki page and follow all the steps in step 3 to create the Google Assistant and Google Cloud Speech accounts and settings. When done, proceed to the post installation below.

(5) Post Installation

cd ~/MagicMirror/modules/MMM-Assistant/
npm install --save-dev electron-rebuild             # ~65 sec
# Do not use the deprecated: --pre-gyp-fix
date; time ./node_modules/.bin/electron-rebuild     # ~25-42 min

The last step will take a very long time, as it need to manually compile the gRPC dependency.
⚠️ This take at least ~25 minutes on a RPi3, with little or no output. Do not interrupt! ⚠️

Module Configuration

To configure the Assistant, you must do the following:

Obtain the credentials (as already described above) and add them in the files:
- ./assets/google-client-secret.json
- ./assets/google-access-tokens.json
- ./assets/google-private-key.json (The OAuth2 certificate file you downloaded and saved.)
Add the Module to the global MM config.js by copying the content of the file: ./assets/config.txt.
Edit the following fields in the config file:
- YOUR_PROJECT_ID
- ~~YOUR_DOWNLOADED_PRIVATE_KEY.json~~

Add the module to the MM modules array in the ~/MagicMirror/config/config.js file by adding the following section. Here you also need to edit and insert the results you obtained from the Google authorization steps above, into the fields marked: YOUR_PROJECT_ID ~~and YOUR_DOWNLOADED_PRIVATE_KEY.json (a filename)~~.

{
    module: 'MMM-Assistant',
    position: 'bottom_left',
    config: {
        assistant: {
            auth: {
                keyFilePath:     "assets/google-client-secret.json", // REQUIRED (Google Assistant API) -- OAuth2 x509 cert
                savedTokensPath: "assets/google-access-tokens.json"  // REQUIRED (Google Assitant API) -- accesss_token & refresh_token
            },
            conversation: {
                lang: 'en-US',
                audio: {
                    encodingIn: "LINEAR16",  // Default. No need to change.
                    sampleRateOut: 16000     // Default. No need to change.
                }
            },
        },
        snowboy: {
            models: [
                {
                    hotwords : "MIRROR"                           // Default model: "MIRROR". (This is not the wake word!)
                    file: "resources/u-models/smart_mirror.umdl", // This file define your MM wake word. (See doc notes.)
                    sensitivity: 0.5,                             // 0.5
                },
                {
                    hotwords : "ASSISTANT"                        // Default model: "ASSISTANT". (This is not the wake word!)
                    file: "resources/u-models/snowboy.umdl",      // This file define your GA wake word. (See doc notes.)
                    sensitivity: 0.5,                             // 0.5
                }
            ]
        },
        record: {
            threshold: 0,                 // Default. No need to change.
            verbose: false,               // Default: true  -- for checking recording status.
            recordProgram: 'rec',         // You can also use 'arecord' or 'sox', but we recommend 'rec'
            silence: 2.0                  // Default. No need to change.
        },
        stt: {
            auth: [{                                             // You can use multiple accounts to save money
                projectId:   'YOUR_PROJECT_ID',                  // REQUIRED (Google Voice API) -- project_id (also shown in key file)
                keyFilename: './assets/google-private-key.json'  // REQUIRED (Google Voice API) -- service_account / private_key
            }],
            request: {
                encoding: 'LINEAR16',     // Default. No need to change.
                sampleRateHertz: 16000,   // Default. No need to change.
                languageCode: 'en-US'     // [en-US]  To set the default GA speech request language.
                                          // See: https://cloud.google.com/speech/docs/languages
            },
        },
        speak: {
            useAlert: true,               // [true]  Enable this to show the understood text of your speech
            language: 'en-US',            // [en-US] To set the default GA speech reply language.
        },
        alias: [{                         // You can use aliases for difficult pronunciation or easy using.
                "help :command" : ["teach me :command", "what is :command"]
        }]
    }
}, // END

The main difference with this full configuration from the simple one, is that it uses the rec vox wrapper for recording, rather than arecord since it works better.
You can change this configuration later when you see that it works.

Available Voice Commands

Command	Who?	Description
`snowboy`	GA	The Google Assistant wake word
`smart-mirror`	MM	The Magic Mirror command mode wake word
-	-	-
help `command`	MM	Explain help about a specific `command`
list all commands	MM	List all available commands
list all modules	MM	List all available modules
hide all modules	MM	Hide all available modules
show all modules	MM	Show all available modules
say "something"	MM	Use voice to say "`something`"
reboot	MM	Reboot your RPi
shut down	MM	Shut down your RPi

Dependencies (Google API keys)

🔴 Unfortunately the Google Speech API is not free. But it does allow you 60 minutes for free per month. But every recognition request is considered to last a minimum 15 seconds. So exactly 240 requests would be available for free. Additional requests will then be charged $0.006 per request, if you have a paid accout. Thereafter, the requests will be blocked. However, if you create an account for the first time, you will be given about $300 USD in credit, to use within 1 year, for free.

Therefore it is very important that we get the Amazon Alexa service implementation going, as they are completely free!

Dependencies (node/npm)

This module depend on the following npm packages:

npm Package	Version	GitHub
@google-cloud/speech	1.3.0	github
google-assistant	0.2.2	github
grpc	1.10.0	here
~~node-aplay~~	1.0.3	Deprecated!
aplay	1.2.0	github
node-record-lpcm16	0.3.0	github
snowboy	1.2.0	github
speaker	0.4.0	github

These are also listed in the package.json file and should be installed automatically when using npm install. However, those may require other packages.

For detailed information, See the Wiki

Bugs and Warnings

Sometimes the GA hangs (issue #25) and the Error icon is shown froozen. If this happens,
just re-load the MM server page from a remote connection and it will reset.

ℹ️ For other bugs, issues, details and updates, please refer to the issue tracker.

Contribution

Feel free to post issues and PR's related to this module. For all other or general questions, please refer to the MagicMirror Forum.

Credits

Most grateful thanks to:

--- - for clarifying and fixing XXXX

License

A license to 💖!

eouia / MMM-Assistant