Technology / Process and Operations / Smart Industry

The Evolution of Industrial Automation in the Food and Beverage Industry

Hard benefits long have provided the financial justification for plant automation projects. As information technology plays a larger role, food companies are groping for other ROI metrics.

By Kevin T. Higgins, Managing Editor

Somewhere in the future, the grizzled maintenance engineer leads his apprentice into the food plant’s electronics boneyard in search of some salvageable parts.

In a dusty corner is a stack of obsolete parts. “Hey, Gus, what are those?” the apprentice asks, pointing to a flat screen monitor and wiring wedged between a fax machine and CD player. Smiling in recognition, the engineer replies, “That there’s an HMI screen and Ethernet cable, Sparky.”

The story of industrial automation is an evolutionary tale, from relay logic circuits to PLCs to SCADA systems and ladder logic to Windows-based software to agnostic systems. Chucking the HMI at this time would be premature, but the groundwork is being laid under the wide umbrella of the Industrial Internet of Things (IIoT).

IIoT and Industry 4.0 describe today’s front lines in plant automation. Whereas the prior battleground was on the operational front, IIoT skews toward IT and communication flow between machines and professionals and professional to professional. Mobile devices and remote access are the exciting parts, and wi-fi networks and cloud computing are the enabling technologies.

The goal is to facilitate information sharing and help a manufacturer run faster, better and cheaper. That’s a big shift from operational automation that produces many soft benefits but is justified on the basis of displaced labor: If the capital cost is equal to the labor cost over three years, then the ROI is 36 months. Faster communication may mean reduced downtime and less physical movement, but those are elusive benefits to quantify.

The new automation battlefront also requires a significant upfront investment, and that gives food and beverage manufacturers pause. Digital field devices and machines with IP addresses are part of the necessary infrastructure, along with a reliable and secure communications network. “It’s expensive and fairly complicated, but it’s not exactly science,” the former CEO of Siemens Digital Factory told Industry Week last year. “It’s doable, we know how to do it, but we have to pay for it and we have to find the people to do it.”

One of the most ambitious IIoT efforts in food to date is SugarCreek's "new" plant in Cambridge City, Ind. A $130 million gut-rehab that will have its ribbon-cutting ceremony this month, the plant stretches SugarCreek out of private-label bacon and meatballs into sous vide, primarily for foodservice.

A wi-fi network, simplified radio dispatch and response, web conferencing, energy monitoring and management and 250 video cameras are among the project components. Real-time locating systems to track objects and people are a key element of a high-performance work team initiative.

The IIoT components represent a $6 million investment. Even a one point increase in yields would more than offset the expense, although CIO Ed Rodden believes tying any ROI metric to essential infrastructure misses the point. “A manufacturer today must have an infrastructure that enables the broad use of technologies to support the business, from collaboration to data collection,” he maintains. “We don’t view (IIoT) as any different from other infrastructure required to be in the business, such as refrigeration in a food plant.”

Sky high sausage

With few exceptions, food companies are taking a wait-and-see approach to production facilities untethered to hard-wired data networks. Which is not to say they are not exploiting opportunities to incorporate state-of-the-art automated processes.

A case in point is OSI Industries’ dry sausage plant in West Jordan, Utah. Opened five years ago, the facility makes extensive use of robotics, with automated guided vehicles (AGVs) and automated storage and retrieval systems (AS/RS) shuttling racks of product between stuffers, drying houses and other points of production.

When OSI embarked on the project, more than a quarter century had elapsed since a new U.S. dry sausage plant had been built. Facing a dearth of new sales opportunities, the domestic vendor base had atrophied, causing OSI design engineers to look abroad for the latest technology, according to Larry Glaser, assistant vice president and director of operations support at Aurora, Ill.-based OSI (www.osigroup.com).

Other than human hands that place sticks holding sausage logs onto racks, product movement from grinding, stuffing, drying and cooking is handled by machinery from European vendors. Three tiers of racks, each holding 72 logs up to 48 inches long, stretch 18 feet high and are moved en masse from point to point. In-process samples also are automatically pulled for moisture measurement and other quality and yield tests.

ROI is calculated by labor reduction. The plant has a staff of 28, according to Allen Forkell, assistant plant manager, adding that a conventional sausage operation with comparable throughput would require more than 100 workers.

Soft benefits in the form of improved food safety, fewer injuries and a smaller building footprint add to the payback. Human handling of products increasingly is viewed as a vector of cross contamination. Also, the towering racks of Genoa salami and pepperoni that are shuttled and stored allowed OSI to build up, not out.

“You don’t have the wear and tear that you have in a normal facility, where forklifts are always running into walls and doors,” adds Glaser, citing another cost savings. Energy use is lower, as well, thanks in part to lights-out operation, a condition that also minimizes the potential for oxidation of the meat. The 50,000-sq.-ft. building envelope occupies less than a third of the real estate a conventional design would have required, he says.

“This type of automation requires someone with ladder logic programming ability,” Glaser continues, and plant engineers are more likely to have a computer background than a mechanical one. Radios controlled the AGVs when the facility opened, but OSI has since upgraded to a wi-fi network that also supports data entry for quality checks and other documentation on tablets, explains Forkell. Digital records have essentially replaced written reports, although paper reports are generated for public health inspectors who require them for HACCP verification.

Since 2014, a data historian has stored track and trace information for every lot, along with data for statistical process control. The camera system also has been greatly expanded, Forkell says. “The coolest thing about our plant is that we know where everything is,” despite a work in process inventory representing millions of dollars in finished-goods value, he adds.

Low-value data

“The talent level you need for that plant is not someone with a wrench and a hammer but someone with programming and IT skills,” Glaser says. Multiple communications networks coexist, including wireless temperature monitoring.

Local temperature monitoring is giving way to sensor networks that relay data to the cloud and are accessed by smart phones and other mobile devices, suggests Rupert Prince, president of Lizard Monitoring (lizardmonitoring.com), Portland, Ore. If companies haven’t already transitioned from point-to-point hardwired networks, the sanitary transportation rules under the Food Safety Modernization Act will provide additional impetus to make the switch.

“Wireless sensor networks have been around for 10 years, but it wasn’t until recent years that they basically did what was promised 10 years ago,” says Prince. He compares the maturation curve to cellular telecommunication, which was dogged by dropped calls and other problems until the technology delivered reliability.

“There’s still a great deal of concern about security,” he says, “but the same technology that is used to secure credit card processing surrounds cloud-based systems.”

Maintaining the cold chain during transport will put an added premium on wireless. Prince’s firm is developing a system that uses the telemetry of the cellular system to provide real-time alarms when trailer temperatures drift above setpoints. Even a trailer load of Willamette Valley strawberries or raspberries is worth $250,000-$300,000, he points out, making wireless sensing an inexpensive safeguard against product loss.

How prevalent such systems will become is unclear. Despite the availability of digital field devices and interest in IIoT, the majority of devices currently sold to the industry still are analog, notes Ola Wesstrom, senior industry manager-food and beverage for Endress+Hauser (www.us-endress.com), Greenwood, Ind. Four years ago, his firm introduced a magnetic flow meter with an Ethernet IP. Coriolis meters and other devices now have the same connectivity, and he expects a gradual migration to such upgrades.

Vendor-managed inventories of chemicals, fertilizer and detergents have been the sweet spot for digital IP devices, since they eliminate unnecessary restocking visits to sites, according to Wesstrom. Although they require hard wiring, smart mass-flow meters can monitor multiple variables, such as density, temperature and viscosity, as well as volume. Reliability and labor savings are two of the benefits, with the devices using the internet to communicate diagnostic and operational data.

Of course, if analog devices are meeting a processor’s needs, upgrades are a hard sell. “There’s no significant premium to go digital today, but comfort level can be a factor in not making the transition,” he says, though he expects that will change as a younger generation of workers replaces senior staffers.

Next-gen hardware

The HMI tethered to a fixed point on the production floor is an obsolete concept, many automation vendors agree. Remote access via smart phones, tablets and other devices can free personnel to troubleshoot problems and alter operating parameters without being in physical proximity.

“Mobile devices are already used as a portable or wearable HMI, mostly with the control function still residing in the machine’s central controller,” points out Daymon Thompson, automation specialist with Beckhoff Automation. “There are even opportunities to directly change parameters on machinery for some level of control through web-enabled devices.”

That freedom and flexibility comes at considerable cost, he cautions, if a manufacturer still relies on PLCs. “Complicated and expensive managed hardware would be required to force a PLC or PAC [programmable automation controller] into this new world of expanded connectivity,” Thompson adds. “The key to making this work cost effectively is by utilizing PC-based control architectures.”

The HMI is obsolete, concedes Alan Cone, WinCC marketing manager for Siemens, quickly adding, “Is it smart to make my phone an HMI?” Cone raises the specter of butt dialing and the inadvertent altering of machine operating parameters.

WinCC—a contraction of Windows control center—is an HMI and supervisory control and data acquisition (SCADA) system. Like Rockwell Automation and many other automation suppliers, Siemens still sells a lot of PLCs, but the firm recognizes the need to nudge manufacturers closer to PC controls in order to leverage the power of IIoT. For its part, Rockwell stopped updating PLCs when Logix was introduced in 1998.

A counterpoint to the increasing sophistication of plant data systems is the emergence of inexpensive hardware with almost a DIY vibe. Instead of retrofitting machines lacking controls with PLCs, MacLean-Fogg, a manufacturer of precision fasteners and other parts, installed Controllino serial devices that use open-source software as part of its IIoT initiative.

Raspberry Pi and Arduino are examples of cheap motherboards that serve as the base of a build-it-yourself PC for under $350. Few food professionals would risk their careers on automation tools that aren’t industrially hardened, but the availability of low-cost options in a fast-evolving sector challenges hardware and software assumptions.

To remain competitive, automation vendors are emphasizing affordability and security of their remote-access and cloud-based systems. “Better information flow is what the digital factory is all about,” Siemens’ Cone maintains, and the data mesh on site is the essential infrastructure for increasingly sophisticated data services.

Digital data in the cloud is one of those services. Similarly, Schneider Electric offers a cloud-based historian to collect and contextualize data. According to Saadi Kermani, project manager for SmartGlance, more than 700 industrial cloud services now are up and running, with more on the way.

“Cellular companies are trying to get on the IoT bandwagon by offering to erect towers adjacent to plants,” says Kermani. “We have conversations going on to form data-moving partnerships.”

SmartGlance provides remote access to OEE, waste rates, error codes and other operating data, either through a laptop or with a mobile app. “We frame it as IT in a manufacturing context.” He describes the individual who takes ownership as “a person in an IT-like role but representing manufacturing."

Similar capabilities are rolled into Rockwell’s enterprise manufacturing system. A pared-down, standalone application for mobile access is rolling out under the rubric Project Stanton (the name is a call-out to Dr. Stanton Allen, one of the firm’s founders). Stanton skips the cloud and plant wi-fi by leveraging the power of the smart phone, which essentially is “a supercomputer in your pocket,” suggests Kyle Reissner, mobility platform leader. The dedicated phones are stripped of their SIM cards, and machine data is accessed through purpose switches in control panels by a server in the phone app itself.

Still in beta mode, Project Stanton currently delivers eight modules: three related to trending and other machine data, five related to human interaction and the leveraging of the knowledge of specialists on staff. “Measurement of impact is the single biggest software challenge,” Reissner allows, “but we’re OK being on the bleeding edge.”

Unfortunately for Stanton and the other technologies, the benefits are primarily time-related. Enhanced capabilities like condition monitoring of equipment are tantalizing but require higher levels of sophistication. IIoT promises to untether the plant from conventional controls and data systems, but without an upfront investment in the necessary infrastructure, it will be viewed as nice-to-have, not need-to-have, technology.