Speech recognition is a technique that allows a computer system to receive input in the form of a spoken word. The words are transformed into digital signals by changing sound waves into a group of numbers and then adjusted to certain codes and matched with a pattern stored in a device. The results of the identification of spoken words can be displayed in written form and can be read by technological devices. The tool in this study was designed to be able to help people with hearing impairment so they can participate in prayer in congregation in knowing the changing movements of prayer. This tool is designed with an Arduino Nano microcontroller board as a voice processing function received from speech recogniton. The results issued by this tool in the form of vibrations that will be directly felt by the user. Presentation of success will increase if the sound around the input is not too noisy. Presentation of success in the words "Allahu Akbar" reached 82%, "Sami'allahu Liman Hamidah" reached 90% and "Assalamualaikum Warahmatullahiwabarakatuh" reached 80%.