By Lisa Lupo

Although it has been more than eight years since the passage of the Food Safety Modernization Act (FSMA), and July brings the last of the general compliance dates for the seven major rules of FSMA, it seems that it will be a long while — if ever — before we can say that FSMA is finally final.

Not only are there some requirements yet to be written (e.g., Produce Safety Rule water standards), small and very small business compliance dates outstanding, and inadvertent non-compliance to be corrected, FDA Deputy Commissioner for Food Policy and Response Frank Yiannas is transitioning the industry from a food safety modernization act to a food safety modernization culture. Following is an overview of the seven major rules of FSMA, where the industry stands with each, and how it is transitioning to a new era.

PREVENTIVE CONTROLS (PC) RULES (FOR HUMAN AND ANIMAL FOODS). As the foundation of FSMA, the PC rules are the most extensive and complex. As such, there are some sections of the rule which seem to be causing the most difficulty for the industry and are, said The Acheson Group Senior Manager of Food Safety Christopher Snabes, the most common areas of inadvertent non-compliance. These include:

• Incomplete hazard analysis such as failure to perform a process hazard analysis; to identify where any re-work is coming from and going to; and/or to include radiological concerns, EMAs, and recognition of inherently intrinsic pathogens for an ingredient.
• Process PC validation studies are either not being conducted, are incomplete, and/or do not have proof that they were reviewed by the PCQI for acceptability.
• Incomplete or non-existent signatures on the Food Safety Plan and verification records.
• Lack of facility ownership of a supply-chain preventive control (if it is called out for a facility). Currently, if a facility is depending on a supply-chain preventive control, the supply chain needs to be fully vetted and documented at the facility with a written list of approved suppliers, along with explanation and documentation of how approval is attained for the hazard being controlled. FDA is currently exploring a two-tier inspection system whereby a corporate headquarters can maintain the supply-chain system in place for all facilities under its umbrella.
• Lack of a recall plan. Facilities often either have no plan at all, have only an incomplete plan, or the business assumes that the traceability SOP is sufficient for a recall plan. (It isn’t.) Additionally, too many corporate headquarters assume the corporate recall plan is sufficient for every facility under its umbrella. Currently, each facility is required to have a recall plan in place. FDA is currently exploring a two-tier inspection system for this as well, whereby a corporate headquarters can have a recall plan in place for all facilities under its umbrella.
• Failure to practice/enforce GMPs. This is particularly challenging for facilities with high employee turnover rates.
• Failure to properly treat the human food to animal food stream. Any human food destined for animal food use needs to be properly identified, handled, and stored.
• Lack of or incomplete letters of assurance: Companies that are not controlling a known hazard and are depending on a downstream entity to control the hazard with a preventive control often have few to no written letters of assurance between the parties detailing this step.

With such failures being seen across a number of industry establishments, it is critical that facilities assess their practices and processes for compliance. To do so, Snabes said, “Take a holistic look at the entire plan, facility, and food safety culture. Don’t depend on an audit checklist; don’t depend on a single food safety scheme auditor to make sure you meet all aspects of FSMA.”

Rather, Snabes recommended that facilities have a trained, well-engaged preventive controls, food defense, or FSVP Qualified Individual (PCQI, FDQI, FSVP QI) or trained produce grower who works with an independent consultant. Together, they should review your facility and plans, openly work with your entire food safety team, and look for opportunities where the facility can improve and close the gaps of compliance.

The PC Rules are not the only ones for which (often inadvertent) non-compliance is being seen:

• Sanitary Transport. “We have seen some companies under the Human and/or Animal Food Preventive Controls Rule not fully complying with or understanding what they need to do with the Sanitary Transport Rule,” Snabes said. This is evident in the documentation of training and understanding of which party is responsible for the training.
• Produce Safety. As the Produce Safety Rule (PSR) had the compliance dates for the agricultural water provisions delayed until 2022 for the largest companies, this rule remains incomplete. However, Snabes said, “It is encouraging to note that several leafy green producers are taking proactive steps now, and not waiting for the compliance date to go into effect, to produce safe crops.”

From the regulatory side of the picture, PSR inspections began this spring, largely conducted by the states, said FDA Deputy Commissioner for Food Policy and Response Frank Yiannas in the Town Hall presentation at the Food Safety Summit in May. He also noted that FDA is taking a hard look at the water standards and is working with trade groups on developing these. Both of the recent romaine lettuce outbreaks had a water connection, he said, adding, “This water connection is very important. I can’t overstate the importance of water.”

• Intentional Adulteration (IA). With general compliance to the IA Rule due in July, large businesses are expected to have a written Food Defense Plan in place along with other requirements of the rule. However, this is currently the most incomplete rule by the industry, Snabes said. Part of the reason is that many companies are waiting for FDA to publish the third and final guidance document for the rule; additionally, the vulnerability assessment training has just been released, which many companies need to have a Food Defense Qualified Individual — as required by the rule.

Because of this lack of readiness, FDA received numerous letters asking for an extension on compliance, Yiannas said. However, FDA decided to keep the date as set, he said, but the agency will hold off on IA inspections until Spring 2020, and inspectors will educate while they inspect.

• FSVP. Although most food companies are aware of FSMA, many importers who do not process food still need a Foreign Supply Verification Plan. “From my experience,” Snabes said, “This is one of the most overlooked rules.”

And it is an area on which FDA is focusing. FSVP inspections began in June 2017 with almost 1,500 conducted to date, and, Yiannas said, “We will do more FSVP inspections this year than last.”

• Accredited Third-Party Certification. Certification bodies are continuing to be accredited under FDA’s voluntary Accredited Third-Party Certification Program. This accreditation means that the companies have been given the authority to conduct food safety audits and issue food and facility certifications for foreign food facilities. The certifications can establish eligibility for participation in the Voluntary Qualified Importer Program (VQIP), which offers expedited review and entry of food, and fulfills the “rare and specific circumstances” under which FDA may require that an imported product be certified to prevent a potentially harmful food from entering the U.S.

FSMA FINAL? While various aspects of FSMA may be complete or nearing completion, FSMA compliance can never really be “finally final” in a food facility. As Snabes explained, “FSMA specifically states that the plans, and all moving parts, need to be reviewed and updated when needed, and all verification records need to be reviewed appropriately by the appropriate trained individual within the proper time frame (usually within seven days).”

Additionally, Yiannas has made it clear that the modernization of food safety will be an ongoing endeavor. “For me,” he said, “food safety modernization was never an act that was passed in 2011, it was a vision. It’s the idea that together we have to continue to improve our food safety approach.”

Additional evidence of this is FDA’s current focus, led by Yiannas, on the Blueprint for a New Era of Smarter Food Safety which, he said, “Will continue to be FSMA-based, but has to be technology enabled.” The world is becoming so digitized and food safety needs to keep pace with that. Digitization won’t solve everything, but it will help, he said. While blockchain, which Yiannas spearheaded during his previous global food safety role at Walmart, is likely to be at least an option in FDA’s traceability initiatives, he also noted the potential application of artificial intelligence (AI), for which a pilot is currently being conducted on imports. The goal is to determine if the system can be augmented to learn dynamically and be a predictive tool — which would then be shared with the industry.

Seeing the lack of traceability as the “Achilles heel of the industry,” Yiannas said, “We will advance track and trace of foods in a significant way.” But, he added, “It’s transparency we’re after, not just traceability.”

The author is Editor of QA magazine. She can be reached at llupo@gie.net.

Photos by Jake Gravbrot

Tree nut consumption has seen significant growth in the U.S. over the last five decades, with overall consumption nearly tripling since 1970. While hazelnut consumption has grown as well, according to USDA figures (https://bit.ly/2Qzw7sk), its percentage of the pot held steady at 0.04%. Additionally, 70% of the world’s hazelnuts are grown in Turkey, with less than 5.0% grown in the U.S. — 99% of which are grown in Oregon’s Willamette Valley, which lies between the mountains and the ocean in the heart of the 45th parallel. Hazelnut trees thrive across the 45th parallel — halfway between the North Pole and the equator — because of the climate. But, hazelnuts (or filberts) are little known or consumed in the U.S. In fact, the vast majority of Oregon’s production is either exported or used as an ingredient in other food products. The Hazelnut Growers of Oregon (HGO), a cooperative of more than 180 growers, is working to change all that.

“There is so much that North America doesn’t know about us,” said HGO Quality Assurance Manager Mark Clute. “We need to introduce them to this nut.”

The goal, said Plant Manager Jason Costa, is to make it a more common household nut. “How do we make it the next almond? We’re basically creating a niche.” And because all HGO’s profits go to its growers, he said, its success provides them with more money, which can then be put into their businesses. “It becomes a cycle.”

A few of the “introductory” things consumers should know is that hazelnuts are a good source of protein, omega-6 and omega-9 fatty acids, Vitamin E, and antioxidants. Research also has shown that the nuts can lower cholesterol levels. As U.S. consumers become more aware of such benefits, and travelers experience the more common uses of this nut in Europe and Asia, the industry is seeing an increase in demand.

BUILDING AN INDUSTRY. HGO addressed this growth potential through its 2016 merger with Wilco and a new facility in Donald, Ore. The merger gave Wilco a growing agriculture business to diversify its grower services and provided HGO with a larger infrastructure and ability to build a new plant with extensive food safety and quality features designed into the facility.

HGO could have done it the old “pack and ship” way with no pasteurization and basic quality and safety checks, then sold it all in bulk to China at a low price (as do up to 70% of Oregon’s producers). “That would be the easy thing to do, but it’s not the right thing to do for the growers,” Clute said. “We’re developing a domestic market for value-added products. We want to create a stability that’s never been there.” HGO does sell some of its product overseas in bulk, but 60% is sold in the U.S., as compared with less than 25% of most others, and it intends to continue to increase this percentage.

HGO had had a plant in Cornelius, Ore., but it was essentially a packing shed, Costa said, and the company had some high-reaching goals: to be SQF level 3 and be the largest value-added hazelnut processor in America. “If we were going to target SQF3, this wasn’t sufficient; we’d need to build a new facility,” he said. “And the old way of processing wasn’t going to be the new way.”

Thus, the new way of doing business came in the form of a 120,000-square-foot facility with a 30,000-square-foot warehouse — with separate raw and ready-to-eat sections. The plant employs 70 to 80 workers during harvest and about 40 year-round.

One of the key aspects of the new facility is its steam pasteurization system, which provides a kill step in hazelnut processing without the use of chemicals. HGO is the only Pacific Northwest hazelnut processor to have an onsite steam pasteurization system. The 200°F dry, saturated steam provides a natural process for a five-log kill of Salmonella and other pathogens. All hazelnuts intended to be sold domestically without further processing are required to be pasteurized, but most conventional processors use propylene oxide (PPO), a toxic gas approved by the FDA to kill bacteria.

The HGO team also attained its goal to be SQF Level 3 certified — in only nine months. As a complete GFSI system, which adds quality on top of the food safety requirements, he said, it took thousands and thousands of hours in development, training, tool and signage set-up, and documentation. The plant now has three SQF practitioners on staff, and all managers are HACCP and PCQI trained. The SQF3 certification also makes HGO one of only 12% of U.S. manufacturers certified to that level.

HGO’s focus on growth can only come as a relief to the growers whose orchards were in peril not long ago (and in some areas still are) from the arrival of eastern filbert blight (EFB). The blight is native to eastern North America, but faced no resistance in the Barcelona and Ennis hazelnut varieties common to Oregon. As explained by Oregon State University, “The blight takes its time killing a tree. It takes seven to 10 discouraging years for the cankers to spread and the leaves to start turning brown, a sure sign of infection. Through management practices like pruning and appropriate spraying, growers can mitigate the disease but not stop it.” It was two OSU professors who literally saved the region — and, thereby, the U.S. hazelnut industry — from the disease which was pockmarking its way through Oregon orchards. After walking tree to tree through orchards and taking meticulous notes, Shawn Mehlenbacher, a professor in OSU’s Hazelnut Breeding program, began his work to identify sources of EFB resistance and develop new varieties.

According to OSU, it takes 17 years from the time two parents are crossed to the release of a new variety. The most promising strain has been Jefferson, developed in 2009 as an offspring of Gasaway developed by OSU Professor Emeritus Maxine Thompson. Many Oregon growers are gradually replacing their EFB-susceptible Barcelona and Ennis varieties with Jefferson, but there is a four- to five-year period between a tree’s planting and the first harvest.

The new facility and EFB-resistant hazel tree varieties also enable HGO to meet its value-added goals and begin to penetrate retail outlets as the industry had not been able to do. “In the past, the industry wasn’t able to guarantee product to U.S. retailers, so it did not get contracts,” Costa said. But now HGO has contracts with Safeway and Albertsons and is talking with other retailers.

“As we built the plant, we were really building an industry that wasn’t there,” Clute said. “There’s an excitement to what we’re doing, but definitely an underlying sense of responsibility to the growers.”

IN THE ORCHARD. In addition to touring HGO’s new facility, QA visited with one of its growers: the fifth-generation orchards of agronomists Tim, Tom, and Kevin Aman in Mount Angel, Ore. The original 10-acre family orchard was planted in 1968 and has expanded since then to 75 acres — all of which is managed by the family. “When you plant an orchard, it’s a heritage,” Tim said.

Hazelnut trees are planted in rows. Although its natural inclination is to form as a bush, commercial growers prune the plants as they grow, removing the sucker branches with chemical or strangling methods. The trees grow 20 to 25 feet high with canopies that form an “umbrella,” keeping the orchard floor clean and vegetation free. It is a “tree that never sleeps,” Tim said, explaining that the trees bloom in the winter, and the pollen released from their catkins floats on the wind to the new growth of neighboring trees, where the hazelnuts begin to form. The nuts ripen in about eight months, but are left to fall naturally out of the husk to the ground. A hazelnut tree can continue producing nuts into its 80s.

Until the 1970s, hazelnuts were harvested by hand, Tim said. Then equipment was developed that “sweeps” the nuts on the orchard floor into compact rows between the trees, after which — much like the pickers that scoop up golf balls at a driving range — a harvester “vacuums” up the nuts, incorporating an aspirator which blows off extraneous material. The nuts ride up a conveyor to be deposited into the nut cart on the back of the harvester. Once full, the nuts are offloaded into a box to be trucked to an HGO wash plant.

To facilitate harvesting, Tim said, “We keep our floors very clean and flat.” The Amans follow the philosophy of “pick early; pick often,” generally harvesting each orchard twice, picking a third time if price and weather are good. If it gets too muddy, the nuts will pick up too much moisture, and it will be difficult to run the harvester. In good conditions, the harvester can run at 6.5 miles per hour. Five people work the Aman acres: two run the sweeper, two the harvester, and one boxes and ships. The brothers also develop and sell 80,000-100,000 tree starts annually and manage 330,000 acres of hazelnut trees for some owners who bought their tree starts. The starts may be derived from the pruned branches of the trees or started from a tissue culture, enabling the release of new varieties.

With hazelnuts’ short harvest window, October can be extremely stressful for growers, as they spend 11 months working toward that one month of harvest. As such, HGO decided to increase its liaising efforts this year, hiring Grower Relations Manager Ryan Flaherty in July. Having worked previously as the grower services manager for Diamond Foods and on the production side in a vineyard, Flaherty brought the experience and expertise needed for the job, for which he said, “I am the voice to the growers from the cooperative and the voice from the growers to the cooperative.” With his primary focus being to ensure the lines of communication are always open, Flaherty said, “Making sure that everything runs flawlessly is helpful for both the plant and the growers.”

FOOD SAFETY AND QUALITY. Such communication and interaction also are critical because hazelnut food safety and quality begin at harvest and continue through to retail:

1. Foreign object detection. Because hazelnuts are harvested from the ground and even shell fragments are considered by FDA to be foreign objects, foreign object detection and rejection begins at the orchard, and the hazelnuts undergo seven methods of detection from field to distribution: mechanical, aspiration, destoners, Xray, laser machines/cameras, metal detectors, and the human eye.
2. Wash plant. From the orchard, the nuts go to one of 12 wash plants, where they are cleaned, tested, and dried to 9% moisture.
3. At the plant. Following a strict chain of custody, the nuts are trucked to the HGO processing plant where they are tested for moisture, and “any defects that might cause alarm.” Samples are sent to a third-party lab to ensure against any HGO conflict of interest. “Most processors test internally,” Clute said. “But we want to ensure we have the full trust of our grower-base that there is no conflict.” The nuts are weighed, and gently off-loaded into underground pits, with each lot coded for traceability, then run through an aspirator and destoner in the second round of foreign object detection and rejection. Once cleaned, the nuts are conveyed up into a silo, each of which holds 2.5 million pounds and has a specially designed chute which allows them to fall gently so as to not incur damage. The nuts remain in the silo until they are scheduled for production.
4. If for export. From the silos, the nuts will be conveyed into the plant where they are electronically sorted by size. Those of large and jumbo varieties are generally set for export, so they are run through USDA inspection guidelines, certified, then packaged in yellow bulk bags. When Salmonella was found in exported hazelnuts in 2017, HGO was not implicated, so its yellow bags have become indicative of high quality, Clute said.
5. If domestic. The small to medium varieties, intended for domestic use, go from sizing to shelling. Some shelled hazelnuts are distributed domestically, but most U.S. product is shelled. X-ray machines are placed before the cracking equipment, and aspirators placed after it to separate out the shells. All equipment is stainless steel for cleanability, and all conveyances enclosed to contain the dust. The resulting nuts (or kernels, as the fruit is more accurately termed) are graded and separated into six sizes; run through two sets of laser sorters to detect and reject any with defects, mold, etc.; then spread on open conveyors for visual quality checks. A USDA inspection also is conducted at this point, by a USDA inspector or Customer Assisted Inspection Program (CAIP)-certified person. “We also conduct our own QC testing — and our checks are much more stringent than USDA requires,” Clute said.
6. Steam pasteurization. The nuts are transferred to a stainless-steel tote for pasteurizing. The totes have a grated plate near the bottom to allow for air circulation and enable pathogenic kill.

FROM RAW TO RTE. At this point, the nuts pass from raw to ready to eat (RTE). “This side of the wall is unpasteurized. Once it passes through the wall, we’re in the RTE area,” Costa said. Samples are again tested with HGO operating under a positive release program, so nothing is shipped until test results are final and a COA generated.

The RTE area is positively pressurized, so air filters out of the room, and external, potentially contaminated, air is not brought inward. After pasteurizing and testing, the nuts can go in several directions:

• Roasting. The nuts are leveled out and roasted at various times and temperatures depending on the desired roast level. They are then cooled, loaded into totes, and queued for packaging. In 2018, HGO conducted a quality test of roaster settings for flavor and texture. After evaluating more than 50 combinations, HGO adjusted its roaster to consistently produce its Gold Standard.
• Slicing or dicing. The pieced nuts may be sent to foodservice or manufacturing facilities sliced or diced to their standards, or they may be packaged for retail.
• Butters or pastes. The HGO plant is a dry facility except for a single segregated room in which industrial butters and pastes are made and packaged for foodservice or manufacturing. With its coated floors and trench drains, it is the only area in which water is used for processing or cleaning, and access is limited.
• Retail. HGO also packages whole roasted and natural kernels for retail. The cooperative is the first to develop and sell a retail line of seasoned hazelnuts; its Oregon Orchard line includes Cinnamon Sugar, Rosemary, Himalayan Salt, Sweet & Spicy BBQ, and Southwest Chili Pepper, and a variety of chocolate-covered nuts.
• Bulk packaging. The pasteurized nuts — sliced, diced, or whole — also may go direct to 25-pound cases or bulk packaging.

With capabilities to store and process all available tonnage into whole, sliced, diced, pastes, butters, and other formats, HGO is well on its way to re-introducing and creating a new niche for this old-but-new-again nut. Whether it will become the next almond is yet to be seen, but if the current pace is retained, the hazelnut will, at least, become known for more than Nutella and flavored coffees.

The author is Editor of QA magazine. She can be reached at llupo@gie.net.

In today’s world of instantaneous internet communication, a single consumer finding a foreign object in your food product can wreak havoc with your brand and reputation. Should the finding be considered potentially extensive, you could be facing a massive recall; and if a consumer is injured by the object, you could easily have a lawsuit on your hands.

According to Food Safety Management, A Practical Guide for the Food Industry (Elsevier, 2014) chapter authors Gilles Demaurex and Laurent Salle, a foreign object is “Any extraneous material in food not intended to be there or be consumed, e.g., field matter (stones, wood, metal) even if a part of the food itself (e.g., bone, shell, pits, stems, etc.); inadvertent processing residues (glass, metal fragments, plastic, personal objects, etc.); intentional contamination (e.g., employee sabotage); miscellaneous particulates and fragments.”

Intrusion can occur at any point along the food chain, and even if the foreign matter is not a hazard, “their finding in food may be indicative of poor hygienic practice and a distressing event for consumers,” the authors state.

Foreign material was the top reason for USDA recalls in the first two quarters of 2019, according to the Stericycle Recall Index quarterly data. While foreign material tends to place somewhere below allergens in FDA recalls, it was the top reason for recalls in at least one quarter in 2018. Additionally, guarding against foreign objects in food is mandated by U.S. regulation. As stated in the Food Code (21cfr110.80[b]8), “Effective measures shall be taken to protect against the inclusion of metal or other extraneous material in food. Compliance with this requirement may be accomplished by using sieves, traps, magnets, electronic metal detectors, or other suitable effective means.”

Because there are many potential stages at which foreign objects can get into food from the field to packaging, it can be advantageous to use a combination of methods to ensure against anything reaching the consumer.

One significant example is that of the Hazelnut Growers of Oregon’s seven methods of detection from field to distribution (Cover Profile. Hazelnuts: The Next Almond?).

DETECTION OPTIONS. So what methods are being used and what is best for your facility? While the specifics will depend greatly on the product, a wide range of controls can be customized to effectively control some of the risks of most processes. North Dakota State University graduate student Daniel Richard Dumas details common methods in his paper, “Foreign Material Identification and Removal in the Food Safety Industry”:

• Filtering and Screening. Screens can separate solid foreign materials from liquid products or separate solid items based on size. Some require mechanical agitation or movement to allow for more effective product flow, but the basic method is that of creating a physical barrier to remove foreign material. The designs of these systems must account for known hazards and production speed. While generally low cost and high efficacy, screens and filters must be regularly inspected and maintained, as they eventually wear and break, potentially introducing metal into the system themselves.
• Magnets. As the most basic level of metal removal used in the food industry, magnets can remove ferrous metals and are often used to identify upstream equipment failure and help protect product from machinery breakdown. Magnets can be ceramic for general use; Alnico magnets for high temperature processes that can exceed 400°F; and rare earth magnets which are the strongest and most effective, as they can remove magnetic stainless metals as well as weakly magnetic materials, but can vary in effectiveness at extreme temperature. Also used at safety control points downstream, magnets can remove many particles, including metal dust and pieces too small for identification by other methods, with little maintenance, cost, or product loss. A major disadvantage is that they can only remove magnetic materials. Other metals will not be removed from the product flow.
• Optical/Laser Sorters. Optical and laser sorters can be used to remove foreign materials or items that do not meet quality parameters. These cameras or lasers can detect surface-level characteristics including color, shape, and moisture, and can distinguish biological characteristics. Once identified, an air jet or mechanical device can remove the nonconforming material. These sorters also can improve product quality through removal of discolored, damaged, and misshapen product.
• Metal Detectors. Many facilities implement metal detectors as a final check of packaged product. Advances have continued with improvements in electronics, computers, and sensors, with metal detectors able to detect metal with high efficiency. The basic operation occurs by one of two methods: a balanced coil system using electromagnetic induction or a magnetic field system. These systems can be limited due to the material that is being tested and the type of metal that is targeted. In materials that have higher conductivity levels, it can be more difficult to identify metal using balanced coil systems, while magnetic field detectors can identify magnetic metal inside aluminum cans. Basic detectors are common in the food industry, and industrial versions continually advance to improve the level of detection and accuracy; remain cost-effective while reducing false trips and product loss; increase operation in environmental conditions such as heat, vibration, and moisture; and reduce cross-signal effects from other machinery or communication equipment. However, if a food process has other inherent risks or the facility wishes to identify non-metal foreign materials, additional or alternative methods will need to be implemented.
• X-ray. X-rays use short wavelengths that can easily penetrate packaging materials and enable detection of foreign materials in packaged products. These detectors use high-powered energy sources and imaging systems to produce images of the material and identify foreign materials. The ability of X-ray devices to detect even small differences between food and foreign material, including metal, can be more effective than that of metal detectors. Additionally, X-ray can detect and identify plastic and glass, and can operate at relatively efficient processing speed. The ability of X-ray to detect foreign materials is continually improving and allowing for more sensitive identification. Some developments, such as the use of dark field imaging, are being attempted to increase their ability to detect organic materials on a smaller scale. X-ray use is currently limited by the cost of the units and maintenance, processing speed, and safety aspects some associate with the radiation. These areas are being improved as advancements are made and costs are reduced.
• Visual Detection. Visual methods include various wavelength or imaging sources. A basic optical method of reflectance measurements uses the visible light spectrum. A light source is used to create a reflectance off the surface of the product and enable differences between food and foreign material to be detected. This method has been shown to help in the removal of sticks, stems, or rocks. Some limiting factors are the heat caused by the illumination and its ability to detect only surface materials, which make testing of sensitive, thick, or bulk materials impractical. Other visual detection methods include testing with near-infrared wavelengths of light which use the absorption ratios of molecular bonds to differentiate the different materials within a sample, and ultraviolet wavelengths using radiation of the product and the signal that various materials will reradiate. These methods vary with efficiency depending on the food product and target foreign material.
• Hyperspectral Imaging. Hyperspectral imaging combines spectroscopy and digital imaging computers. The visible spectrums and multiple-spectral wavelengths can be customized to detect foreign bodies that do not share similar patterns and to identify quality parameters such as bruised produce and ripeness. Due to its speed and wavelengths that can be incorporated, there are many other potential avenues for which this method may be helpful in the future, however additional research into computations and filters will be needed.
• Thermal Imaging. Thermal imaging typically occurs in one of two ways: If the product and/or foreign material gives off heat at a suitable range, infrared energy can be applied and the difference between the two used to create a digital image, in which foreign material can be detected. The other method is the application of heat in a short burst, with the distance into which the heat is able to penetrate measured. This creates a contrast that enables the detector to identify the foreign material. These methods may not be suitable for all food products but may be able to be adapted to processes where heating or heat treatment of products is already occurring.

EQUIPMENT SELECTION. With such options, how does the food facility determine the best equipment by which to detect and prevent foreign object intrusion? As Demaurex and Salle explain in their chapter, several aspects will determine the choice of detection equipment, such as contamination risk, packaging and product characteristics, line layout, environmental conditions, or process speed. But, the authors add, no technology can detect all types of physical hazards. Following is the authors’ recommended selection process (with updates for relevance):

1. Determine the source and nature of physical hazards along the production line. Your HACCP (and FSMA Preventive Controls) plans, plant records, maintenance reports, and consumer complaints can all be helpful.
2. Apply all possible prevention measures to reduce foreign bodies (GMPs, hygienic design, best practices, sorters) according to the principle “first prevent, then detect.” Working on the hygienic design of equipment with the supplier will deliver better results than simply using a metal detector to detect equipment issues.
3. Select the most appropriate technology and location according to the type of product and the packaging. This might require modifications to the line layout. For example, the installation of a metal detector before filling instead of an X-ray device on packed product might require a redesign of the filler infeed or height of the building. Keep in mind that metal detectors and X-ray technology are more effective on small products (e.g. bulk before filling or individual packaging), than on grouped products or shipping cases.
4. Select a short list of potential suppliers, based on criteria such as best detection performance for the specific application; ability to provide turnkey solutions (infeed, product handling, reject, etc.); support level available; and coherence with other equipment brands already installed in the plant to optimize maintenance and operational cost.
5. Build a user requirement specification (URS) and send this with products and contaminated test samples to the preselected suppliers for detection capabilities trials. It is recommended that you build your own sets of contaminated test samples to ensure a fair performance benchmarking between suppliers (with spheres and real fragments). Indicate the testing procedures to follow. Ideally you should be present during the trials.
6. Build a selection matrix to make the choice, assessing the key elements of the URS. For example, equipment with slightly inferior performance might be chosen for its ease of cleaning and hygienic design.
7. Conduct a detailed performance qualification with the selected supplier to assess and document the detection capabilities for each relevant foreign body type.

Regardless of the number and types of methods and equipment implemented, many food facilities choose to supplement them with human visual quality and safety checks. While this cannot detect internal, imbedded material, the idiomatic phrase “an extra set of eyes” can be the very difference between customer satisfaction and a federal inspector knocking at your door.

Stored product insects (SPI) can contaminate stored foods; cause increases in temperature, moisture, fatty acids, uric acids, and molds; and decrease weight, all of which can result in economic loss, regulatory penalties, customer rejection, and adverse consumer reaction. But the insects are often difficult to detect until populations build because of their small size and ability to live, feed, and breed within the product itself or in small cracks and crevices. They also are difficult to control because of their high reproduction rates, mobility, and ability to survive on very small amounts of food, such as residues and fines in trucks, bins, and equipment. But new technologies are continually being developed for SPI detection, prevention, and elimination. Following are some of the most prominent:

DIGITAL DETECTION. This technology began with rodents and is now moving into the realm of SPIs. Electronic pest detection systems come in many forms, but the basic premise is that of detection, notification, and data collection. A sensor installed in a trap is triggered by touch or movement. A signal is relayed to the system hub, and a capture alert is sent to a specified device (e.g., of the pest control technician or plant associate). Capture data is stored and provides a historical map of pest activity.

“New technology always starts somewhere, and when that technology starts to provide value, we start to see it branch out,” said Bayer Environmental Science Technical Service Lead Joe Barile. Such is the case with digital monitoring. It began with rodents because the hardware and technology were there. But now, he said, “The ice has been broken.” When he was introduced to the technology, “one of the first things I thought of was stored product pests because of the complexity of gathering information,” Barile said. SPI monitoring requires a lot of work in putting the traps in place, and checking and swapping them out regularly.

When digital monitoring is combined with automatic visual monitoring, the potential for detection is even greater. These sticky traps have cameras built into them which take photos at specified intervals. Rather than simply denoting the entry of a pest, these traps help in insect identification and a count of the number collected.

Analysis of captured insects also can provide a “heat map” of insect counts in the facility, indicating where hot spots are and inspection time should be spent. “You don’t have to count and analyze it yourself, you can know where to focus,” said Insects Limited Vice President Tom Mueller. For example, SPIs may be gathering where product has been spilled, a machine that was turned off has infested debris, etc.

Such information not only helps track the source of the problem, it can shorten the time of action, enabling proactive prevention rather than reaction, and reducing recalls, Barile said. “We can intercept problems if we have this type of information out there.” Monitoring also frees up time for pest control technicians. Rather than spending time manually counting insect species in traps, they can be inspecting the facility, making recommendations for problem areas to provide for prevention rather than reaction.

However, FSMA’s preventive controls requirements can pose a downside for digital monitoring. “If you are aware of insect presence, when and how you address it could take you out of compliance,” said Trece Market Manager James Miller. For example, if a technician determines that a count of insects of which he is notified is not significant, he could decide to wait until his next service. But a federal inspector or third-party auditor could see it differently. This same issue exists with rodent digital monitoring, he said, “And we won’t find out until they start enforcement.”

False positives also can cause issues, as technicians will not know a notification is not a capture until they arrive at the facility. But even with the negatives, Miller sees the compiled data as improving integrated pest management (IPM) with the value being “how well you use that to better your food safety program to reduce your risk of stored product pest infestation.”

Knowledge attained through the monitoring can be further enhanced with smartphone apps that are being developed for pest identification, Mueller said. “You won’t have to gather a specimen and send it to an entomologist as much as in the past.” Integrating these with other technologies, such as routing, can enable a technician to get updates and review a facility’s history and be completely informed at each visit.

Though still in testing, camera monitors, which are implanted into traps, could be very beneficial, but the cost may be prohibitive. However, in particularly sensitive areas or high-risk applications, the value could be worthwhile, said USDA ARS Research Entomologist Jim Campbell. Advantages would include the ability to monitor hard-to-reach traps, get data in near real-time, providing for more proactive prevention, and being able to place traps in hard-to-reach areas without worrying about worker safety in accessing them.

INSECT LURES. To enable wider insect capture, multi-lure traps are becoming more prevalent and verstile as technology enabling attraction of up to 32 species of stored product beetles has become almost mainstream, Miller said.

Although multi-lure traps are not really new, more people are asking for and talking about them, Mueller said. Pre-loaded traps also are becoming more available, he said. A technician or facility worker need simply open the trap and set it out. Care needs to be taken with these, however, because there can be a difference in quality and/or quantity of lure. “So with ease of use, there may be a little sacrifice in quality.”

Mating disruption, though currently available only for the Indian Meal Moth, is another area to which food processors are becoming more acclimated, Mueller said, noting that this used to be seen as an added cost. But, with FSMA’s increased prevention, facilities have become more willing to spend money in this area, as the pheromone attractants are being seen as ways to monitor, predict, and prevent future infestations.

THE FUTURE. While system-wide linkage and cost are current barriers to complete integration of some technologies, such as digital detection, the applications of these and the traps and lures are continually evolving. And that evolution is benefitting food processing facilities by enabling the pest control industry to provide the service that is promised. Rather than spending time at a facility checking traps and counting insects to gather data, the data is collected for them, and they can arrive at the facility ready to provide further inspection or recommendations, handle issues, or proactively increase prevention.

Additionally, Campbell expects that digital technology will continue getting smaller and cheaper, bringing the cost of implementation down as well. “I think it’s really an exciting technology and, eventually, will be the way we go,” he said.

Expressing similar sentiments, Barile said, “I’m really excited about this. It’s long in coming. There is some resistance — but remember when cell phones first came in,” he said. “After 40 years in the business talking about IPM, I think these will validate the pest control industry as being IPM-focused. Stored product pest application is a slam dunk; it fits the FSMA support requirements that we’re all involved with.”

The author is Editor of QA magazine. She can be reached at llupo@gie.net.