Networked sensors aid robots building keyless entry units
Michael Williams -- Control Engineering, 2/1/2002
Fitting keys and other accessories onto key rings has never been easy. Spreading the tight-fitting spring often takes people a couple of attempts, and it's no easier if you happen to be a robot assembling automotive keyless entry units. That's why system integrator GSMA Systems (Palm Bay, Fla.) recently helped one of the nation's largest automotive suppliers to design and build a new automated robotic assembly system for its keyless entry transmitters and receivers.
"In designing the vision and verification system," says Mark Senti, GSMA's president, "the main requirements were modular vision systems that could be easily programmed and networked." Consequently GSMA developed a system based on eight DVT (Norcross, Ga.) Series 600 SmartImage Sensors networked over Ethernet and six GE Fanuc (Charlottesville, N.C.) LRMate200i robots. Unlike traditional machine vision systems, SmartImage sensors are completely self-contained with all image acquisition electronics, microprocessor and communications in a palm-sized unit.
The client's system automatically inspects and assembles keyless entry units from a number of different parts, including plastic cases, keypads, printed circuit boards (PCBs) and key rings." The final product is a fully assembled set that consists of two transmitters and one receiver. Up and running for about two years, the system helps GSMA's client assemble one keyless unit every 6 to 8 seconds, or approximately 2.5 million units per year. The system was expected to have a 1.5-year payback.
To build the keyless units, the robotic production line's separate workcells conduct an assembly and inspection process. First, a vibratory feeder feeds two transmitter case shells to a date stamping device. Next, keypads are inserted into the shell and inspected by a DVT Series 600 SmartImage Sensor that checks that the right keypad has been used and that its position is correct. "To perform this inspection, DVT's Series 600 systems are taught images of the different keypads used in the product. The smart camera is then programmed to perform pattern matching on a specific part of the image to decide if the right product is in place," says Mr. Senti.
Meanwhile, the PCB controller used in the transmitter's shell is inspected at another workcell that checks the ID code of each board. Once verified, these boards are placed in the transmitter assembly. Using another Series 600 smart camera, the back shell of the plastic transmitter case is imaged for specific fiducials that check which type of shell is in place. After inspection, both halves are conveyed to a snapping station that assembles the complete keyless entry transmitter. When assembled, the final workcell attaches a keyring to the transmitter, and checks placement with another Series 600 camera.
To develop this assembly system, customized machine vision software was created off-line on a laptop PC using DVT's FrameWork software. This meant teaching the many different part types to the system for label verification, part presence, and part-type inspection before deploying it on the production line. Other advantages GSMA found in DVT's SmartImage Sensors were their ability to troubleshoot easily and to add more products through an Ethernet port. These changes were easily tested off-site with the SmartImage Sensor emulator, which allows users to apply FrameWork software without having a SmartImage Sensor connected to the PC by running an emulation program on the PC and using previously recorded images. GSMA used the emulator to program SmartImage Sensors, and then transmit inspection configurations to the manufacturing line via the Internet.
-
Very well written and informative article on the use of sensors in automotive electronics assembly.
Eleanor T. West - 2007-1-12 06:09:00 CST
Machine Vision
08/01/2006
View All Control Engineering Webcasts from CTL
