TJBot [Beta]

Node.js library that abstracts several functions for TJBot.

Encapsulates basic functions for TJBot such as making your bot see, listen, speak, play sounds, etc.

Some of the functions exposed work with certain IBM Watson Cognitive Services. For example "seeing" is powered by IBM's Visual Recognition Api. Similarly, speaking and listening are powered the IBM Text to Speech and Speech to Text. To use each of these services, you will need to add a config.js file to your application and add your credentials for each of the services you are interested in using.

• tjbot.shine(color)
• tjbot.pulse(color, duration,delay)
• tjob.speak(message)
• tjbot.wave()
• tjbot.listen(callback)
• tjbot.converse(workspaceId, message, callback)
• tjbot.see(mode)

Installation

This module depends on your raspberry pi having an installation of several tools.

ALSA tools for sound recording.

  $ sudo apt-get update
  $ sudo apt-get upgrade
  $ sudo apt-get install alsa-base alsa-utils

FOR DEVELOPMENT PURPOSE: to develop using TJLib is possible to test your conversation recipe in Ubuntu. Just verify if alsa-base alsa-utils are installed and set configuration.microphone config with card/device found at arecord -l

After the above dependencies are installed, you can proceed to install the module using:

$ npm install tjbot

Next, please setup credentials for the IBM Watson Cognitive Services you will be using in your application. A template for your configuration file (config.js) can be found in the library root folder (config.default.js).

API

Below is an example of how to use the module.

var tjbot = require('tjbot');
var constants = require('./config'); // user configuration file (samples/config.default.js)

// obtain credentials from config.js, see samples/config.default.js for more info.
var credentials = constants.credentials;

// these are the hardware capabilities that our TJ needs for this example
var hardware = ['led', 'microphone', 'speaker'];

// turn on debug logging to the console
var config = constants.config ;

// instantiate the TJBot library!
var tj = new tjbot(hardware, config, credentials);

// Change LED color to red.
tj.shine("red");

tjbot (hardware, config, credentials)

Returns a tjbot object instance that can be used to control the various hardware add-ons (e.g. microphone, speaker, led, servo arm, etc) and exposes some additional helper functions.

• hardware - string array containing strings for the various hardware components you would like access to. Valid options are led, microphone, speaker, servo, camera

• config - JSON object containing configuration parameters. It could be set at config.js with default parameters and replaced at your TJ script.

  var config = {
      'attentionWord': 'TJ', // attention word for STT
      'ledPin': 8,    
      'servoPin': 7,
      'voice': 'en-US_MichaelVoice',
      'verboseLogging': false,  
      'ttsReconnect': true,    // reconnect to STT service on error
      'visionConfidenceThreshold': 0.5, // Confidence threshold for tags from visual recognition service. Tags below this will be ignored.
      'visionTextConfidenceThreshold': 0.1,
      'cameraParams': {
          height: 720,
          width: 960,
          vflip: true,   // vertical flip
          hflilp: true    // horizontal flip
      } //set camera params
  };
  

• credentials - JSON object containing credentials for various IBM Watson Cognitive services that can be integrated with TJBot.

tjbot.shine(color)

Changes the color of the LED to the specified color.

• color - Color which may be in string or hex format. E.g red, yellow #ffff

tjbot.pulse(color, duration, delay)

Continuously pulses a color until stopPulsing() is called.

• color - Color which may be in string or hex format. E.g red, yellow #ffff
• duration - integer representing pulse duration in seconds
• delay - integer representing delay in seconds

tjbot.speak(message)

Speaks a given message. This method returns a promise with a string which is the path to the audio file that is played.

• message - text to be spoken out

tjbot.wave()

Waves robot arm (up-down-up).

tjbot.listen(callback)

listens for utterances from the microphone, takes a callback function as an argument.

• callback - function that is called whenever a new transcript arrives.

tjbot.converse(workspaceId, message, callback)

Starts a conversation turn using the watson conversation api, takes a callback function as an argument.

• workspaceId - workspaceid, used to keep track of the conversation context.
• message - message sent to the conversation service.
• callback - function that is called whenever there is a response from the conversation service. This callback is called with a response object
  • response - JSON object {object: obj, description: string}
    • object - JSON object returned from the conversation api
    • description - String containing text returned from conversation service.

tjbot.see(mode)

Captures an image, and analyzes it using the IBM Watson visual recognition service.

• mode - String variable that specifies the type of service used to analyze image. Values are

  • classify - classify image across known tags
  • text - recognize text within the image

License

This library uses the [Apache License Version 2.0 software license] (LICENSE).