ECE 5990: Final Project

Voice Controlled Car


Ashkan Ravani (

Xiaoyu Guo (

Vishnu Govindaraj (


The main objective of this project is to exploit the “IoT” potential of Raspberry Pi computer to control a Remote Car using the Voice Commands. The Raspberry Pi with a Microphone connected to it will receive the Voice Commands and will control the Remote Car by Processing the received Voice commands.


With voice recognition becoming more and more realistic and accurate with the help of cloud based voice recognition APIs, the applications based on voice commands are also getting realistic. The Raspberry Pi single board computer which is well known for its excellent embedded system portability, can be used as an intermediate processing unit which can easily connect the cloud based Voice recognition APIs with the the real time hardware like Cars, Lights and Rovers to come up with Voice controlled applications. This idea enabled us to build this project to control the Remote Car using the Voice Commands. This project provided us the opportunity to experience the techniques like FIFO(linux named pipes) and PWM which we learned as part of ECE 5990 course to interface three different technologies Google Voice API, Raspberry Pi Computer and Remote Car into one working model, “Voice Controlled Remote Car”. The Voice controlled Remote Car we built is able to recognize the voice commands but not as consistently as we expected. Though the Voice recognition softwares have improved over time, the perfect voice recognition is still not available. By training our program with training data for each of car’s control signal, we were able to achieve 75% consistency in controlling the Car.


The overall system design