Customer Solutions

Automating a Color Fuse Vision Inspection System with NI LabVIEW and Machine Vision

Author(s):

Industry:

Product:

The Challenge:

The Solution:

At CNN Technology, we developed an automated vision inspection system for a European automotive components manufacturer. The device under test (DUT) is a printed circuit board (PCB), measuring approximately 50 cm by 28 cm that consists of 60 fuse-box holders and other electronic components. There are a total of 48 color fuses that need inspection. The other fuse-box holder is left blank intentionally.

There are two models of the DUT that differ in color and dimension regarding the fuses. The first model of DUT contains fuses with colors, including red, light brown, dark gray, blue and purple. The second model of DUT consists of red, green, blue, and orange color.

The region of color detection for the two models is slightly different because of the DUTs physical construction. The height of the fuse for the newer model is higher than the old model. In addition to checking the fuse color, the new system must verify the position of the triangle and circle shape that appears on the side of the labeling sticker.

Test System Components
The test system includes of the following components:

• National Instruments PCI-1411 color image acquisition cards and PC-DIO-24 digital I/O card. For image acquisition, we used three channels to capture the various color fuses and one channel to capture the side label. We also used a digital I/O card to control the start and stop button on the fixture.
  
• The system includes four units of a CCD color camera, which serve as a robust and compact image-capture device.
  
• We developed the test fixture to accommodate two DUT models with a fixed camera position. There are three camera holders mounted directly on top of the DUT that cover the entire region of interest for color fuse detection. We mount another camera on the side of the fixture to capture the side label. The fixture also consists of lighting and enclosure to provide a constant light source on the DUT.
  
• We used custom software to develop the system and to provide the user a virtual instrument to perform tests. We used LabVIEW software to compose two modes – production test mode of operator use and engineering calibration mode for engineer use.
  

Production Test Mode
We designed the interface as simply as possible, making operation easy and minimizing training of new personnel. Both of these factors are essential in a manufacturing environment. The operator graphical user interface (GUI) consists of the graphical display of the region of interest on the DUT as well as start test and print button. The start button duplicates on the fixture to provide a ‘mouse free’ environment if desired. Operators only press the start button to commence testing. When the unit fails, pop-up screens display the location of the fuse that failed, as well as the actual color fuse that should be in that location. An indication sound also generates the unit fails. Operators can press the stop button to stop the indicator sound. This step alerts the operator that the unit has failed and prevents the operator from skipping the failed unit. The result of every test logs to a disk file, and the user may select to print the test result with the print button.

Engineering Calibration Mode
The engineering calibration mode interface provides a flexible tool for calibrating the pass/fail tolerance. The adjustment of the tolerance is necessary if the fuse’s color varies slightly from one production batch to another. The changes of the supplied material can cause this variation. However, normally it will not need any adjustment once it is set up if the fuse’s color is constant.

Increasing Productivity by 200 Percent
By developing an automated vision inspection system using National Instruments LabVIEW and machine vision products, we increased productivity by 200 percent compared to the previous manual inspection method.

For more information, please contact:
Fok Kah Weng
Software Manager
CNN Technolgy (M) SDN BHD
737-1-10, Kompleks Sai Sungai Niedng
JLN. Sultan Azlan, 11900 Penang
Malaysia
Tel: 604-646113
Fax: 604-6448181
E-Mail: fok@cnn.com.my