Detecção em tempo real de danos causados pelos consumidores em placas de circuito impresso usando dispositivos móveis e detectores YOLO

Abstract

The identification of consumer-induced damage is essential for electronics manufacturers’ warranty programs. Consumer-Induced Damage (CID) refers to any damage caused by an unauthorized person, including the consumer. The product warranty does not cover these damages, avoiding expenses in the manufacturer’s revenue. The warranty process for consumer-induced damage is usually carried out manually by technically qualified individuals. However, this task demands a lot of attention to detail, can be time-consuming, and is susceptible to human errors. With this in mind, this work presents an object detection model for low-computational-cost devices using computer vision and deep learning methods with YOLO detectors embedded in mobile devices to identify consumer-induced damages on printed circuit boards (PCB). Sixteen experiments were conducted with four YOLO neural network architectures, and a mobile application for CID detection was successfully developed. The best model achieved a mAP@0.5 of 33.1% and an average of 5.7 FPS on real mobile devices.