Manufacturing Inspection System Uses Machine Vision to verify assembly and labeling

Reducing human error with automated inspection

Client – Automotive Component Manufacturer

Challenge

Our client already had an end-of-line tester in place.  However, preventing incorrect product shipments drove them to add machine vision capabilities to verify that the part being packed is of the correct physical configuration and that the part was labeled correctly.  They also wanted a more automated way to track which serial numbers were being shipped.

Solution

Viewpoint enhanced the existing end-of-line tester by adding machine vision capabilities to verify correct part assembly and part labeling.  This capability also allowed for automated tracking of which parts went into which shipping container.

Benefits

  • Automated part assembly verification to reduce human error from manual visual inspection
  • Automated label verification to reduce the chance of shipping the wrong product

System Overview

The enhanced system added machine vision-based capabilities to an existing end-of-line manufacturing test system.  New hardware (cameras, lighting, fixture) was selected and integrated by the client.  Viewpoint developed the image analysis routines using the Cognex In-Sight software.  These routines were then downloaded and controlled using LabVIEW software developed by Viewpoint.  In addition, the LabVIEW GUI contained the image acquired by the camera and the results of the image analysis.  The tester can inspect four different part types.

The software essentially performs the following functions:

  1. Look up the expected characteristics of the part being inspected.
  2. Populate the on-camera In-Sight “spreadsheet” with configuration information used in the image analysis/inspection.
  3. Trigger the image capture and read results from the on-camera spreadsheet.
  4. Use the on-camera image analysis to check a critical angle of the part as the part is set in the nest fixture.
  5. Check the information laser etched on the part and compare the results with what should be on the part (relative to the barcode read in for the lot and the 2D barcode on the part) using the OCR/OCV capabilities of the camera.
  6. Perform other physical part characterization image analyses to verify the part was correctly labeled & assembled.
SOFTWARE FUNCTIONS
Look up expected part characteristics
Trigger image capture
Read results of on-camera image analysis
Display image taken by camera and show if test passed or failed
Monitor contiguous part failures & initiate shutdown
Log vision test failures to database
HARDWARE USED
Existing end-of-line tester
Test Fixture
(qty 2) Cognex camera
Lighting for camera
INTERFACES / PROTOCOLS
TCP/IP