Abstract

People with speech impairments experience communication barriers due to their limited ability to speak, while the general public generally has limited knowledge of sign language. This study aims to design and develop a prototype of a hand-sign alphabet translator based on the Indonesian Sign System (Sistem Isyarat Bahasa Indonesia/SIBI) as a communication aid between people with speech impairments and the general public. The research method employed was a design and development approach, consisting of hardware and software design, followed by overall system performance testing. The prototype is designed as a smart glove equipped with five Flex Sensors, one on each finger, to detect finger bending when forming alphabetic signs. The system uses two ESP32 microcontroller units connected via a Wi-Fi network. The transmitter unit processes sensor data and wirelessly transmits it to the receiver unit, where the translated results are displayed as text on a 16×4 LCD in real time. The test results show that the Flex Sensors have good measurement precision, with an average voltage reading error of 1.24% when the fingers are straight and 0.63% when the fingers are bent. The system successfully recognized 16 out of 26 SIBI alphabet letters, namely A, B, D, E, F, G, H, I, K, L, M, N, O, Q, R, and Y, with an accuracy rate of 61.53%. In the word-formation test using five words, namely ADA, HALO, MAKAN, BUKU, and BOLA, the system successfully displayed only the word ADA. The letters J and Z could not be recognized because they require dynamic movements that cannot be detected by the Flex Sensors. Overall, the prototype can be used as a basic communication aid for people with speech impairments; however, further development is required, particularly in detecting dynamic letters, forming words, adding correction and pause features, and improving battery life.

Keywords: Hand Sign Translator, Flex Sensor, Alphabet Letters, SIBI, Speech Impairment