Document Type

Conference Proceeding

Publication Date

7-28-2026

Abstract

Mixing of microcontroller unit (MCU) integrated circuits (ICs) during the final testing stage of semiconductor manufacturing can lead to material waste, production delays, and customer dissatisfaction. This issue often occurs when standard JEDEC Matrix Trays (JMTs) are reused without confirming that all ICs have been removed after testing, a process typically performed through manual inspection and therefore susceptible to human error due to high test volumes, small IC package sizes, and visual similarity between IC packages and tray surfaces. This study develops an automated IC Detection Test System using embedded vision to determine whether JMT trays are empty prior to reuse. The system employs a TinyML-based Convolutional Neural Network (CNN) deployed on a low-power OpenMV Cam sensor board to classify tray pockets as either containing an MCU IC or empty background, while an MCU-based control unit automates tray positioning and communicates inspection results. Unlike conventional machine-vision inspection systems that require high-cost computing platforms, the proposed approach performs real-time IC detection directly on a resource-constrained embedded vision device. Experimental evaluation shows that the system achieves an accuracy of 97.33% in detecting IC presence on JMT trays. By automating tray inspection, the proposed system reduces reliance on manual checking, prevents IC mixing, and improves manufacturing reliability. From a business and operational perspective, the system reduces material waste, minimizes rework and retesting, improves process efficiency, and supports sustainable semiconductor manufacturing practices. The approach demonstrates how embedded vision and TinyML technologies can enable intelligent, low-cost quality assurance systems for Industry 4.0 manufacturing environments.

Download

Share

COinS

United Nations Sustainable Development Goals