Vision sensors play key role in lock assembly
A manufacturer and supplier of cam, switch and special-purpose locks for move then 50 years, Fort Lock (River Grove, Ill.) recently implemented new machine vision sensor technology to ensure proper lock functionality. The system checks for the presence of springs during the assembly process. Up to six springs per lock are inserted during the automated assembly process, and each one needs to be ...
A manufacturer and supplier of cam, switch and special-purpose locks for move then 50 years, Fort Lock (River Grove, Ill.) recently implemented new machine vision sensor technology to ensure proper lock functionality. The system checks for the presence of springs during the assembly process. Up to six springs per lock are inserted during the automated assembly process, and each one needs to be present and correctly positioned to keep a lock's tumblers in contact with an inserted key as it turns.
"In the past, we mechanically checked for the springs. However, this process involved contact between the probe and the part, which occasionally caused a spring to bounce out of the assembly," says Mark Mazanec, manufacturing engineer at Fort Lock, a division of CompX International Corp. (Dallas, Tex.). "This would require reworking parts, which was ultimately affecting our throughput. We knew we needed a non-contact method of checking parts, but, until recently, optical methods were too costly and required exacting lighting and positioning."
Low cost, high speed
After learning that a PC-based vision system would have been overkill for the application and that a laser beam system would have required consistent spring placement, which the assembly could not provide, Mr. Mazanec and his colleagues looked into low-cost vision sensors. They selected an In-Sight 2000 from Cognex (Natick, Mass.). In-Sight is a vision sensor that features an industrially hardened vision processing unit; separate high-speed digital camera; onboard light control; built-in discrete I/O points; and a standard VGA display output.
During the manufacturing process at Fort Lock, springs are automatically loaded into a lock assembly, which moves from station to station on an indexing table, where subsequent assembly operations are performed. As the lock moves into the inspection station, it stops under the vision camera, located 3.5 in. above it.
The camera, which is set in continuous acquisition mode, captures an image of the lock assembly, and transfers it to the vision processor. There, the image is analyzed by the sensor's PatFind part location software tool, which verifies the presence/absence of each spring. The entire inspection cycle time, between the initial camera trigger and inspection result, occurs in a fraction of a second. A key attribute of the tool is its ability to locate springs even when lighting is poor, or when springs vary in their anchored position from part to part.
Upon verifying the presence and proper position of each spring, the vision sensor sends a "pass" signal to a PLC, and the indexing table moves the part along to the next station, where the next assembly step is performed. If the part fails the inspection, no further assembly processes are performed on the "failed" part, and it is indexed through the remaining stations back to the operator for removal. Green and red LEDs appear on the user interface screen, making it easy for operators to monitor activity.
Inexperience? No problem
Even with no prior machine vision experience, Mr. Mazanec reports that he was able to set up the vision application alone with the sensor's spreadsheet interface. The process involved selecting vision tools and parameters from drop-down menus using a handheld control pad. The vision spreadsheet then automatically generated tool results into worksheet cells, which were then linked to perform the assembly verification.
Since installation, the vision sensor has been running for eight hours per day without any failures. Mr. Mazanec adds that Fort Lock is now able to better ensure product quality, and can realize higher throughput rates. "The In-Sight sensor has greatly improved our production process," he says. "It allows us to efficiently use automation that has inherent assembly shortcomings by catching an error before subsequent operations are performed. And, unlike the mechanical checking process, it won't cause bad parts. Our initial goal was having a non-contact system that would allow us to get our maximum throughput, and we have finally achieved it."
For more information, visit www.cognex.com .