Raspberry Pi Cardboard Robot TJBot With IBM Watson API

In another article we have shown easy example of using IBM Watson’s power of machine learning as a WordPress Plugin which can analyse the emotion or tone of post’s text. Instead of server, if we use a hardware, the thing becomes as smart object and basic example of Internet of Things (IoT). In this guide we point how to use Raspberry Pi like single board computer & IBM Watson API to create Raspberry Pi cardboard robot named TJBot. The original project is created by Maryam Ashoori at IBM Research and available on GitHub.

Raspberry Pi Cardboard Robot TJBot : List of Things To Buy

You will need the following things for various projects :

1. Raspberry Pi 3 Model B or Compatible Single Board Computer
2. USB Mini Microphone
3. USB speaker for Raspberry Pi for Raspberry Pi
4. Raspberry Pi camera
5. NeoPixel RGB LED
6. Female to female jumper wires
7. Any cheap Made in China Servo Motor like Tower Pro SG 5010 Servo Motor
8. Raspberry Pi 16GB NOOBS Card

IBM used Tower Pro SG 90 Servo Motor which costs $2.

NeoPixel RGB LED is Digital RGB Addressable LED of 8mm size, of through hole technology (not SMD), this one :

Sadly NeoPixel RGB LED costs very high – $5 and one piece not available. This is the library :

For the body of IBM TJBot, you can 3D print, order 3D print service or cut the robot from cardboard after printing.

Raspberry Pi Cardboard Robot TJBot : Resources to Build

Here are useful resources to create the thing and watch others projects :

Errors are common. For NeoPixel LED, you can follow this guide to test :

Basically the cardboard box aka robot’s body is not so important to you or me. We need the development. Initial setup can deliver you pain and total number of things needed is not cheap to many. IBM Watson’s service is cheapest among all.