Advanced control solutions enable processors to satisfy regulatory demands while gaining a competitive edge.
Last year was full of all sorts of bad news from the FDA: product recalls due to E. coli, Salmonella and other nasty pathogens; problems with allergen contaminations; and labels that failed to list potential allergens, missed some ingredients altogether or promised certain levels of vitamins or minerals that simply weren’t up to the level specified on the label. While these problems screamed for attention, consumers continued to demand more creative, tantalizing foods and drinks. Food engineers had to meet these challenges with fewer resources, while tracking the ingredients used, the processing steps taken and the customers who ordered the products.
Meeting these demands by manually recording data from various “islands of automation” is becoming more and more difficult. As governing bodies increase the information they want and decrease the time allotted to procure it, manual methods will no longer cut it. It’s time to start thinking about plant-wide advanced process control.
Supplements: On guard!
According to David Haase, vice president, operations for Wild Flavors North America division, 2008 marks the first year of compliance to the new FDA 21 CFR Part 111 for dietary supplements (nutraceuticals). Subpart E requires the establishment of a production and process control system that tracks and traces all ingredients and provides information guaranteeing all ingredients were tested. The process and production control system must ensure that a dietary supplement is manufactured, packaged, labeled and held in a manner that will ensure the quality of the supplement and that the product is packaged and labeled as specified in the manufacturing record. Quality control in all phases of production, packaging and storage is mandated.
While the Bioterrorism Act of 2002 began to implement the tracking of animal and plant shipments, the European Commission’s EC 178 of the same year is more stringent, and food suppliers to, and in, the European Union are still working on meeting the demands of this ruling. Iconics President Russ Agrusa describes the ruling as giving the customer in the grocery store the ability to scan a bar code label on a meat package to discover where the cow was raised, what grains it ate, where it was slaughtered and what processing and packaging steps were used. The UK General Food Regulations 2004 enforces EC 178 and especially Article 19, which can impose fines and jail terms on processors who flagrantly disregard this set of food safety laws.
Too much to do, too little help
No question about it, tracking and tracing is on everybody’s mind-and at every stage of the product’s lifetime-literally from farm to fork.
“As the food industry is governed more and more by lot traceability, automation through bar codes and scanning will provide the biggest benefit,” says Ricki Reves, CFO, Juice Harvest Corp. “Product data contained in a bar code will give the entire history of the product from receipt to shipment.” Reves further explains that real time is of the essence for electronic record keeping, noting that “product moves through our facility within 24 hours, so it must be tracked in real time on the plant floor from point of receipt through production to point of shipment of processed product.”
Terry McCorriston, director, CSB-System International, says that along with determining whether a process is safe, it’s essential to have the ability to track and trace throughout the entire plant. He notes that in Canada, federal and provincial governments are providing grant moneys to processors to the tune of $20,000 or so to help them put track and trace systems in place to identify what ingredients from what suppliers were used for a batch, what production line was used, what packaging line, etc.
Brandon Henning, GE Fanuc Intelligent Platforms food and beverage global industry manager, points out four concerns his customers are facing. First, the FDA is implementing stringent requirements on track and trace, not only on finished products, but on raw materials as well. Second, processors are dealing with more product introductions, which means more variation, recipes, packaging, etc., requiring additional agility and flexibility in the manufacturing and packaging processes. Third, retailers are forcing quality back into the supply chain by implementing their own tracking and tracing, which further mitigates risk surrounding product quality and safety. Fourth, the high cost of energy and retailers’ green initiatives are forcing processors to reduce energy usage and cut down on unnecessary packaging.
According to Rusty Steele, food and beverage business development manager, Industrial Solutions Center, Schneider Electric, the challenges are complicated by the fact that processors have to cope with fewer skilled, trained and experienced personnel in key areas such as production, engineering and maintenance. Interviews with nearly two dozen processors attending Food Engineering’s ProcessTechnologyXchange 2007 revealed nearly half of them were concerned about finding qualified engineering help. Steele also says that many processors are facing very high costs per minute or hour of downtime, and one method to help alleviate issues at the process control level is to provide hot switchover (within one scan cycle) to a back-up controller when the primary controller fails.
Pulling the loose ends together
Process control isn’t just limited to CIP, a single batch process or a packaging operation, or other isolated processes such as a bake or pasteurization operation. According to Steve Ryan, director, Proficy process solutions, GE Fanuc, “Process control will pay off in all these areas, and anyone looking at process control should make sure they are including these opportunities as part of their solution.” Ryan adds, “The best cost/benefit ratio can be achieved by an integrated process system that has information integrated into the architecture. It’s about the flexibility in process control.”
Matt Ruth, food and beverage group director at Advanced Automation, says the biggest force in driving information out of an integrated process control system is the lean culture and its methods. “This drive to reduce inventory and waste has created a huge demand for ‘data-driven decisions,’” he says.
These data-driven decisions typically manifest in initial areas such as product traceability, OEE, downtime and waste/yield. Following these in importance, says Ruth, are areas like quality, HACCP, inventory control and integration to the ERP system. This connectivity is predicated by the overall desire to remove paper from the factory floor to gain real-time information without the natural latency of paper-based systems. Some examples include HACCP, CIP reports and product genealogy.
To reach this level of connectivity, Steele advises building a common network with open architecture that will allow the highest number of field devices to communicate at the process control level. Additionally, new tools are available to allow users to design an automation system (hardware and software) around their actual process rather than the typical topology of the control system. This approach will prevent the “add more stuff on over time” mistakes, and provide a unified platform for collaborative control.
The good news is that old, proprietary process control platforms do not necessarily have to be ripped out to fit into today’s connected system. “Manufacturers like Siemens and Rockwell have created platforms to convert old communication technology to Ethernet,” says Ruth. “Once on Ethernet, the sky is the limit from an information and control perspective.”
According to Ryan, batch control provides plenty of bang for the buck, but the industry is asking for simple batch where the control system, historian, etc. are all integrated so processors have full S88 compliance. Beyond batch, technology has shifted, and it is important to focus on combined platforms and multidisciplinary control. “Processors should take advantage of today’s technology that allows a plant-wide architecture. If you try to solve today’s problems with yesterday’s tech, that is not going to work. The new ARC Collaborative Production Systems (CPS) manufacturing model is designed to provide a new, holistic vision for how processors view their manufacturing and production operations, and how those operations can best collaborate with the rest of the enterprise.”
ARC Advisory Group’s CPS model helps combine its collaborative process automation systems (CPAS) and operations management models. According to Craig Resnick, ARC’s research director, “Collaborative production systems prevent assets from being niche islands of information and ensure that all assets are delivering their maximum return on investment.” CPS also eliminates manufacturers’ internal barriers that may exist between, for example, plant floor and IT personnel.
The model provides direction to processors regarding how to apply the principles of collaboration to ensure each production system is contributing valuable relevant information to the other, which enables the best possible management decisions to be made. The CPS architecture provides a real-time, unified perspective to support enterprise-wide operations management, making collaborative manufacturing possible.
According to GE’s Henning, “If a food company actually adopted the CPS model, it would be more beneficial than implementing single-minded tools.” Henning says there is a need to update systems that can be managed and maintained by the new workforce and to take businesses successfully forward for the next 10 to 15 years. It’s no longer very effective to update or maintain a single controller at a time. By and large, businesses are looking to tie their control systems into their business information systems, and the CPS model can help them define a path to success.
An example of a system that seems to fill the description of a CPS is a bakery ingredients system in the UK that automates the entire process. British Bakels (Bicester, UK) has managed to coordinate the elements of its production enterprise into a flexible batch system that orchestrates powder blending, wet production, warehousing, inventory, scheduling and fulfilling orders. Wonderware’s InBatch and other FactorySuite components manage hundreds of raw ingredients and hundreds of finished ingredient products. The batch software produces daily high volumes of products with many differing orders-all without error. According to Nick Luxemburg, engineering manager, “The system that has been installed provides us with major competitive advantage and the ability to deliver right on schedule.”
Statistical process control (SPC) is not a new tool, but many processors don’t make use of it, and when they do, they’re not using it for real-time process monitoring. For example, Eric Samp, Molson Coors brewing QA manager, has seen companies that use SPC for end-of-batch monitoring. Samples are taken upon completion of the batch and sent to the lab, but by the time the chemist has run various tests on it-for example, HPLC (high-performance liquid chromatography)-and accumulated all the data, it’s often a couple of days past the completion of that particular batch. By the time this data is entered into an off-line SPC system, a few more days will have expired. By the time QA has the test results, the product may have already been distributed through the supply chain where consumers may perceive variability at the retail store.
There are other advantages to using an SPC system. According to Richard Morris, Zontec marketing manager, an on-line SPC system can provide manufacturers with proactive results should it discover an up-tick in bacterial contaminants in a food product. This gives a processor immediate information that halts production before a contaminated product can get out the door. With a program like this in place, it’s entirely possible insurance premiums would be lower simply because the risks are demonstrably lower.
Rob Wiersma, Lawson food and beverage industry manager, adds that SPC information can be funneled to those who actually need it. For example, slightly off-spec color or taste is a concern of the QA department, and a CEO or CFO might not want to be bothered about a tiny QA anomaly in an SPC chart on a product’s flavor. However, he or she would certainly want to know the specifics of a contamination. Real-time SPC could narrow the window down to hours of production (not days or months), so the company doesn’t have to recall a month’s worth of food. And if the processor has a “hold” policy, then no contaminated food will have left the premises, keeping CNN’s coverage relegated to the ongoing political debates.
Real-time SPC saves millions
For PepsiAmericas, an outdated quality tracking system lacked the ability to be networked and had ceased to provide the company with a real enterprise-wide solution. As a result, various plants had their own sets of specifications that oftentimes didn’t match that of their PepsiCo parent. A custom-designed software project, PASLAB, was implemented by Zontec. This sophisticated lab testing and quality control system acquires and manages test data from PepsiAmericas’ lab measurement equipment to satisfy its parent company’s strict product testing standards, master specifications, clinistics and regulatory requirements. It features a quality control component centered around Synergy 2000 SPC software, and is currently being used in 21 plants in the US and the Caribbean.
According to PepsiAmerica’s Director of Quality Al Warner, “PASLAB ensures that all facilities are playing on the same field, working from one set of master tables.” The application’s front-end uses Zontec’s S2K DLL and ActiveX toolkits, which allows the system to be flexible, provide an open architecture for the future and interface with various data collection devices using standard off-the-shelf hardware.
The system captures data such as CO2, Brix, Titratable/pH, net content and Total Disolved Solids/Turbidity and transfers it to PepsiAmericas in Rolling Meadows, IL. “It allows us to prove to PepsiCo that we performed a test, what time we performed it and what were the results,” Warner says. He notes that they also incorporated closed-loop logic that prohibits lab technicians from proceeding until the test requirement is completed, thus guaranteeing full traceability. “Now we have a system that presents the data in a format that the beverage industry is familiar with, and can output reports any way we wish,” he adds. “In the first year, we’ve reduced our out-of-standards by 50%, improved our compliance level from 96.3 to 98.5 and reduced waste by more than $2 million a year.”
Borrowing from pharma
In 2001, the FDA had a good idea for its pharmaceutical constituents. Termed process analytical technology (PAT), PAT’s goal is “to understand and control the manufacturing process, which is consistent with our current drug quality system: Quality cannot be tested into products; it should be built-in or should be by design.” A desired goal of the PAT framework is to design and develop processes that can consistently ensure a predefined quality at the end of the manufacturing process. Such procedures would be consistent with the concept of quality by design and could reduce the risks to quality and regulatory concerns while improving efficiency.
Gains in quality, safety and/or efficiency will vary depending on the product. PAT benefits include:
This set of procedures doesn’t have to be owned by the pharma industry; it’s a good description of what any food or beverage processor would like to accomplish.
Maaike Van den Bergh, marketing communications, production & quality for Siemens Belgium said that Siemens SIPAT software (PAT software for pharma) could be used for food applications, but since the software has been available for a little less than a year, the company has not yet marketed it for the food industry. SIPAT is a completely integrated PAT solution and works with chemometrics, analyzers, process equipment and process control systems.
According to Ingrid Maes, senior consultant-innovative technologies at Siemens’ Competence Center pharma (also in Belgium), the University of Copenhagen has begun a PAT Master of Science program for the food, pharma, biotech and general production industries. (More information can be found at www.models.life.ku.dk.) Maes, who was involved in the food industry for 15 years before coming to pharma, had been applying PAT principles during those years, but the process was called in-line process monitoring and control. In the food industry, the majority of process analyzers tend to be NIR (near infra-red) and are used to measure moisture, fat and protein levels. NIR can also be used for milk powder, chocolate, bakery, dressings, meat, starch, beer, soft drinks and other applications. In terms of NIR applications, Maes says the food industry is actually far ahead of pharma.
Umetrics is the software company that worked with Siemens to develop SIPAT, and has also worked with GE and ABB in developing similar products for pharma applications. Its manager of process analytics, Chris McCready, actually thinks the pharma industry is learning from the experiences of the food industry, noting that, “a couple years ago [Sept. 3, 2003], The Wall Street Journal pointed out how poor and inefficient manufacturing processes were in pharma.” Clearly, the time was right for pharma to revamp its manufacturing processes. “Foods have to operate with very small margins and still be profitable,” McCready continues, “and pharma actually found some success stories from food and updated with the latest and greatest technology that was available. PAT in pharma was a design turn on what had been done in food.” According to McCready, some time ago, Frito-Lay used what might be a predecessor of PAT technology to monitor in real time the amount of seasoning applied to potato chips. This closed-loop control system was easily applied to food, but pharma wasn’t typically operating in closed-loop control.
PAT for different reasonsPharma and food production have different needs, but PAT meets them all. According to Gerry Broski, director of marketing for bulk weighing and product inspection at Thermo Fisher Scientific, “In the pharma industry, the value proposition is in saving small batches of very expensive materials, whereas in the food industry, it’s in saving large quantities of relatively inexpensive materials.” Broski says the drivers are different, but the goals are the same: to improve the process and to improve product safety.
While it didn’t affect human food, the recent melamine contamination is a good application for a NIR analyzer. Rick Cash, marketing technology manager for Thermo’s process instruments division, explains that while NIR can also be used as a real-time monitor to measure moisture, protein and fats, it can be used to check for adulterant levels of melamine mixed in with wheat gluten. The company’s Spectra-Quad was set up successfully to recognize melamine’s signature in real time as product passes before it. Molson Coor’s Eric Samp has experimented in a few trial runs with PAT at the brewery. He notes that the real-time nature of PAT for monitoring the ongoing process can let a brewmaster compensate for varying qualities of ingredients and hold the batch to color, flavor and quality. “There are several opportunities to take advantage of PAT. What I think has more value [than SPC] is that you get the feedback immediately, so you can start the root-cause analysis. Was it upstream, was it the actual media that you’re using for fermentation?” Samp says if PAT can be used effectively to control the brewing process, the savings on keeping a large fermentation batch alive are very significant.
While PAT techniques have been applied to food applications in the past under monikers other than PAT, there is no reason not to take a serious look at this technology now. An industrial PhD thesis written by Christian B. Zachariassen in 2007 entitled, “Process Analytical Chemistry and Technology in Pectin Production,” for the University of Copenhagen and sponsored by CP Kelco, provides an extended look at applying PAT in a food application. It can be freely downloaded at www.models.kvl.dk/research/theses/.
Batch technologies, collaborative production systems, real-time SPC and PAT are all tools that can be applied to food and beverage plants today to stay competitive in 2008 and beyond, and to maintain the electronic records necessary to meet the demands of governmental regulators. Can your company afford to overlook the advantages offered by these powerful tools?
For more information:
Rob Wiersma, Lawson,
Matt Ruth, Advanced Automation, 610-458-8700 ext. 274, firstname.lastname@example.org
Brandon Henning, GE,
Steve Ryan, GE, 800-433-2682, email@example.com
Chris McCready, Umetrics,
Richard Morris, Zontec,
513-648-9695 (ext. 204),
Rusty Steele, Schneider Electric,
Russ Agrusa, Iconics,
Rick Cash, Thermo Fisher Scientific, 763-783-2638,
Gerry Broski, Thermo Fisher Scientific, 763-783-2600,
519-579-7272 (ext. 203),
Ingrid Maes, Siemens,
Maaike Van den Bergh,
What is PAT?Process Analytical Technology (PAT) is a system for designing, analyzing and controlling manufacturing through timely measurements (i.e., during processing) of critical quality and performance attributes of raw and in-process materials and processes with the goal of ensuring final product quality.
It is important to note that the term analytical in PAT is viewed broadly to include chemical, physical, microbiological, mathematical and risk analysis conducted in an integrated manner.
There are many current and new tools available that enable scientific, risk-managed pharmaceutical development, manufacture and quality assurance. These tools, when used within a system, can provide effective and efficient means for acquiring information to facilitate process understanding, develop risk-mitigation strategies, achieve continuous improvement and share information and knowledge. In the PAT framework, these tools can be categorized as: