The question is no longer whether you should check your products, but when and how often.

Key Technology’s onCore system uses laser scanning and vision systems to sort and remove cores from chopped-up cabbage and spinach. Source: Key Technology.


X-ray images of a loaf of bread in an aluminum tray show defective (top) and passed (bottom). The X-ray system can detect SS fragments down to 0.6mm, SS wire (0.3mm x 2mm long or larger), glass down to 1.5mm and ceramic down to 2.0mm. Source: Anritsu.

If you’ve been checking your finished products with vision, metal detection and X-ray inspection systems before they go out the door, you’re certainly heading off some potential catastrophes, should you find a mislabeled or contaminated product. But while you can save money by avoiding recalls and potential lawsuits, are you really being efficient when you find tramp metal or stones in finished, packaged product? Maybe you can afford to throw away a little rejected food, but what about wasted packaging-which for some food products, costs more than the ingredients that went into them?

While labeling, packaging, palletizing, storing and shipping costs are high, the fact a product made it through production and packaging before foreign material (FM) was discovered adds up to wasted material and production time. Checking incoming ingredients is paramount. Can you trust that your suppliers checked their products, especially if they’re in South America or China?

“We just shipped an X-ray machine to a pepper processor that’s getting its products from Central America,” says Ray Spurgeon, Eriez Magnetics product manager. “Everything imaginable is in these incoming [ingredients]-from stone to bone, glass and cans.” Spurgeon says checking ingredients from these countries is crucial to producing contaminant-free products.

Borne out of an application from the plastics industry, a pneumatic line metal detector for checking for metal contaminants in bulk materials in food-grade applications was developed by S+S Inspection Systems. “If you have sugar and flour and other ingredients in silos and pneumatically convey them, an in-line metal detection system can catch a problem before you mix the ingredient with other ingredients, cook it and package it,” says Doug Pedersen, S+S sales manager. Consequently, processors can recover overhead by rejecting an incoming ingredient before the product is made. “By the time you get the flour to bread or cookies, you’ve added a lot of extra cost.” In addition, says Pedersen, processors can identify suppliers with repeat problems and take appropriate action.

While checking incoming ingredients makes a lot of sense for a lot of processors, there is still a number without end-of-line inspection systems, says Robert Scott, Mettler-Toledo Safeline product manager for X-ray inspection systems. In today’s litigious climate, processors need end-of-line inspection as a liability policy. “Today, the vast majority has protection, but there are still a few that don’t. They just stick their heads in the sand and say they never had a problem in the past,” says Scott.

But processors shouldn’t throw technology at a problem without doing their homework. “If you don’t know what kind of contaminations you might have, you can’t really justify a piece of equipment to do the checking,” says Robert Rogers, Mettler-Toledo Safeline senior advisor for food safety and regulations. “But if you go through a detailed process of determining types of hazards and where they exist, and then put practices in place to prevent them, the inspection equipment becomes a tool to identify a potential problem and its solution.” Rogers says detecting is the easy part; what’s vital is how processors use the data the equipment provides to locate contamination points, prevent further occurrences and implement procedures to eliminate contamination sources.

Inspection of two parallel product lines with foil packaging runs at the rate of up to 600 products per minute on RAYCON X-ray equipment from S+S Technology. Source: S+S Technology.

Vision sees product and packaging quality

The two stereotypical images most people have of vision systems is bottle-cap and label inspection. While both of these are strategic, vision systems have a much larger universe of uses-even within these two applications. Starting with the latter, label inspection goes beyond making sure the label is on and in the right place. The most important issue, according to Mike Gay, Rockwell CPG industry manager, is to make sure what’s on the label reflects what’s in the product and agrees with any packaging. This includes any information on potential allergens.

This information and more concerning lot tracking, ingredient sources and equipment used to make the food product can be contained on labels-often in 2D barcodes-and is key to traceability, says Jeff Schmitz, PPT Vision director of marketing. Vision systems are a key element in traceability schemes and can be tied into labeling systems to verify the right labels are put on the package.

“A product recall is quite often the result of a bad package, meaning the wrong label,” says Tom McLean, Mettler-Toledo CI-Vision general manager. Often the product quality is fine, but the processor forgot to indicate it was made with peanut oil, adds McLean. “Label mix-up prevention is a very important tool for vision.”

When a food processor runs a special time-limited product run, for example, supporting a sporting event like a NASCAR race, it wants to be sure the label matches the packaging (internal and external) and is being shipped to the right distributors/customers at the right time, says Gay. Here is where the processor’s ERP system coordinates data from both the vision and labeling systems.

Another issue, which may not be so apparent to hurried plant floor production workers, but is crucial to the marketing department, is the quality of the label itself. Vision systems can continually check for off-color, fuzzy or improperly placed labels. “I would contend faulty labeling-as much as food contamination-is a real problem; the one that gets people every day is seeing a wrong-colored or wrinkled label on the grocery store shelf,” says McLean. If the label isn’t right, consumers automatically think the food isn’t right either, and they won’t buy it.

The issues with bottle caps or cup seals are two-fold, says McLean. First nobody wants to buy a container of cottage cheese with particles hanging out of the cup. It doesn’t look good, and there’s no guarantee the food isn’t contaminated. Closely allied to the latter is: If a seal doesn’t appear tightly and perfectly affixed to the container, consumers are likely-and rightly-to expect product tampering.

For the traditional bottle fill/cap inspection, vision sensors are typically employed. In the past, a complete vision system (camera, lens, lighting and electronics) was priced well beyond $30,000. Today, all components have been placed in a single, small package at a fraction of the earlier cost. Services provider Hamilton/ITS specified a Checker 101E sensor from Cognex to be installed at Meridian Beverage Company of Atlanta. “The new line runs at 375 bottles a minute so downtime on our system wasn’t an option,” says Ray Hamilton, president. The vision sensor has a built-in PLC that accepts encoder feedback, which allowed easy integration on the line to kick out rejects.

Mettler-Toledo Safeline PowerPhasePRO metal detector checks fresh, incoming fish for metallic contaminants before further processing. Source: Mettler-Toledo Safeline.

Vision sorts out food quality

Vision systems can make qualitative judgments about food quality. For example, Schmitz looked at an application using color vision sensors with white lighting to sort tuna for grading purposes. Tuna is graded into various product-use categories according to color, for example, sushi-grade, canned tuna or cat food.

Bret Larreau, Key Technology optical product manager, points to another sorting application for vision sensors. Historically, processors of kettle-style potato chips had to rely on hand sorting because earlier automated inspection systems couldn’t detect problematic clusters of chips, which are common to kettle chip production. To address this application, Key modified its existing Optyx sorter with a specialized off-belt scanning zone that measures opacity to detect and remove clusters of kettle chips in addition to standard on-belt scanning that identifies defects and FM. Removing clusters and FM from the line enhances product quality and maximizes food safety.

Another application for color vision sensors along with a laser system is sorting and removing the cores of chopped-up heads of cabbage and lettuce, says Larreau. After mechanical separation in the onCore machine removes 85 to 90 percent of the core, a combination of laser technology and color cameras detects differences in the structural properties of the product to remove pieces of core and FM, and determines each object’s size, shape and color to identify and remove defects.

The metal is a tramp

Two general metal detection/removal technologies are widely used today, and they can often be deployed together-at least for ferrous metals. Permanent magnets (PM) collect only magnetic (ferrous) metal contaminants, and metal detectors can generally detect ferrous and non-ferrous metals (aluminum, copper, bronze, brass, lead and most SS), says Martin Lymn, Loma general manager, sales and service. While PM technology is generally limited to homogeneous, free-flowing powders or liquids, electronic metal detection can be applied in varying degrees to most food applications.

Knowing when to mix and match these technologies and X-ray systems to the application is important. “If you’re looking for fine pieces of ferrous dust in flour, metal detectors won’t be any help-nor will an X-ray machine,” says Pedersen. A rare earth magnet will grab those pieces of dust-sized metal particles. For hot dog applications, Pedersen suggests a metal detector on the emulsion feeding the linker and one at the final package inspection-both for less than the cost of one X-ray machine.

In many cases, combining rare earth magnets and metal detectors does a good job of detecting and removing tramp ferrous metals and finding non-ferrous metals, including some SS, depending on the alloy. But processors need to be aware that locating powerful, rare-earth magnets near sensitive metal detectors may cause unexpected results in metal detector performance. In addition, metal detectors running in close proximity should be operated on different frequencies to avoid interference issues and inaccurate detection.

Generally speaking, metal detectors can’t be used for checking packages with metal foil lids, but there are exceptions. Steve Gidman, president of Fortress Technology, describes one application. “The processor provided airlines with juice cups and was required to inspect the product after the foil lid was installed. Using a pipeline detector upstream of the filler [checked for] small contaminants, but the stainless steel components of the filler could still fall into the process. Since the filler components would rapidly fall to the bottom of the cups, Fortress designed a system using a sensor beneath the belt inspecting the lower area of the filled cup to a high degree of performance while ignoring the foil lids.”

Metal detectors can serve other purposes as well. In some instances metal detectors are also capable of being used as temperature gauges. For example, a frozen product will run as a non-reactive product. As the product thaws, it becomes reactive. Some food processors use metal detection at the back end of freezer tunnels, or for product coming out of storage freezers, not only to inspect for metal contaminants but also to monitor/regulate freezer temperature.

Metal detectors can also be used to determine if metalized foil pouches are present (one processor uses inverted reject metal detectors to insure metalized cheese pouches are actually included in the packaging). Ferro-search metal detectors are designed to detect ferrous metals and SS in products contained in aluminum trays or metalized foil packets. But if a processor has an issue with non-ferrous metals as a potential metal contaminant in its process, X-ray detection will most likely be a better option.


X-ray, the almost all-seeing

“X-ray creates a picture of what is otherwise out of sight,” says Allan Anderson, InspX LLC president. “X-ray inspection does not lose resolution or sensitivity with changes in condition (temperature/salt content, etc.) of the product it is inspecting.” However, while X-ray works in opaque situations, not all materials can be seen by X-ray. The more the contrast in the molecular make-up of the contaminant from the food material in which it is mixed, the more detectable it is. So while an X-ray system can detect contrasting amounts of metal, stones and bone, it will not “see” aluminum and soft cartilage-like fragments mixed in with chicken breast meat. But, since it can see through aluminum foils and several types of cans, it can find metal contaminants where metal detectors won’t work.

One application where X-ray is effective is in the detection of broken hypodermic needles, sometimes found in pork carcasses, says Erik Brainard, Anritsu Industrial Solutions USA general manager. Metal detection systems have problems with this application for several reasons. First, SS (depending on alloy) can be difficult to detect; second, there is an orientation effect that may cause the needle to go undetected; and third, the meat is reactive, which places a further burden on metal detection sensitivity.

While an X-ray system is not a replacement for a vision system, it can be used in some applications that can be troublesome for vision. Spurgeon recalls a bean processor that uses a vision system to help sort out foreign objects, including stones. The problem is two-fold, he says. First the conveyor bed must be singluated, i.e., adjusting the process flow for a single layer of beans-because vision can’t see through several layers. Second, stones with the same color and specific gravity as the beans won’t be rejected. An X-ray system can see through several layers of beans to find stones, metal or glass contaminants.

If a processor supplies Walmart or Costco and uses a foil pouch, bag or lid-anything foil-it should expect to use X-ray as the final inspection step, says Kevin Jesch, Heat and Control product manager for metal detection and X-ray systems. Both retailers expect top quality in the products they buy-especially if they put their names on the product. While X-ray technology hasn’t changed much, recent design/application challenges are being met, says Jesch. Because washdowns are harsher, an IP-69K-rated enclosure is a necessity, and a refrigeration unit is used to cool the system. Equally important, user interfaces are becoming available in several languages, and set-up procedures often include auto-learn functionality so plant floor workers can set up the equipment for new products and start using it immediately.

Nutrichem, a European producer of specialized dietary liquids and powders, already had magnets and conventional metal detectors in place, but it wanted to take quality to a higher standard as well as meet all food safety regulations. With aluminum foil packaging and metalized films, the company purchased RAYCON X-ray scanners from S+S Inspection for end-of-line checks. The X-ray systems inspect 600 products per minute on two parallel lines.

Peter Jonigkeit-Burke, FBL primary and secondary production, powder and pouch production, Nutrichem Diät+Pharma GmbH, is satisfied with the X-ray scanners from S+S: “The reliability of deliveries to customers is of paramount importance to us, so we constantly seek to exceed requirements of our quality management systems and increase productivity. Contaminant detection and removal has been an important subject at Nutrichem for a long time, and presently X-ray systems offer the best technology to meet our demands. What we want is a final inspection of aluminum-packed products, because this eliminates any possibility of contamination at a later stage.”

For more information:
Doug Pedersen, S+S Inspection, 716-297-1922, doug.pedersen@sesotec.com
Robert Scott, Mettler-Toledo Safeline, 813-889-9500, robert.scott@mt.com
Jeff Schmitz, PPT Vision, 612-819-6545, jeffs@pptvision.com
Tom McLean, Mettler-Toledo CI-Vision, 630-446-7700, tom.mclean@mt.com
Bret Larreau, Key Technology, 509-529-2161, blarreau@key.net
Ray Spurgeon Jr., Eriez Magnetics, 814-835-6298, rspurgeon@eriez.com
Martin Lymn, Loma Systems, 763-656-7808, martin.lymn@loma.com
Steve Gidman, Fortress Technology, 416-754-2898, steve@fortresstechnology.com
Robert Rogers, Mettler-Toledo Safeline, 813-342-9138, robert.rogers@mt.com
Allan Anderson, InspX LLC, 510-226-6686, anderson@pecocontrols.com
Erik Brainard, Anritsu, 847-419-9729, ext. 301, ebrainard@us.anritsu-industry.com
Kevin Jesch, Heat and Control, 800-227-5980, kevinj@heatandcontrol.com