Desenvolvimento de um protótipo de sistema de alerta para pessoas com deficiência auditiva

This paper aims to simulate an alert system for people with hearing impairments, focusing on the creation of a proof of concept capable of identifying specific sounds and converting them into alerts. In this system, a computer program checks whether the sound signal has a dominant frequency within a defined range of interest and, when this condition is met, publishes a message to an alert topic on an MQTT server. This message is then consumed by an ESP32-C3 microcontroller, connected to a Wi-Fi network as a subscriber to that topic. Upon receiving the message, the microcontroller activates a vibration motor, notifying the user through a tactile alert. In addition, signal processing was carried out using the Fast Fourier Transform (FFT), employed to assist in identifying the dominant frequency of each sound signal. The proposal falls within the field of Assistive Technology and also explores principles of Wearable Computing, aiming to demonstrate how technological devices can be applied to promote accessibility and autonomy using low-cost resources. The tests showed that the system was able to correctly identify sound signals whose dominant frequency was within the defined range. The system also responded effectively to both synthetic audio and sounds representative of real-world environments. The results confirm the effectiveness of the solution as a proof of concept, with the developed structure potentially serving as the basis for future implementations, such as the construction of a wearable vibrating bracelet capable of delivering tactile alerts, in line with the perspective of Wearable Computing.