Vision Sensors Error-Proof Oil Cap Assembly
Supplying parts to the world’s leading automotive companies leaves no room for error. That’s why Miniature Precision Components (MPC) uses three vision sensors to error-proof the automated assembly of oil-caps at its Prairie du Chien, WI, facility. With 41 molding machines ranging in size from 25 to 550 tons, this 100,000 sq.
Supplying parts to the world’s leading automotive companies leaves no room for error. That’s why Miniature Precision Components (MPC) uses three vision sensors to error-proof the automated assembly of oil-caps at its Prairie du Chien, WI, facility.
With 41 molding machines ranging in size from 25 to 550 tons, this 100,000 sq. ft. facility employs about 450 people. MPC’s four manufacturing plants generate approximately $167 million per year supplying the automotive and commercial industries with high-quality injection-molded parts and assemblies.
Quality is a given at MPC. In fact, the manufacturer has been a Ford Q1 preferred supplier since 1989, and has received numerous supplier awards from the likes of GM, Nissan, Harley-Davidson, and Chrysler.
In addition to oil caps, the facility manufactures a host of other thermoplastic parts and assemblies, including PCV valves, thermostat housings, and quick-connect ports for emission control systems.
“We achieve quality through automation, and machine vision has been a key component of our automation strategy for the last seven years,” explains Shane Harsha, MPC manufacturing engineering manager.
An automated oil-cap assembly system is a case in point. MPC automation and tooling engineer Brian Champion recently augmented traditional tooling and sensor technology with Checker vision sensors from Cognex Corp. MPC says that the upgrade was very cost-effective and delivered greatly improved repeatability to provide more efficient production of defect-free oil caps.
“Because vision sensors are so simple to set up and easy to install, they offer a very cost-effective solution for inspections where traditional sensors are not reliable and a full-blown vision system is too expensive,” explains Harsha.
The MPC oil-cap assembly system installs O-ring seals into molded thermoplastic caps, and then prints on top of the caps. The system uses two vibratory bowl feeders, about four feet in diameter. One feeds O-rings and the other feeds oil caps into the process.
A vibratory bowl feeder consists of a large bowl with a spiral ramp up the side. As the bowl vibrates, the parts work their way singly up the ramp to an inline conveyor. At the end of the inline conveyor, a pick-and-place arm transfers O-rings to the first station on a rotary assembly dial.
After an O-ring is loaded onto the dial fixture, it indexes to the second station. There, another inline conveyor feeds blank caps from the other vibratory bowl feeder to a second pick-and-place arm that presses the caps on top of the loaded seals.
Then the assembled cap and seal continue on the rotary dial through the pad printing and final inspection stations to complete the process.
Tight control of O-ring and cap orientation is critical to ensure the seal is properly installed so that the finished oil cap will function as intended. The cap must also be in the right orientation prior to pad printing in order to meet stringent quality requirements.
Because the hard tooling and traditional sensors in the cap assembly system proved unreliable, MPC selected three Checker 202 vision sensors to ensure proper O-ring and cap orientation.
“The small size, built-in lighting, variable working distance, ladder logic and free-running capability made the vision sensor units simple to install. There was no need to wire them to a PLC, no need to install and wire a trigger sensor, and the four-step set up makes it by far the easiest vision sensor that I’ve ever used,” says Champion.
The first vision sensor detects inverted O-rings between the vibratory feeder bowl and inline conveyor. A second vision sensor checks that the O-ring is positioned properly on the dial fixture before the cap is pressed on. A third ensures cap orientation is correct prior to assembly and printing.
Each O-ring has a sealing bead on one side. The bead must be placed face down when the seal is loaded onto the assembly dial. If it’s not, the machine shuts down. The operator must then access and reposition the seal before restarting the machine.
Mechanical tooling on the feeder bowl was designed to prevent inverted O-rings from entering the process. However, the tooling was unreliable, according to Champion. O-rings that were very slightly warped or not perfectly flat occasionally made it past the tooling, and were loaded upside down causing the machine to shut down.
“Having the operator flip these seals and restart the machine was really eating into our efficiency,” says Harsha. “If the production rate dropped from 360 to 200 caps per hour, it cost us about $20,000 a year in downtime. As we approach full production volumes, that cost could increase to as much as $120,000 per year.”
How vision sensors work
Image analysis software uses multiple software agents that extract information from selected areas in the image. The process starts with selecting the built-in, part-finding sensor, and then placing inspection sensors on the features to inspect. The device includes three types of inspection sensors that can address a wide variety of applications:
Brightness sensors look for light and dark areas;
Contrast sensors check features that contain light and dark areas, such as date codes, threads and barcodes; and
Pattern sensors trained to know what a feature looks like and then signal when it is spotted.
To detect the inverted O-rings in this application, Champion first trained the part-finding sensor to look for the O-ring in the image. Then he positioned a pattern sensor in the correct location to verify the presence or absence of the sealing bead. The pattern sensor looks for the pattern of the sealing bead on the O-ring and then signals when it is detected. The pattern sensor remains in a fixed position relative to the part-finding sensor so that it is always in the correct location to look for the shape of the sealing bead. If the sealing bead is missing, the vision sensor sends an output through an optic coupler to a pneumatic solenoid that blows the inverted O-ring off of the line and back into the feeder bowl to be recirculated.
Because it was such an easy and cost-effective solution and took less than an hour to set up and install the first vision sensor, Champion decided to completely error-proof the oil cap assembly process by adding two more. Both are used at the next station where the cap is pressed onto the loaded O-ring. One is mounted on the moving pick-and-place arm. The other is fixed above the inline conveyor that feeds caps to the process.
Champion set up the vision sensor on the moving arm similarly to how he set up the one looking for inverted O-rings exiting the feeder bowl--first using the part finding sensor to look for the O-ring in the image, then a pattern sensor to verify the presence or absence of the sealing bead. This allows the vision sensor to ensure the O-ring is properly loaded before the cap is pressed on.
The final vision sensor mounts above the inline conveyor feeding caps to the process, just upstream of the pick-and-place arm that presses caps onto the loaded O-rings on the assembly dial. This vision sensor checks cap orientation. However, it was set up in much the same way by first training the part-finding sensor to recognize a corner radius of the oil cap, then training two pattern sensors to recognize the oil-can handle and oil drop graphics.
By training on two patterns, the vision sensor can determine cap orientation. If it is not in the correct orientation for installation, the vision sensor signals the pick-and-place arm controller to rotate the cap 180 degrees before placing it on the assembly dial.
“The system has helped us achieve zero-defect rates in the manufacturing process,” notes Harsha, “while lowering scrap and is the perfect solution for many of our inspection and error-proofing applications.”
John Lewis is market development manager at Cognex. Contact him at firstname.lastname@example.org .
|Search the online Automation Integrator Guide|
Case Study Database
Get more exposure for your case study by uploading it to the Control Engineering case study database, where end-users can identify relevant solutions and explore what the experts are doing to effectively implement a variety of technology and productivity related projects.
These case studies provide examples of how knowledgeable solution providers have used technology, processes and people to create effective and successful implementations in real-world situations. Case studies can be completed by filling out a simple online form where you can outline the project title, abstract, and full story in 1500 words or less; upload photos, videos and a logo.
Click here to visit the Case Study Database and upload your case study.