pepper-google-assistant

The basic objective of this project was to enable Google Assistant on the Pepper Robot. Because of a lack of root access on the robot, we had to find alternative ways to make this work such as using the notorious PYNAOqi framework which provided no actual guarantee of working at all times.

We had also developed a simple front end to either follow the interaction between the robot and any other person and also to synthesize speech using several Text-to-Speech API's.

What we have achieved

• Build a custom Google Assistant solution for the Pepper humanoid Robot.
• Train a simple voice trigger command on which the Pepper robot will start listening when its name ("pepper") has been said.
• Combine both these solutions in one solution.
• Build a front end for both the Google Assistant solution and the Text-to-Speech functionality.

Notes

Please refer to the my bachelor thesis Teach Robots to speak: Text 2 Speech Solutions for Robotics for more details about this project.