Real-Time Industrial Defect Detection on Edge Hardware Using Fine-Tuned YOLOv8: A Systematic Benchmark on the NEU Surface Defect Database and MVTec AD with Automotive & Battery Manufacturing Extensions
Researchers have developed Industrial-YOLO, a framework using a fine-tuned YOLOv8 model for real-time defect detection on edge hardware. This system was benchmarked on the NEU surface defect database and MVTec AD, with added automotive manufacturing extensions. The framework achieves over 120 FPS on an NVIDIA Jetson Orin platform with a 98.5% mAP, demonstrating robust, zero-latency performance suitable for automated optical inspection systems. AI
IMPACT Enables high-speed, low-latency defect detection in manufacturing environments, potentially improving quality control and reducing costs.