Robotic Soldering Vision System

Cognex Insight 2800

The In-Sight 2800 vision system combines artificial intelligence (AI) with traditional rule-based vision tools to solve a wide variety of applications. From presence or absence detection to classification and character reading, the In-Sight 2800 offers an easy-to-implement solution to eliminate errors.

Features

• High Performance Vision Sensor With Adjustable RGB Lighting
• Presence/Absence Detection of Features, Patterns, Edges, and Circles
• Edge Learning-based Sortation with ViDi EL Classify
• Ethernet communication with TCP/IP, Profinet, Ethernet/IP, SLMP, and FTP for saving data

Demo: Solder Quality Inspection Using ViDi EL Classify

• In-Sight 2800 can be trained to accept or reject PCBs based on solder quality and presence of defects.
• Using the ViDi EL Classify tool, can inspect a region containing a single solder pad or multiple pads.
• Train your system to make a confident model.

Keyence IV4

Beyond its predecessors. It excels in detecting parts without the need for precise positioning adjustments, ensuring correct part quantities and placements, and accurately identifying and tallying targets even in challenging conditions. Renowned for its simplicity and reliability, the IV Series has been further fortified in the IV4, boasting increased brightness, versatile field-of-view options, and finely tuned built-in hardware that ensures robust performance in varying environmental conditions such as ambient lighting and subtle surface changes. With its enhanced AI tools, the IV4 effortlessly tackles applications that traditional vision sensors find challenging.

Features

• Inspection at 100% is feasible.
• It accommodates a broad range of detections.
• The sensor is designed for easy operation by anyone.
• It ensures consistent and accurate detections.
• It reviews the entire surface, ensuring stability even with misaligned objects.
• Clear, distortion-free images can be captured.
• The IV Series can be set up in three different ways: Via a dedicated panel, through a display, or with a PC.

Demo: Vision Results

The automatic teaching inspection tool uses the image sensor to learn the variations and individual differences among non-defective parts and recognizes those that differ as defective parts. These algorithms, which are considerably close to human sensitivity, eliminate unstable elements to successfully resolve the inspection.

The setup is done simply by presenting non-defective parts, reducing the conventional need for specialized knowledge and complicated configuration. This is an inspection tool that makes it possible for anyone to achieve and maintain stable inspections.

The range of variability of defect-free parts is determined by learning color information per pixel. What cannot be determined with a black and white image, such as color irregularities of non-defective parts, is also learned accurately.