Desenvolvimento de um dispositivo para a coleta de dados de vibração em drones utilizando uma FPGA Kria Kr260, acelerômetros ADXL345 e envio via Bluetooth

Abstract

This work presents the development of an embedded device for vibration data acquisition in drones, focused on predictive maintenance and aligned with Industry 4.0 principles. The system was built using a Kria KR260 FPGA, leveraging its parallelism to integrate multiple MicroBlaze cores, each dedicated to reading ADXL345 accelerome ter sensors via the I²C protocol. The acquired data was stored, processed, and finally transmitted via UART to an HC-05 Bluetooth module, enabling wireless communication with external devices. The goal of the project is to enable efficient real-time vibration signal collection, allowing for the identification of mechanical faults through subsequent analysis. Experiments were conducted on a MARK4 drone operating under four distinct condition classes, including a healthy state and various simulated faults. A measurement range of ±16g was used, with a sampling rate of 400 Hz. The results demonstrated that the proposed solution is effective for embedded applications requiring low latency, high reliability, and scalability, serving as a viable tool for real-time diagnostic and monitoring systems.