Automatic speech recognition (ASR) has gained a lot of popularity in the mobile robotics, where the commands could be provided to the robot wirelessly to maneuver. A navigation system combined with ASR is a complex system to carry out, because the system has difficulty in recognizing the voice commands when the environment involved already has disturbances like road noise, air conditioner, music, and passengers. The objective of this research is to operate a mobile robot with a single-arm manipulator, where the robot can perceive the speech and it can react to the individual speech commands provided by the operator swiftly and precisely. In order to recognize the speech, mel-frequency cepstral coefficient (MFCC) speech recognition algorithm is chosen and implemented in MATLAB. Various training and testing have been done in MFCC algorithm where it has to carry out the real-time processing of speech data and respond to it. Based on both the training and testing the voice commands collected from the five test subjects both male and female, the speech recognition system achieved 89% efficiency for the test database. © 2021, Springer Nature Singapore Pte Ltd.