Plant automation: Finding the right fit
Do what makes sense when installing automation in your facility, considering your budget and business goals. Understanding your application can help you fine-tune your equipment purchase.
While the trend for large manufacturers is to automate a new plant in its entirety, small and medium-sized food and beverage processors rarely have the funds to automate their plants, turning them into lights-out facilities. Even some tier-one processors may find areas in certain operations where it may not make sense to over-automate. But learning to spot the areas where automation can be the most effective and bring the best return on investment is a step in the right direction.
- You can’t control what you don’t measure
- Updating the process: Where are you?
- Packaging—Get it right!
- Networks—Are we connected?
- How to cope—Get some advice
Large controls suppliers provide a variety of complete automation hardware/software solutions. But budget-conscious manufacturers have plenty of alternatives to automate in selected areas where it is needed—whether to meet federal record-keeping requirements or improve quality, throughput and efficiency. Rockwell Automation solutions enable the development of a connected enterprise, says Dave Sharpe, Rockwell director of consumer packaged goods industry. “A fully connected enterprise [provides] the ability to connect and share valuable information among people, devices and machines. It has a single, coherent environment, built from disparate automation systems that can communicate with business-level systems to increase visibility across an organization and arm decision-makers with better information.”
Of course, any controls/sensors you purchase today will probably be network ready. But, knowing your application can help you fine-tune your equipment purchase.
“What underpins all our products is the concept of ‘fit for purpose,’” says Curtis Uhll, West Control Solutions engineering manager. “By this, we mean a [processor] should be able to choose exactly the right combination of features and functionalities for a specific application—at a commensurate price. Not every application needs every bell and whistle; some requirements are very simple and straightforward. By developing solutions specifically for those requirements, we can deliver not only a much more cost-effective system, but one that is simpler to deploy and use—and that can contribute to error minimization.”
Whether a piece of process equipment works in conjunction with other equipment or is operated standalone, it’s important that it can work intelligently offline or online with other systems. “Our Polar Processing systems offer an ‘island of automation’ to ensure the customer’s system functions correctly and can capture data at all critical stages of the process to ensure accurate batch reporting both on and offline,” says Chris Mason, JBT Wolf-tec director of sales and marketing. These systems include controls at several stages in the process and work orders that refer to a master recipe for a product that, for example, includes brine formulation; injector, maceration and massager settings; and all product transfer speeds and feeds. Thus, a line of Polar Processing systems can be run by a small number of operators.
Understanding the differences between Industrial and office Ethernet data transmission
Industrial Ethernet is becoming the plant floor network of choice in both processing and packaging areas. EtherNet /IP (Ethernet Industrial Protocol), managed by ODVA (Open DeviceNet Vendors Association), is an application layer communication protocol that is commonly associated with many Rockwell Automation control products and is based on the Common Industrial Protocol (CIP) standard.
The differences between typical front office Ethernet networks and industrial networks—and the familiarity of traditional IT staff on how to configure and maintain these networks—can create challenges. There are three types of traffic on an Industrial Ethernet network. Unicast traffic routes directly from one point to another point, while multicast traffic routes from one point to many points. Broadcast traffic routes from one point to all points.
Two primary differences between front office and manufacturing applications are the utilization and configuration of managed network switches specifically for multicast traffic, which typically comes from smart devices on plant floor process networks in a connection-oriented producer/consumer-based protocol. In this context, a connection is simply a relationship between two or more nodes across a network. Typical items in a multicast group include flow meters, variable-speed drives and scales. Each of these produces process data and consumes configuration data.
A device must be a member of a multicast group to receive group data, and all members of the group receive data. However, a device doesn’t need to be a member of a group to send data to the group. The main problem with multicast traffic in a producer/consumer model is that traffic grows exponentially with the number of hosts (devices). This is where the managed switch comes in.
A managed switch can turn on Internet Group Management Protocol (IGMP) Snooping. When enabled, IGMP Snooping sends out broadcast traffic to determine the members of any multicast groups. Using this information, combined with its media access controller (MAC) address table, a managed switch routes multicast traffic only to those ports associated with members of a multicast group. A non-managed switch (or improperly configured managed switch) treats multicast data the same as broadcast data and sends it everywhere.
Understanding the above is necessary for an IT staff to properly configure and manage industrial networks over time. Connecting the front office to industrial networks sometimes results in broadcast storms from the front office, which can bring the industrial network down to a crawl. Knowing how and what to connect between these functional areas is essential for the continued good performance of each.
Source: David McCarthy, TriCore.
With JBT systems, measurement accuracy plays a critical role in getting processes right. For instance, to make a brine solution, incoming water is measured for weight and temperature. The ingredients specified in the recipe are put into the system and mixed only when the water has reached the proper temperature and been mixed for the correct period of time. Succeeding processes also depend on strict parameters of upstream and downstream conditions. “Manual processes often allow process checks only to be ‘adjusted’ rather than recording actual data points, making it unlikely to meet quality standards,” notes Mason.
In an automated process, temperature, pressure, flow, conductivity, percent solids, level, turbidity and pH are commonly monitored, says Steve Malyszko, president and CEO of Malisko Engineering Inc., a certified member of the Control System Integrators Association (CSIA). “Operational parameters that are sometimes monitored include valve position, motor speed, motor current, motor vibration, presence or position, and the rotation of a device such as an agitator or mixer,” adds Malyszko.
“For many years, there wasn’t a lot of true data acquisition going on in a process,” says Jason Stricker, Shick Solutions national accounts manager. “If it worked, [processors] left it alone. If they had some kind of problem, they just started physically and mechanically trying to troubleshoot things. The movement today is toward more data-based decisions. That brings into play a lot of monitoring—whether it’s in a drive, liquid or mixing system.”
While new equipment typically features complete control loops, older systems may not have the sensors or the controls, and the result can be costly. “We still see a lot of inaccurate ingredient delivery,” says Timothy S. Matheny, president of ECS Solutions, Inc., a certified CSIA member. “For example, we recently saved a customer $500,000 annually just by using our statistical predictive delivery technology to reduce the overshoot of expensive ingredients. A $15,000 instrument that allows the delivery of two materials at once can often pay for itself in weeks, destroying the argument that small manufacturers cannot afford to automate.”
In beverage plants, monitoring temperatures for presterilized or pasteurized products is important, but so is the measurement of the liquid flow of product in sterilizers and container flow through pasteurizers, suggests Tom Braydich, senior consultant for Matrix Technologies, Inc., a certified CSIA member. “Processors need to be equally concerned with product hold times and temperatures. With some products, like soup or pasta, bacteria kill steps may be dependent upon specific ingredient fill weight. Checkweighers ensure precise fill weights and proper process duration, while inline instrumentation may be used to measure product concentrations and viscosities.”
Improving process control is a two-step process, according to Ola Wesstrom, Endress+Hauser senior industry manager, food and beverage. Instrumentation and automation can be thought of in what Wesstrom calls “two buckets.” One is for traditional process control (e.g., what is the pump pressure, tank level, heat exchanger temperature?), while the second bucket is inline quality monitoring or control. “The starting point is to visit the plant’s lab and determine what measurements can be supplemented or replaced with inline measurements,” says Wesstrom.
Using inline analyzers helps manage many issues. For example, the amount of disinfectant used on a hydrocooker for canned food needs to be closely controlled, as overdosing can cause corrosion and the waste of chemicals, while too little can compromise food safety. Wesstrom recalls one plant that had monitored disinfectant by taking grab samples to a lab for analysis twice an hour before installing inline analyzers to measure free chlorine, pH and the disinfectant’s conductivity. Real-time measurement eliminated overdosing, saving the plant $13,000 annually in disinfectant costs. The measurements also allowed the automation system to add makeup water based on measured values, saving on energy and water usage, and producing less wastewater. In addition, the inline analyzers eliminated the need to send a lab worker to the hydrocooker to take grab samples. The bottom line was a payback period of just seven months.
Tracking disinfectants isn’t the only way to reduce costs. By continuously tracking weight fluctuations on the production line, operators can make a fix to a specific ingredient to ensure finished product is at just the right weight, says Rockwell’s Sharpe. “A company like Hillshire Brands is concerned with weight, because any off-weight Jimmy Dean one-pound, fresh sausage roll is manually inspected and either put back into the system or trashed. The Federal Meat Inspection Act requires packaging to represent content quantity, and if an inspection finds these packages are mislabeled, the product cannot be sold, and fines can be incurred.”
Michael Griffith, Stellar automation manager, cites another place where sensing and control can potentially bring about savings. “Food processors could do a better job of tracking plant utilities [WAGES, i.e., water air, gas, electricity, steam] usage, not necessarily to increase throughput or quality, but to better manage costs.”
If you’re still using paper, you might not be alone—but you could be in a bind when FDA asks for your records. “Food and beverage regulations continue to evolve as producers move to more dynamic production, increasing the complexity of production processes,” says Rockwell’s Sharpe. “Many food and beverage manufacturers rely on paper-based information-gathering methods that can be time consuming and subject to error. Information software built to work directly with the automation already on the plant floor helps by proactively identifying and alerting workers to issues that could affect product quality and lead to recalls.”
The lack of a sufficient level of instrumentation or sensing devices to provide robust control and monitoring in processes is the most obvious stumbling block to improved automation, according to Malyszko. Most processors agree philosophically that they need to invest in more instrumentation, but reality sets in when the discussion comes around to the cost of improvement. “Too often, the primary factor for not moving forward is the high cost of the sensing devices. When this happens, a hearty discussion on ROI and risk avoidance is necessary, if the initiatives are to move forward,” adds Malyszko. Sometimes, the systems that are in place are not being used to their full advantage. “Facilities could do a better job of leveraging the connected automation devices they have,” says Stellar’s Griffith. “In many facilities, a well-integrated system is on a common network and platform, yet there is little interaction between various systems.”
When PLCs are in place, some processors fail to take advantage of their data capabilities. “Food processors that use automation to weigh, dispense, heat and blend a recipe frequently fail to capture and scrutinize the data that is readily accessible from PLCs,” says Matrix’s Braydich. He adds one of the reasons this happens is the processors’ antiquated systems lack Ethernet connectivity or MES software.
“Low utilization of equipment is a common problem,” says ECS Solutions’ Matheny. “The [processor] may not be using an ISA-88-compliant batch-management system and/or may lack simple equipment, such as a staging scale tank that reduces batch times on the bottleneck unit[s]. Additional or appropriate instruments can often reduce processing times, producing significant capacity for minimal capital investment. We almost never see basic process control software that is truly ISA-88 compliant, which would allow processors to add new products by adding new recipes without modifying the PAC/PLC programming.”
Melissa Topp, ICONICS director of global marketing, reports several processors have requested the kind of assistance with product quality/consistency and recipe control the ICONICS BatchWorX ISA-88-compliant batching system can provide. The processors have the hardware (PLCs, temperature controllers, etc.) in place, but need software to get a complete view of their processes.
While inadequate automation infrastructures and sensors—or low-level industrial networks—may have been a problem in the past, new automation technologies are driving software solutions, says David McCarthy, president and CEO of TriCore, Inc. In addition, the low cost of storage and bandwidth, along with the rise of smart devices, is creating an information explosion in manufacturing systems. Today, many systems are being implemented to move data between the plant floor and the boardroom, creating actionable information to produce more efficient operations.
According to West Control’s Uhll, ease of use remains a problem. “While the accuracy and quality of automated solutions are pretty much taken for granted, [processors] find the largest area of frustration is ease of use.” West Controls remains sensitive to these issues and designs systems accordingly. However, Uhll says there is a very real tendency for systems to be over-specified, perhaps driven by fear of not having the “best” solution or a desire to have future flexibility. “It’s important to realize that over-specifying comes at a cost.”
While many people may believe OEE is one of the performance metrics that’s well understood, it can be hampered by issues like poor flow control, a lack of monitoring line performance, multiple control systems, changeover issues, bottlenecks and broken bottles, or damaged cans or packages. “The most common problem on the packaging side is poor flow control,” states ECS Solutions’ Matheny.
This may start with the front end of the packaging line. As Shick Solutions’ Stricker contends, the problem isn’t always in the ingredient processing area; it’s the issue of consistently making product faster than it can be packaged.
Packaging problems often go back to the system design stage. “The number-one [issue] is not having a well-thought-out line performance and monitoring system in place to capture raw operational data and produce meaningful information,” says Malyszko. “The system must capture data and produce information that will directly help the production team. To accomplish this, these production stakeholders must be included in the early designs, specifications and selection processes.”
“Typically, food processors lack the concept of good line control,” offers Matrix’s Braydich. “For example, if you walk into a facility where glass bottles are being filled, and all you hear is the clanging of glass, that’s your first clue the plant needs help with line control. The second clue is broken or damaged containers. Older systems lack a PLC that orchestrates the flow of the line from the depalletizers to the full-case palletizers. Each island of automation passes analog and digital signals up and downstream to adjacent equipment. This design worked years ago at lower speeds, and when glass and metal container weights were much heavier, but increasing speeds and lighter containers require [faster] automation.”
Implementing effective line automation requires a full evaluation of the line, starting with a critique of each piece of equipment for run and surge speeds and efficiency. Then, a “V-graph” should be built to determine which machine is causing the bottleneck, adds Braydich. This evaluation may reveal newer packaging equipment and additional sensors are required, along with more conveyors, to improve line efficiency and throughput.
Yet, there are other issues to consider when evaluating a line. “A lack of connectivity between equipment in the packaging process affects many manufacturers,” notes Mike Wagner, Rockwell Automation global business manager of OEM packaging. “When machines are purchased individually and not as parts of a line solution, communication between the machines can be affected when new solutions are integrated.”
Multiple control and network solutions also require additional integration, says Wagner. Using IT and OT (operations technology) solutions with the same architecture can make this integration easier. To address low-machine productivity, direct-drive linear and rotary motors should be specified, when possible, since they can reduce maintenance costs and downtime.
Changeovers and machine safety are two additional design considerations, according to Brad Schulz, Festo industry segment manager, end of line packaging. “Machine changeovers are both cumbersome and costly, and should be automated if possible.” Schulz also points out that machines—whether for the process area or packaging—must be designed for safe operation. Festo and other suppliers offer several machine components to both processors and machine builders to meet safety regulations.
Yet another area of concern in packaging systems is the maintenance of tight temperature control and timing to produce consistent heat sealing, says Uhll. Today’s films require a temperature range of +/-1°C, a far cry from that of +/-5°C a few years ago. Tracking and tracing is also an important aspect of packaging. Topp cites a customer, a large soft drink bottling company that uses ICONICS HMI/SCADA and ReportWorX enterprise reporting, charting and analysis software to trace ingredients from bottles to their original source. A code is stamped on the outside of the bottle, so employees can quickly trace what batch of syrup was used and determine the origin of the water used, as well as the cap and the bottle itself. A report is then generated, showing where other finished products containing any of the source ingredients have been shipped.
“It takes more than a network to integrate manufacturing equipment and business processes,” says ECS Solutions’ Matheny. “But many food processors are slow to automate, because they have trouble imagining how automated processes can be agile enough to withstand today’s constant change.”
“An increasing number of food and beverage corporations now have access to massive amounts of data that should be able to provide a ‘complete picture’ of production,” says ICONICS’ Topp. “However, not all companies have adopted software technology solutions to allow them to turn that big data into actionable intelligence. Leading-edge manufacturers have it down to a science, with every stage [from recipe planning to mixture to packaging to shipping] interconnected, and each has its own set of shareable data that affects the other stages. In short, most plants have the necessary network infrastructure, but only some are utilizing big data analytics solutions.”
While many processors have recently recognized the value of a more integrated system, they find it difficult to connect processing with packaging, says Stellar’s Griffith. “Unfortunately, some plants were built with such an array of hardware and software that integration at the current juncture is not a straightforward affair.”
“Often, from an integrator’s perspective, the issue is that plant automation systems have outgrown the capabilities of legacy networks,” says Malyszko. Simple communications among controllers is not sufficient to support the need for more data within the automation system. On the other hand, fast-update HMIs, data historians and event loggers can be large communication resource hogs that tax the throughput of legacy communication networks. Plus, you need greater security, says Malyszko. “The result is replacing 20-plus-year-old communication networks with systems containing much higher throughput, versatility and security.”
Unless you have your own engineering staff, getting outside assistance may be necessary to update legacy systems, since automation issues must be tackled on an individual basis. Suppliers, A&E/C firms and system integrators are good sources of help.
“The implementation of automation solutions in a food and beverage plant must carefully assess and map out how to capture the right data and turn it into meaningful information,” says Rockwell’s Sharpe. The connection of operation and information technology across the enterprise can be a strategic asset. Sharpe outlines three important points:
- A critical first step in implementing a solution is a baseline assessment of the plant with a focus on the goals related to quality, downtime, productivity and OEE.
- Once a goal is identified, an implementation plan can identify the objectives, problems and metrics.
- A complete assessment enables processors to create a consistent plan for implementing automation solutions.
Control System Integrators Association, www.controlsys.org
ODVA (Open DeviceNet Vendors Association), www.odva.org
Profibus Association, www.profibus.com