Copyright © 2018 Authors. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. International Journal of Engineering & Technology, 7 (3.14) (2018) 242-249 International Journal of Engineering & Technology Website: www.sciencepubco.com/index.php/IJET Research paper Develop and Implementation of Voice Recognition Robotic Car Hairol Nizam Mohd Shah 1* , Zalina Kamis 1 , Mohd Fairus Abdollah 1 , Mohd Shahrieel Mohd Aras 1 , Faizil Wasbari 2 , Nursabillilah Mohd Ali 1 , Clement Chia Kuan You 1 , Zairi Ismael Rizman 3 1 Center for Robotics and Industrial Automation, Faculty of Electrical Engineering, Universiti Teknikal Malaysia Melaka, Malaysia 2 Faculty of Mechanical Engineering, Universiti Teknikal Malaysia Melaka, Malaysia 3 Faculty of Electrical Engineering, Universiti Teknologi MARA, 23000 Dungun, Terengganu, Malaysia *Corresponding author E-mail: hnizam@utem.edu.my Abstract The idea in this paper is to develop a voice recognition system that can recognized five commands to control a robotic car. The focus area is mainly on voice identification and recognition system. The aim of the system was not recognizing sentences but only isolated a word then demonstrates the action on a simple built robotic car. The system allows user to deliver voice commands through a microphone for control the movement of the car. Voice command is sent to computer and the process to compare the signal with signal stored in database using Vector Quantization (VQ) technique. Mel-wrapping filter bank in feature extraction was used to reduce the root mean square am- plitude noise amplitude and also improve signal to noise ratio. Result showed that the robotic car can be controlled by 5 basic voice command which is stop, forward, reverse, turn left and turn right by integrating source code in MATLAB with Arduino UNO microcon- troller. Keywords: Voice Recognition; Vector Quantization; Arduino. 1. Introduction Nowadays, vehicles are very important in order to ease daily job and improve the quality of life. Most of the vehicles are not friendly for physically disabled or handicapped user. Besides that, some operation such as police, military, rescue operation need unmanned vehicle to do the job as the situation they face daily is dangerous and sometimes inaccessible by human [1-4]. Such job with high risk needs control in distance like voice control instead of hand control, so that job can be done without risking human life or limb. Living in this century full of development, world‟s economy, mil i- tary, healthcare, entertainment and transportation has been changed by the advanced technology which exists among all of us. With today technology, there are different ways to control appli- ances and devices without going near to the controlling button on the devices such as using remote control. One of the ways of con- trolling devices is by using voice recognition technology. When voice control is mentioned, speech recognition is the first word to be considered. The term "voice recognition" is used to refer as speech recognition where the recognition system is trained to a particular speaker, hence there is an element of speech recog- nition, which attempts to identify the person speaking or to recog- nize what is being said [5]. However, there are differences be- tween voice recognition and speech recognition. Voice recognition is a system relates to identifying voice of a particular user based on his or her unique vocal sound. On the other hand, speech recognition identifies almost anybody‟s spoken words in the cor- rect sense and then converting them into machine-readable lan- guage. In voice recognition system, although different recordings of the same words may include more or less the same sounds in the same order, the precise timing or the durations of each sub word within the word will not match. Therefore, the efforts to recognize words by matching the speech to pre-recorded speech templates will give inaccurate results because there is no temporal alignment. Besides that, noise that occurred in a sample of speech would affect the accuracy of recognizing a voice signal. As noise energy in a signal is more than the energy of a signal, the signal to noise ratio (SNR) is decreased. Once SNR is lower, the accuracy of recognizing words can be degraded. 2. Related Work Speech is a natural source of interface for human–machine com- munication, as well as being one of the most natural interfaces for human–human communication [6]. Speech recognition or voice recognition technology promises to change the interaction be- tween human and machines (robots, computers, etc.) in the future. This technology is still improving and scientists are still working hard to cope with the remaining limitation. Nowadays, this tech- nology has been introduced to many important areas. There are two categories of speech recognition, which are speaker dependent and speaker independent. Speaker dependent is a sys- tem that trained by the user who will use the system. This system only responds accurately to the user that trained the system. The advantage of speaker dependent system is that it can achieve high- er command count and better accuracy than speaker independent system. Meanwhile, system independent is a system that responds to a word regardless of who is the one that speaks. Due to this reason, the system needs to respond to different kind of speech patterns, inflection and enunciation‟s of the target word. Com- mand count for speaker independent system is usually lower than speaker dependent system, but the accuracy can be maintained within processing limits. Normally, in the field of industry, speak- er independent voice system is required compare to speaker de-