0.1mm Precision Automatic Grasping, SingTown Technology Empowers New Scenarios in Electronic Assembly

On the automated production lines of electronic assembly manufacturers, the feeding and placement of loose SOP chips remains one of the critical precision processes. Due to the random orientation of chips, with varying front/back sides and angles, manual identification suffers from high error rates and slow speeds, affecting placement accuracy. The manufacturer aims to utilise SingTown Technology's OpenMV smart camera, employing embedded vision algorithms to swiftly identify the position, angle, front/back side status, and Pin1 marker point of SOP chips. The data will then be relayed to a robotic arm for automated grasping and placement. The identification area is approximately 200×200mm, within a distance of 200mm, primarily targeting SOP14/SOP16 chip types. The system must operate stably under complex lighting conditions, achieving millisecond-level identification feedback to meet both real-time and high-precision requirements, thereby providing an intelligent visual identification solution for electronic production lines.

Electronic assembly manufacturers have achieved significant results in the automatic identification and positioning of bulk SOP chips by utilising SingTown Technology's OpenMV intelligent camera.

The system employs an OpenMV intelligent camera installed above the material tray to capture chip images in real time. Its edge-side AI utilises machine vision algorithms to perform contour detection, morphological analysis, and orientation determination for SOP chips.

Initially, the image preprocessing module denoises and enhances edges, segmenting overlapping or rotated chips. Subsequently, the deep learning model, combined with template matching methods, swiftly identifies chip models. Finally, the coordinate extraction module transmits the precise X and Y positions along with orientation information of each chip to the robotic arm, enabling automated grasping and placement.

The operational results indicate that this OpenMV intelligent camera solution can achieve over 99% recognition accuracy in high-density bulk material scenarios, with single-chip positioning precision reaching the 0.1mm level, significantly enhancing assembly line efficiency and stability.

By introducing SingTown's intelligent cameras, manufacturers have not only achieved automated identification and precise positioning of bulk SOP chips through edge deployment but have also effectively reduced labour costs and defect rates, providing a reliable and intelligent upgrade path for electronic assembly production.