Voice Control for openHAB

This is a simple but extensible voice control tool for openHAB. It is rule based and no data needs to be sent to any external API.

Examples will use the Main UI.

Usage and Setup

This voice control tool has been published as rule template for openHAB.

Further documentation on usage and basic setup is maintained here. Setup can be done manually as described below.

Manual Installation in openHAB

1. Copy Files and Customize

Copy the whole repositoy to the following folder in your openHAB installation: <OPENHAB_CONF>/automation/personal/voice-control-openhab

It can be any folder outside <OPENHAB_CONF>/automation/js, but in the following the above folder is assumed.

Optional: Add your custom rules to the end of voiceCommandRules.js.

In your shell, change the working directory to <OPENHAB_CONF>/automation/js and run the following command to install the package, such that it will be found by the JavaScript Scripting addon:

npm install --omit=dev ../personal/voice-control-openhab/

This will install the package into the node_modules subfolder. Note that only changes to these files will take effect after a reload, not the files in the copied repository at <OPENHAB_CONF>/automation/personal/voice-control-openhab ! So if you need to change anything, either change it directly in the node_modules subfolder, or in the copied repository and install the package again (bump a new version or use npm uninstall before; please refer to the official npm documentation).

2. Add Script Rule to Call the Interpreter

Add a new ECMAScript VoiceControl in Settings->Scripts with the following content:

(function (data) {
  let vc = require('voice-control-openhab').voiceCommandRules;
  vc.interpretUtterance(items.getItem("VoiceCommand").state);
})(this.event);

3. Add Rule triggered by Updates to VoiceCommand Item

Add a simple UI rule VoiceCommandUpdated, triggered by updates to the VoiceCommand Item that executes the created script rule VoiceControl:

Here's the YAML code if you prefer this:

triggers:
  - id: "1"
    configuration:
      itemName: VoiceCommand
    type: core.ItemStateUpdateTrigger
conditions: []
actions:
  - inputs: {}
    id: "2"
    configuration:
      considerConditions: true
      ruleUIDs:
        - VoiceControl
    type: core.RunRuleAction

4. Enjoy!

Done! Now you can use e.g. the openHAB Android app to send speech commands, either by the microphone button in the app (not visible in Main UI):

or by the openHAB speech command widget, which you can place on your home screen:

Development

After checking out this repository, run

1. npm install to install all dependencies and
2. npm run test to run all tests.

Before pushing changes to the repository, run npm run update-yaml to automatically update the rule template YAML file.

Note that for development, the openhab-js library is installed. It is only used for convenience in mocking/testing and can be omitted in openHAB, as it is available there anyway. Since it relies on an openHAB installation, running and testing the files in this repository is only possible when mocking the openhab-js library, otherwise you will encounter errors.

License

This tool is available under the terms of the MIT License.