Process and Operations / Technology

When it Comes to Plant Automation, Smart Machines Mean Smarter Production

Whether a food company is striving for smart factories or simply wants the agility to respond to new-product demands, automation can help.

By Kevin T. Higgins, Managing Editor

Munich’s annual Oktoberfest is rightly regarded as a bacchanal with 6.3 million of your closest drinking buddies, but there’s also a county fair component, with dozens of antique tractors and carnival rides like Eva’s Trip to Paradise.

An early 20th Century tilt-a-whirl without the whirl, Eva’s Trip is powered by its original 11-hp Siemens motor, in service since 1939. The Germans call it a roller coaster, but compared to the thrill rides elsewhere on the fairgrounds, it’s a tame relic.

A similar analogy applies to the machines that make food. Mixers, ovens and even robots have the same functionality today as they did decades ago, but automation advances create a night-and-day comparison in terms of performance. And the best, as they say, is yet to come.

The programmable logic controller is the automation equivalent of Eva’s Trip. No automation vendor will say it out loud — they collectively sell millions every year — but the PLC is obsolete technology. Sure, it provides the logic needed to safely and efficiently run a machine, and its industrial hardness is unquestioned, but PC-based controls are essential in the era of Big Data.

The Achilles heel of automation is loss of flexibility, an easy trade-off when production schedules are static. The trend in food and beverage is in the opposite direction, points out Eckard Eberle, CEO of process automation at Siemens Process Automation in Nuremberg, Germany. The flexibility challenge is further complicated at companies with multiple facilities, none of which is ever identical. That works against operational efficiency.

Extracting data from sensors and field devices as well as controls is well and good, but unless it can be converted to useful information, the effort is wasted. Collecting data isn’t hard: Siemens’ Craig Nelson, senior product manager-motion control, cites the example of a manufacturer who currently consolidates 30,000 data points a minute from its facility. “People who are on the leading edge of the digital factory view data as their competitive advantage in asset utilization, waste reduction and peak-demand shaving to reach the optimal point to produce,” says Nelson.

Product genealogy

The data beast drives supply chain optimization. The Barilla Group recently launched an experiment that matches raw material genealogy with product serialization. Working with the Italian division of Cisco, Barilla is printing QR codes on farfilla pasta and tomato and basil sauce containers sold in Italy.

By scanning the code with a smart phone, shoppers are linked to a website where details on where the raw materials were harvested, where the grain was milled, which plant processed it and when the finished good was distributed. The “digital passport” embedded in the code ostensibly advances food safety and answers the “where did this food come from?” question posed by consumers -- although the more significant advantage may be the supply-chain visibility it should provide Barilla.

“The Internet of Everything changes the way we farm, produce, distribute and consume food, making it more transparent and therefore safer,” Agostino Santoni, CEO of Cisco Italia, maintained in a prepared statement. But the pilot project is really a baby step in the more ambitious goal to track the more than 1,000 raw materials Barilla sources, all the way from the farm field to the supermarket. Data manipulation on a massive scale will be required, and farfilla and sauce are the shakedown test.

Big Data and its sister, the Internet of Things, also play a central role in another Cisco project involving Sugar Creek Packing Co. , Washington Court House, Ohio. The company recently commissioned a brownfield project in Cambridge City, Ind., a facility acquired from a bankrupt food processor. Harvesting large amounts of data and feeding it back as actionable information was considered essential for establishing a high-performance work team structure at the new plant, explains Ed Rodden, chief information officer.

“High-performance work teams are quite different from what you see in traditional meat processing facilities, where top-down management is typical,” says Rodden, The semi-autonomous teams work with little supervision, with production, maintenance and HR issues handled by team members. If they are to hit their production targets, they need meaningful feedback. Video monitors with KPI numbers are insufficient.

A sous vide cooking system is the centerpiece in Cambridge City. “That’s a disruptive cooking technology and a highly automated system, with hundreds of sensors to control the process,” Rodden says. To access the data, team members use a Cisco mobile app called Jabber, “essentially an IP phone,” he adds. Conventional cell phone coverage in an industrial facility is spotty at best, and installing a booster system would have added $300-500 million to project cost. Jabber radios essentially function like a desk phone and integrate easily with plant software.

The wireless network also will track worker locations within one meter via RFID tags embedded on protective headgear. “It sounds like Big Brother,” allows Rodden, “but it’s primarily about safety. And knowing where people are working is a problem in meat plants, where they move around a lot.” A bigger issue will be filtering data flows that could be overwhelming. “Software improvements address some but not all of the issues,” he adds.

Data glut absolutely is an issue, concurs David Sharpe, director-consumer packaged goods for Rockwell Automation Inc., Milwaukee, but there’s light at the end of the data tunnel. “Data needs to be conceptualized, and different tools to aggregate data and help understand what is happening in the process are becoming available,” he says. “But you need to know what the problems are before you try to resolve them. The more connected the enterprise is, the more options you have to drive efficiency.”

Compact foodbots

Other than notable exceptions like Pepperidge Farm, robotic applications were few and far between in food manufacturing until the late 1990s, when articulated-arm robots began cropping up in palletizing. Applications have spread upstream, but fencing around the work cell continues to limit use. The advent of collaborative robots that can work side by side with line workers is easing those space restrictions and expanding opportunities for automation, assuming the machines are robust enough for the industrial environment.

End-of-arm tooling, the presentation of materials to the robot and other factors enter into the equation, but risk assessments at both the OEM’s shop and the end-user’s specific installation enable these free-range bots. Four parameters — force, power, speed and momentum — are considered in those assessments, and if the assessment determines that the robot poses no threat to human health or adjacent equipment, the application is considered in compliance with U.S., Canadian and global safety standards.

Affordability and ease of programming are highlighted by collaborative robot suppliers like Universal Robots USA Inc., Ann Arbor, Mich. Denmark-based Universal has shipped 4,000 “cobots” in the past 10 years.

Sensors and safety devices are built into the robot’s controls, according to Brent Bartson, senior technical support manager, and systems integrators can plug in light curtains, safety PLCs, laser scanners and other components to create “a virtual wall around the robot.” Payloads are limited — Universal’s biggest machine can only handle 22 lbs. — and although the controls are proprietary, Universal employs open architecture to facilitate data transmission to EtherNet/IP or other communications protocols.

“It’s virgin ground, and if I was a food manufacturer, I would like to know the training credentials of the integrator and what he knows about the safety standards,” Bartson cautions. Risk assessment by independent parties such as Pilz Automation Safety is essential before deploying a cobot.

Universal’s European heritage reflects a new approach to machine safety that began on the Continent and is slowly migrating to the Americas. Instead of safety PLCs and e-stops, European machines are incorporating drives with built-in logic and motion controls, slashing response times with “aggressive deceleration” to prevent collisions and safe torque to lessen a possible impact, according to Monte Swinford, an engineer with Bosch Rexroth Corp., Charlotte, N.C.

Depending on operating code and machine speed, conventional safety controls can take up to 100 milliseconds to stop a machine, allowing several inches or even feet of motion. When logic and motion are built into the drive, only 2 milliseconds would elapse.

As important as safety, this type of controls automation can promote big productivity gains, Swinford points out. If power is cut off, capacitors must recharge before operations can resume. By switching to safe mode, advanced drives avoid that downtime.

Even a general purpose VFD can contribute regenerative energy to today’s manufacturing facilities, adds Siemens’ Nelson. Instead of dumping kinetic energy to a resister and dissipating it, today’s control technology can recover breaking energy from motors and return it to other power users or store it in batteries for later use.

Energy efficiency, supply chain traceability and safer operating environments are important, agrees Joe Martin, regional sales manager with Beckhoff Automation in Savage, Minn. However, greater flexibility and easy changeover are the mantra of modern manufacturing, and “flexibility is one of the trademarks of an open, PC-based control system,” Martin emphasizes.

“By moving to a PC-based system, manufacturers boost their ability to gather and repurpose vast amounts of plant information and production data,” he notes. “Traditional PLCs, while proven as wonderful products to replace relays, are no match for the processing power of modern PC-based controls.” PCs also facilitate the use of cloud databases to send and access encrypted plant data, facilitating corrective actions, predictive maintenance and faster changeovers.

Automation won’t replace all manual processes, at least in the lifetimes of Earth’s current residents, and levels of automation needed are as varied as the sizes of food companies. Regardless of whether they are intent on building the plant of the future or are content with the equivalent of Eva’s Trip to Paradise, food manufacturers need to be aware of where improvements are occurring and which best fit with their operations.

Keep up with Putman Media's Smart Industry Event via our Twitter Feed