Skip to content

Smart Metrology, Smart Choice

By Karen Haywood Queen Contributing Editor, SME Media
NoahofExactscanningpart.jpg
Exact Metrology’s Noah Branscum uses a Hexagon Absolute Arm to measure part features. (Provided by Exact Metrology)

Gone are the days when metrology was considered a necessary cost, to be avoided if possible. Manufacturers today embrace the philosophy of “you get what you measure,” or the more apt adage, “You get paid if what you deliver measures up.” But that doesn’t mean providers of metrology devices aren’t competing to deliver lower cost and more capable devices. For manufacturers it might be time to upgrade.

Why? Over approximately the last 10 years, metrology providers have embraced the digital revoluion. That means the quality and speed of laser scanners, video cameras, computers and software has improved. That is according to Mark Arenal, general manager of Starrett Metrology Systems Division in Laguna Hills, Calif., with corporate headquarters in Athol, Mass.  From handheld calipers to handheld scanners, or vision systems and touch probes, it’s worth looking at new technology that could save time, cost, and frustrations in your shop.

There is much to consider.

Take video cameras for example, a mainstay of many metrology applications. “Video cameras are much more capable at lower price points,” Arenal said. “Computers have much more processing ability. And software has evolved over the years. We’re now trying to find areas where we can further improve quality control, reduce scrap, and improve throughput in inspection.”

Laser light scanning also has gotten faster and more accurate, said Noah Branscum, short range applications engineer for Exact Metrology with offices in Cincinnati, Ohio; Brookfield, Wisc.; and Molin, Ill. Branscum is a fan of Hexagon’s Absolute Romer Arm and a recently-released Hexagon scanner that is, he said, “absolutely phenomenal at picking up tons of data.” The scanner delivers both higher quality and quantity of data and can be deployed as quickly as five minutes, Branscum said.

“In our departments of computed topography, short-range scanning, and long-range scanning, every piece of hardware in our arsenal has been redesigned or upgraded in some way with speed, quality of data, and being able to pick up other types of materials easier than before,” Branscum said.

Single-fixture means “one and done”

Improved metrology systems allow manufacturers to measure and inspect parts in one fixture position instead of having to measure once, then rotate the part and measure again.

Zeiss-O-inspect-at-customer.jpg
The Zeiss O-Inspect system includes multiple sensors, with several tactile sensors as well as a camera in order to inspect surface details with little or no manipulation of the part. (Provided by Zeiss)

“Everyone wants to measure everything in one fixture position,” said Christoph Stark, product manager for O-Inspect at Zeiss Industrial Quality Solutions in Oberkochen, Germany. “Manufacturers might have one system with multiple fixtures. If you use only an optical camera on top, you can measure only things you can see, but not behind or on the sides. You measure the part in one position, then rotate it. When you rotate the part in the fixture, you can measure only one part on one device; rotating 30 parts takes time. The solution is to have multiple sensors with several tactile sensors as well as a camera. You reduce time because you don’t have to manipulate the part. Manufacturers can measure more in less time with increased quality.”

One goal is to move away from having the first inspection take place in the quality-control department, Arenal said. Instead, parts and components are being inspected during manufacturing production to ease that quality assurance burden in the lab.

“Advances in technology, software and ease of use have allowed us to put inspection systems right on the shop floor,” Arenal said. “An operator can take a part during production, put the part on an inspection system, and put the part back on the machining center for more work if the part did not pass inspection. All of that is done within one cell on the manufacturing floor. It is not always necessary for the operator to rely on the quality-control department and get in line for that first article inspection.”

With that goal in mind, manufacturers increasingly are looking to measure only what matters—for example measuring only one or a handful of features on a 500-feature part instead of all 500, Stark said. For such entry-level measurement tasks, he said that the company’s new camera-based O-Detect measurement system is a good option for minimal investment.

For example, TRW in Germany manufactures safety systems for seatbelts on airplanes, which obviously have rigorous quality requirements, Stark said. For the final quality inspection, every feature must be measured. But during production, it’s necessary to measure and check just a few features. If those parameters are correct, the rest of the parameters will be fine, he said.

“Statistical process control focuses on the most critical characteristics,” Stark said. “You’re not able to measure all 500 characteristics on time during the cycle time; that 100-percent control also can cost more money than the manufacturer wants to spend. You have to reduce it to a reasonable number of characteristics and choose the most critical. TRW is using O-Inspect, our entry-level, multi-sensor solution.”

In other cases, it’s beneficial to measure one key component, such as a cellphone camera sensor, early in the manufacturing process, Stark said.

“There are 20 to 30 parts in a phone camera,” Stark said. “Some camera lenses are not good in the optical axis. If you measure only at the end, you may have to throw away the whole device. If you measure a single part earlier, you have to throw away fewer parts. That leads to less production loss.”

Overall, “The quality and complexity of camera components are rising almost every year,” Stark said. “We have to keep up with this trend. Therefore, we also need more quality in our sensors. We are constantly optimizing our optical sensors to measure the camera parts. For example, higher resolutions and magnifications are possible.”

Easier user interfaces is an ongoing goal, Stark said. “People who work on a shop floor are not really trained in metrology,” he said. “We are working to make it as easy as possible for beginners. We’re focusing on the entry budget as well as the entry user. What people are searching for is a one-button solution.”

Automatic inspections are another trend, Stark said. While manual optical measuring solutions in the past required that the machine table be moved by hand, now Zeiss’s fully-automatic CNC-based systems such as O-Detect and O-Inspect execute the inspection automatically. “They can measure 24–7,” Stark said.

Moving forward, metrology manufacturers face challenges as they try to achieve competing goals—for more exactness of medical and aerospace parts, more ease of use in many sectors, more affordable options for manufacturers who are extremely budget-conscious or not 100-percent sold on the technology.

Some manufacturers get one metrology system, let that system prove its worth, then implement the technology throughout the factory, Arenal said. Similarly, more industry sectors are accepting metrology technology as use becomes more widespread in industries such as biomedical.

“The value keeps accelerating to higher levels for metrology and inspection,” Arenal said. “The acceptance of these technologies also is accelerating. It takes a leading industry to set the pace. Biomedical is setting the pace, especially in the United States.”

Optical + digital

The combination of optical and digital takes the subjective manual element out of the inspection decision, producing more consistent results with data that can be tracked and traced, Arenal said.

Before digital metrology systems, an operator would look at the part, magnified on an optical comparator screen with multiple profiles on a mylar overlay, Arenal said. The operator would select the correct overlay, line it up with the part, and then make a judgment call on whether the part was within specifications. If not, the operator would mark on the physical part areas where it was outside tolerance, send the part back to be reworked, then check the part again against the mylar overlay, Arenal said.

AVX550_2360Starrett.jpg
The Starrett AVX550 Multi-Sensor Vision System has two cameras and lenses for sequential use in the same program along with touch-probe options. (Provided by Starrett)

With this system, “There’s no data involved,” Arenal said. “The operator can’t feed back any data on the machine part. If they are producing parts out of tolerance, the whole process shuts down until they fine-tune it enough to confidently proceed with producing parts. It’s an artisan process rather than a technically-driven process.”

In contrast, horizontal digital video (HDV) comparators can put a digital overlay on a video screen over the part brought in from a CAD file, and determine specifically how much deviation is present, Arenal said. Also important, the operator can feed that information back to the machine tool to improve the process.

For example, Starrett HDV300 and HDV400 Benchtop Digital Video Comparators enable tremendous measurement throughput.

“Subsequent parts come out without needing adjustment or rework,” Arenal said. “Part of it is throughput; part of it is having accurate data to make reliable decisions. For a manufacturer of orthopedic implants, there was previously a lot of subjectivity in the inspection process. By implementing the HDV system the process increased in speed by a factor of 10 to one. We can also track the data; the data resides with the part forever.”

Remaining challenges include movement, reflectivity, and size, Branscum said. “The most common, movement, is when the part you’re scanning moves in relation to the scanner, which causes misalignments,” Branscum said. “This allows us to get creative and quickly design mounting for the part we’re scanning.”

Reflectivity can be difficult when chrome or mirror-like finishes reflect the laser and will read incorrectly, Branscum said. Newer scanning software is making reflectivity less of a problem, Branscum said.

“Size is the last challenge, dealt with by deciding which scanner is appropriate for the size and tolerances needed or using reference points or spheres that are used to align larger pieces of data more accurately than just using curvature of what you’re scanning,” Branscum said.

Another challenge comes in meeting vastly varying needs of different manufacturing sectors.

“One the one side, the electronics industry wants us to make it as easy as possible,” Stark said. “We also need to include as many sensors to meet the requirements of medical and aerospace industries. Everyone wants to have traceable measurement results. Our goal is to have one software suite and manufacturers take the functionality they need.”

For example, according to Arenal manufacturers of medical products must validate any process on the factory floor, including and especially quality control.

Zeiss’s reporting software suite, PiWeb, can compare results machine-to-machine, plant-to-plant, and scale globally, Stark said.

“The biggest thing manufacturers are looking for is traceability, to really understand where a result is coming from and if they can trust it,” Stark said. “Manufacturers not only want to measure parts but also to inspect. For example, they want to see scratches on the glass. They want to combine all this information in production.”

This summer, Zeiss is beginning a pilot program to combine measurement and inspection, Stark said: “The demand is there.”

  • View All Articles
  • Connect With Us
    TwitterFacebookLinkedInYouTube

Always Stay Informed

Receive the latest manufacturing news and technical information by subscribing to our monthly and quarterly magazines, weekly and monthly eNewsletters, and podcast channel.