Improved inspection process pays for itself in less than a year
A new vision system integrated into the inspection process at a car battery manufacturing plant improved the error detection rate to 99% and saved the manufacturer both money and labor.
Car batteries are made up of six books with two cells per book. Orange paint is applied to the battery book to indicate polarity. The orange alternates with white to ensure proper alignment and therefore proper performance of the battery. The manufacturer was using a photo eye that inspected each cell for a specific shade of orange in the location where color was expected.
The color was well defined before the battery was molded with hot lead, but a new process caused the orange to diminish, and burned cells turned the orange to brown. Inspection instruments were no longer picking up on cell alignment due to a change in the color. As a result, operators were spray-painting cells to reapply the orange. Besides being inefficient and messy, the manual intervention could lead to human error.
Further complicating the process, battery books sometimes were pulled from different lines where the photo eye was trained to identify a different shade of orange. Another problem was overspray of the orange paint. Customers seeing orange paint on the exterior of the batteries mistakenly perceived the batteries as seconds.
The user requirements were:
- Remove the need for manual operator intervention
- Maintain current “reversed cell” notification and processing
- Create an ability to identify multiple shades of orange and brown
- Inspect both sides of all 12 cell areas (colored and non-colored)
- Eliminate spray paint
Solution is high resolution camera
Matrix Technologies engineers introduced a 7900 series Cognex camera, featuring ethernet connection, color and 12 megapixels. To integrate the camera into the manufacturing process, the team installed the Cognex In-Sight Explorer software to set up and train the cameras. Next, they created data tags and defined the inspection area. The color palette was created to detect multiple shades of orange and brown. Next, logic was established that could alter recipes for different sized batteries.
To utilize the new camera, the programmable logic controller (PLC) to add on profile (AOP) specific to the Cognex cameras was installed. Then the PLC was programmed to process input/output data, count pixels, select recipes and send confirmations. The team tied a strobe light to the position switch, which acted like a flash on a traditional camera to illuminate the inspection area. Finally, the team set pass/fail criteria to distinguish there was enough of the acceptable color and, conversely, enough white. With a few tweaks of the color palette, the system was ready to go.
Results exceeded expectations
In a trial run, the vision system performed perfectly. The camera positively identified the inspection areas 99% of the time without the need of the added spray paint. From the operators’ perspective, nothing changed except they were responsible for one less task. The battery manufacturer implemented the new inspection process across all remaining production lines.
The result is the manufacturer reduced the cost of paint and the labor to apply and clean up the paint. The project paid for itself in less than a year.
Matrix Technologies is one of the largest independent process design, industrial automation engineering, and manufacturing operations management companies in North America. To learn more about our manufacturing operations management capabilities and manufacturing process control solutions, contact Michael Helmstadter, Senior Consultant.