![](/FE/Home/Images/FE-May-cover-08.jpg)
Higher ingredient costs-thanks to conversion en masse of corn to ethanol-and tightening food safety requirements by not only the FDA, but also large retail chains, have food processors caught in a wringer, unable to loosen the tension without the right software tools. But before acquiring these tools, processors need to do a little homework. The assignment includes a familiarization with FDA’s new Food Protection Plan (published in November 2007); a look at Wal-Mart’s new set of food safety standards; and a re-evaluation of HACCP plans, internal process and supply chain.
Government and retail clampdown
After last year’s frightful series of food-borne illness outbreaks, allergen problems and mislabeled foods-and with public pressure riding high-FDA issued a food protection plan that addresses both food safety and food defense for domestic and imported products. In it, FDA requests authority to order food safety recalls and issue additional preventive controls for high-risk foods. FDA’s integrated strategy provides three elements of protection: preventing foodborne contamination, intervening at critical points in the food supply chain and responding rapidly to minimize harm. For food processors to be effective in this scheme, several software tools can be employed, and several are covered in this manufacturing software guide. (See chart beginning on page 34.)Wal-Mart Stores Inc. has become the first nationwide US grocery chain to require its suppliers of private label and other food products such as produce, meat, fish, poultry and RTE foods to have their processing plants certified against one of the internationally recognized Global Food Safety Initiative (GFSI) standards. GFSI standards provide real-time details on where suppliers fall short in food safety on a plant-by-plant basis. According to J.P. Suarez, Wal-Mart senior vice president and chief compliance officer, and a GFSI board member, “We are taking this step to ensure the integrity of our products throughout the entire food supply chain. We encourage other US retailers to follow our lead and to also endorse these standards.”
![](/FE/Home/Images/FEX0508softwareguideCS1.jpg)
A production line at Arla Foods’ Brabrand dairy has seen a 5% increase
in OEE over a period of two years since management began looking at the
process. Wonderware’s suite of products was used, including DT Analyst,
Production Events and Enterprise Integration. Source: Wonderware.
Start at the beginning
To meet all these safety standards, processors should begin, of course, with a good HACCP (hazard analysis critical control points) plan. According to Karen Theel, Oracle senior director of process manufacturing, there are several tools in today’s software systems to assist in the process of establishing quality assurance procedures, specifications with critical control points and critical limits, test methods for CCP, environmental monitoring and corrective actions. Workflow can be used proactively to route quality samples, results and disposition activities to the appropriate personnel/approvers.Johann Heydenrych, director of industry solutions at itelligence Group, has 25 years of experience in food science and microbiology. He says that while all these software tools are available, several folks are still struggling with the concept of HACCP. “There is a misunderstanding about what food safety and HACCP are all about, especially with CCPs.” Take, for instance, the refrigerated truck driver who accidentally shuts down the refrigeration unit over a lunch break. While the truck’s temperature may be automatically monitored by a wireless temperature monitor, the more likely scenario is no automated monitoring at all. Will the driver make note of any wide temperature variations in his log? The same could apply to a freezer failure at a plant, where all too often there is no automated monitoring-just an analog thermometer and a person with a clipboard who is expected to notate an out-of-temperature reading.
Getting a HACCP plan pulled together can be an information nightmare for processors, says Kathleen Ley, co-owner of Norback, Ley & Assocs. Ley, whose consulting company assists processors with HACCP plans and provides HACCP planning software, says HACCP entails much more than temperatures at key process points. Processors need to know how hazards flow through the system, what they are, what organisms such as E. coli may be present, what ingredients can be affected, what can go wrong with products such as milk or pork, and a host of other things. She adds that processors who sell to McDonalds must have a HACCP plan in place or simply not sell to McDonalds. And processors should expect to be audited by McDonalds. In fact, today most customers are in the know and will ask to see a HACCP plan before they buy food and beverage products. When asked if there is still a lot of HACCP teaching that needs to be done, Ley replies, “I wish I could say no. We just taught a class of 35 people last week from a company that should know better on some of these issues.”
HACCP and other process data
Getting good data starts with HACCP. But besides pleasing regulators, good process data makes business sense, and monitoring critical temperatures doesn’t require specialized software. According to Darren Riley, segment business lead, software and applications, at Rockwell, “If you have the right infrastructure, you don’t need any special software. HACCP is looking at specific control points, and I’ll bet you that plant people are realizing there are a lot more control points than just the temperature of the oven.” It’s the temperature of the facility that might be causing further temperature fluctuations. Or the temperature and pH of the CIP solution might also be important to prove HACCP data during an audit.Charlie Rastle, solutions marketing manager at Rockwell, suggests having access to the right data is important for improving the process. Looking at Riley’s oven example, Rastle suggests measuring gas flow into the oven and looking at the combustion mixture to see if there are fluctuations that affect efficiency, quality and reject rates.
According to Dr. George Cheng, CTO of CyboSoft, automatic control of process and quality variables is much more important today because of high commodity and energy prices. Any waste in raw material and energy affects the bottom line. Some of these variables include temperature, flow, level, pressure, pH, density, moisture, etc. The goal is to eliminate over/under heating and over/under drying so products are manufactured within the quality specifications at minimal material and energy consumption. “PID controllers are commonly used in the food industry. However, they cannot effectively control difficult process and quality variables such as steam temperature, pressure, density and moisture.” Unfortunately, Cheng says, these loops are often left in manual control, resulting in quality variations as well as energy and material waste.
For some applications, having the right data and a process model allows advanced control techniques to improve the process. Rastle suggests that for high-value applications (e.g., extracts, alcoholic beverages, oils), advanced control systems beyond the traditional PID loop can be used to optimize extraction or yield, or to make sure a quality standard is met.
Cheng says that model-based control such as model-predictive can help but is often too expensive to launch and maintain, so it is rarely seen in the food industry. Another approach, says Cheng, is to use adaptive control without models (i.e., model-free), which can provide the performance needed with no tuning for applications such as steam generators, evaporators, ovens, dryers, tomato hot-breaks, etc.
Always room for improvement
One popular key performance indicator (KPI) is overall equipment effectiveness (OEE). OEE includes machine uptime/downtime, availability, throughput and other information. According to John Oskin, Informance executive vice president, while OEE is an important metric, capacity utilization is also a strong metric to show how a processor is managing assets. Coupling these technologies, where appropriate, with a strong initiative to reduce cost/time for an operation goes along way to improving KPIs.Often, processors don’t know how really bad their OEE is until they install the tools to measure it. For example, Weetabix Food Company recently installed CDC Factory to grow the branded side of the business. “Prior to CDC Factory, we were unknowingly operating at 48% OEE,” said Paul Whitehead, plant manager. “We now have full visibility of the factory floor and have dramatically reduced production line stoppages, as well as reduced our labor and administrative costs.”
Yves Dufort, Wonderware director, manufacturing industry business development, thinks OEE is really important for processors that have a proliferation of SKUs, especially in newer, hot-selling product lines. On one hand, Wrigley’s Juicy Fruit has been around for years and years, and the process is well understood. But, on the other hand, Orbit is relatively new and very popular, and is available in several flavors. The last thing Wrigley needs is a stock-out on Orbit. So for a high-revenue, demand-driven line, OEE makes sense. Squeezing a couple of percentage points out of an older, existing revenue line won’t be as meaningful as on the Orbit line.
Obviously, the more automation that’s in place, the easier it is to measure OEE. But processors can get started without automation, says Brandon Henning, GE Fanuc Intelligent Platforms global industry manager for food and beverage. “It depends on what the processor’s perspective is concerning the metrics it needs, where it’s at in the OEE journey. If the processor is focused on major breakdowns and tracking that data, it may not necessarily need to invest in control systems.” Henning suggests the processor’s need to be better educated with what’s going on in the plant coupled with picking off some of the low-hanging fruit can get it started on an operational excellence journey, to which OEE belongs.
Picking off the low-hanging fruit is just what Ahold Coffee Company (the Netherlands) was able to do right after installing a Citect Ampla MES, which will have gained an ROI in less that 12 months. ACC’s operations manager, Jeroen de Jager, is now able to objectively identify the most readily achievable OEE goals (the “low-hanging fruit”) that should receive the highest priority, and also manage the resources for eliminating waste and stoppages more efficiently. In addition, MES has helped de Jager identify the poorer-performing lines and provide more insight into actual production costs.
More than track and trace
With the complexity of today’s supply chains, track and trace can be a matter of life or death for a brand. “Recent food safety scares have heightened consumer and retailer focus around product safety and traceability,” says Oracle’s Theel. “The risks to brand equity associated with product contamination are significant, and food processors are focusing resources on product safety, in terms of designing safety into new product development and manufacturing as well as ensuring product can be rapidly traced and recalled.”For software vendors, providing a track-and-trace module is a given, says Vernon J. Spaulding, VP integration services at Data Specialists. For the processor, Spaulding says track and trace is like buying an insurance policy; it doesn’t contribute to the bottom line, and it probably subtracts from it. Until a negative event occurs, a processor just doesn’t see its merits. Unfortunately, several organizations are still recording information on paper, placing it into files and trying to retrieve it manually when an auditor asks for it.
For some fast-growing processors with older, disconnected software, an upgrade to a product that automatically handles track and trace is the way to go. La Tortilla Factory had growing pains as sales of its heart-healthy tortillas took off. According to CFO Stan Mead, the recall procedure “was that of a paper trail located in several areas.” It could take several people the better part of a day to locate a particular batch and track it through inventory and distribution. Since putting in a Deacom ERP system, the 175-employee company now has instant access to supply chain, batch, costing and lab information. Mead says tracing ingredients takes less than half the time it did before, and he has seen a sharp increase in efficiency.
It’s time to get over thinking about track and trace as a means toward risk mitigation, and instead think of it as a way to focus on quality and brand enablement, says Rob Wiersma, Lawson director of industry strategy, food and beverage. Today, brand is often the most important aspect of a product, and maintaining the quality of the brand far outweighs the cost of the tracking and tracing of it.
Ali Ergün, a system controller at Arla Foods’ Brabrand Dairy in Denmark, says traceability sells product. “Customers who see the way we register the arrival of raw material and how we follow it all the way to the carton can relate to the process, and they are impressed. When it comes right down to it, it [traceability] is a good selling point.” The dairy uses Wonderware’s family of products.
“There are some processors that recognize they get more value than just the traceability from their involvement in track-and-trace projects. These companies are seeing that they can flow the cost of the product through production and get more accurate costing and gross margin reporting,” says Terry McCorriston, director, CSB-System International. They are seeing they can analyze the production times and quality issues from different suppliers to improve their operation and reduce costs. Some are using traceability to show compliance from separation of organic or other specialty specifications of the food.
Rastle says that in most cases, his customers are putting in track and trace not for regulatory compliance, but for providing the next level of information so they can track the products through manufacturing and optimize their entire production process.
“Before track and trace helped with the regulatory ‘who, what, where and when,’” says Beth Berndt, CDC Software director, product management, industry solutions consumer products. Now, she says, it helps processors understand the “why.” Why was there a problem? Was an ingredient off-spec? Was it the processing? Was it the operators? Knowing the answer to “why” can help solve quality problems, and should there be a recall, it can narrow the window to a much smaller opening.
Batch management, says Heydenrich, is key for monitoring and tracking and tracing because batch management gives an identity to the raw materials that were procured and stored in the production of a food product. Batch management allows for an identity to be given to a semi-finished good and a finished good that is ultimately going into the supply chain. Good MES and ERP systems have the fundamental ability to identify materials via batch management. But should MES or ERP be the guardian of track and trace?
Track and trace as supply chain
Think of track and trace as supply chain and an ERP responsibility, says Harold Katz, Syspro technology enabling manager. Track and trace-just like supply chain-is concerned with receiving inventory into an ERP system, passing it on to work-in-progress and the processing environment and then fulfilling the order. “A lot of people make the mistake of looking for their track-and-trace functionality in the systems that sit around the ERP system. You need it in the ERP system to manage it effectively.”Another way to look at the supply chain, says Spaulding, is as a division of lean manufacturing. It’s all about getting the right quantity to the right client at the right time, and in the quantities the customer desires. But not all industries lend themselves to this model. In particular, the dairy industry runs productions on demand, but there’s often a disconnect with other areas of the supply chain, e.g., raw materials, procurement and inventory.
Today’s supply chains, where part of the process is outsourced, present some new challenges. Berndt thinks of these as more logistical or business challenges than technical. She asks: In today’s world of outsourcing and copacking, who is actually responsible for the product at any given time? “If I’m outsourcing product, I’m kind of responsible for it, but it isn’t in my plant anymore.” So, today’s ERP systems need to manage all of these relationships for products the ERP system doesn’t necessarily “own.”
Risk management
Today’s increasing costs of commodities and energy call for some form of risk management. One weapon is the ability to forecast correctly. Katz says forecasting within an ERP system provides the ability to manage both the supply and vendor side of the supply chain.Jelly Belly, a family-owned confectionery business, produces 1,200 SKUs of products including 49 flavors of jelly beans from ingredients with widely fluctuating prices from all over the world. One method of risk mitigation is a good forecasting system. According to Jelly Belly’s Ryan Schader, VP business development, “Even though the commodities can be more expensive, it’s very difficult to get a price increase at the retail level. So supply chain for us starts with forecasting. We have to build a forecast.” It takes 14 to 21 days to make each flavor, and the company turns orders in 24 hours. “So if you don’t have the right product in the warehouse at the right time, you missed the order.” Lawson’s ERP handles 7,000 structures, each consisting of a plant resource, routing and bill of materials.
While no one has total control over commodity pricing, there are steps that can be taken to minimize risk, says Berndt. Commodity-driven manufacturers are buying bulk (futures market), they’re contracting for either livestock or crops so they can predict the costs, and some have actually become owners and growers who are producing their own commodities.
Heydenrych sees another approach. Processors should form buying groups to negotiate better commodity pricing, although he warns that special software will be needed to handle this approach. This method of forecasting has a problem-it’s somewhat one-dimensional. It doesn’t necessarily consider all the variables.