There’s nothing like a massive affliction of the labor force to make you think about labor-saving techniques.
The pandemic, along with all its other disruptions, gave food and beverage processors an incentive to at least consider automation. Workers were getting sick or fearful of getting sick on the job – a fear that was all too justified in many cases. Automation is being seen as one of the best ways to deal with that fear, both immediately and, especially, going forward.
The pandemic revealed other weaknesses, especially in how production and supply chains handled shifts in demand, that could potentially be eased by automation. It has renewed or intensified processors’ interest.
“The usual attitude pertaining to automation, robotics, and artificial intelligence, in particular in manufacturing, was the devil” because they kill manufacturing jobs, says Max Versace, CEO of Neurala, a supplier of inspection software for machine vision. “But when the pandemic hit, all of a sudden, there is no level of automation near the level required for us to continue production in the face of additional requirements the pandemic put in place.”
Those requirements take many forms, but they can be divided into two broad categories: labor and flexibility.
Labor was the most dramatic, attention-getting problem – at least in the early stages of the pandemic. Workers kept getting sick, especially in meat and poultry processing plants, which rapidly became centers of infection in their mostly rural communities. Between actual cases and fear of sickness, processors couldn’t find enough people to staff their plants.
Flexibility also came into play a great deal during the pandemic, thanks to the shifts it caused in consumer demand. In some cases, processors had to ramp up production of certain mainstream products that were suddenly in high demand due to pantry stocking. In others, single-serve and other SKUs were needed, calling for higher rates of packaging changeover.
Ben Maddox, Western food, agriculture and wine executive for Bank of America’s Global Commercial Bank, thinks the pandemic redirected some of the momentum for automation that was already there.
“In general, the industry had been moving towards automation well before the pandemic occurred,” Maddox says. “I would say that it has caused an acceleration in the automation process more from a flexibility standpoint than anything else.”
The potential for automation, for both labor savings and flexibility, varies by industry, points out Suzanne Kopcha, vice president, consumer products & retail for Siemens Digital Industries Software.
“For instance, the meat processing includes a high degree of manual work from slaughtering and dissection until slicing, and therefore less automation. Whereas packaging within the meat processing industry already includes a lot of automation,” Kopcha says. In contrast, chocolate production usually has fully automated process lines. "But here packaging is less automated, especially for seasonal campaign products, e.g. Halloween chocolate pumpkins. With specific packages and blisters, these include a lot of manual work.”
AGVs Run Into Trouble at Water Bottler
Automation may have great potential to help the food & beverage industry, but it doesn’t always go smoothly. Case in point: the lawsuit of BlueTriton, formerly Nestlé Waters North America, against John Bean Technologies (JBT).
JBT supplied and installed automatic guided vehicles (AGVs) designed to transport pallets in Nestlé’s two U.S. bottling plants. The breach-of-contract lawsuit, filed in Delaware Superior Court, charges that one of the AGVs caught fire while recharging last September, leading Nestlé to ground all 51 of the vehicles. Nestlé terminated the contract with JBT the following month.
The lawsuit also charges that the AGVs repeatedly broke down and that they collided several times with storage racks and other fixtures, on one occasion causing thousands of pounds to fall from a fifth-level rack. BlueTriton is seeking for JBT to remove the AGVs and related equipment from its plants and pay compensatory and punitive damages.
Nestlé filed the lawsuit in March, shortly before it spun off the North American waters division to investors including C. Dean Metropoulos. The new company, renamed BlueTriton, is carrying on the suit.
A spokesperson for JBT declined to comment on the litigation, but said, “We believe our track record speaks for itself, having successfully designed and installed hundreds of AGV systems around the globe for our customers.” BlueTriton did not respond to a request for comment.
Determining which aspects of a plant’s operation are most amenable to automation is one of the trickiest, most challenging decisions a processor can make. Every plant is different, and there are almost infinite variables. That said, there are certain general principles that can be applied, taking into account both needs and the kind of equipment available.
Realistically, especially in plants that are at a relatively low level of automation, it’s best to go after low-hanging fruit. That means solutions that can be implemented quickly and show fast returns.
“Wherever existing solutions exist off-the-shelf and/or where clear ROI can be calculated, we tend to see these solutions adopted first and faster,” says Decker Walker, partner and global leader of agribusiness for Boston Consulting Group. “AGVs [automatic guided vehicles], robotic arms processes, sorting, visual inspection and foreign object detection are all examples where we have seen early and rapid traction.”
To determine the ROI of potential automation, it pays to look at three factors, says Mark McGinnis, director of systems engineering at Applied Manufacturing Technologies: “Throughput, ergonomics and quality are the three main factors that help determine the automation’s ROI.” Obviously, as throughput rises, so does the need for automation, he says. “If you are running two shifts, you should be heavily considering automating. If you are running three shifts, it might be too late.”
Ergonomics has many aspects: Are change parts too heavy to reasonably expect workers to lift? Do parts or tools have to be stored overhead due to floor space constraints? Do employees need to bend or stoop repeatedly, either in the course or normal operations or for cleaning and maintenance? These are signs that a machine may have to be either enhanced, for instance with servo-driven changeover parts, or replaced.
As for quality, automation can bring consistency to operations, but its biggest challenge is dealing with inconsistent product. If products are irregularly shaped or otherwise variable, it’s difficult for “hard automation,” which works mostly through mechanical means, to handle products with consistency. What’s often required in such cases is more sophisticated vision systems to locate and sort product, detect flaws and perform other functions, paired with sophisticated end-of-arm tooling for robotic arms, such as rubber grippers or vacuum cups.
Automation and packaging
Packaging tends to be the most automated part of a food or beverage plant. It’s a high-speed, iterative operation that usually lends itself well to being done by machine.
However, a couple of factors can hinder the development of automation in packaging. One is that, because it’s so pervasive, it has achieved legacy status in many plants – meaning it’s due for an upgrade.
“A lot of the automation in packaging facilities has been there for a while,” says Mark Ruberg, packaging industry manager for Beckhoff Automation. “If it’s been there for more than 10 years, there’s an opportunity to increase productivity.”
Another complication happens when packaging patterns proliferate, due to SKUs growing or customers wanting customized case or pallet loads. In such cases, there’s a temptation to keep doing things manually as much as possible.
That’s one reason why end-of-line packaging, like case-packers and palletizers, have a high potential for automation, says Steve Mulder, regional industry manager for CPG at Rockwell Automation. “We’ve seen over time, as you work your way into packaging and into end-of-line and logistics, automation kind of drops off step by step.” As customers demand more specialized loads, it is becoming increasing prevalent to create those loads in the plant, as opposed to breaking down and repacking them in a warehouse or distribution center.
The iTRAK from Rockwell Automation uses autonomously powered parts to move loads independently.
“As channels to market have started to get really dynamic, we’re starting to see activity that could be handled at the DC manually now being moved back into the manufacturing plant, in a production line,” Mulder says.
Fast or flexible?
In operations, especially in packaging, there often is a basic choice to be made between flexibility and high throughput. More of one means less of the other – at least according to conventional wisdom This problem was intensified during the pandemic, when processors had to both make quick pivots between SKUs, and run some of them extra-long, to meet heightened demand.
For decades now, equipment manufacturers have made progress against the flexibility vs. throughput dichotomy by incorporating automatic changeover, often using servomotors to reposition machine parts. But servo-driven changeover is expensive, and in any case, changeover is only part of the time loss involved in putting out multiple SKUs. The basic truth is that the more sizes a machine can accommodate, the slower it’s going to execute any one of those sizes.
Mulder gives the example of a cartoner with a pitch, or distancing between units, of 12 to 15 inches. The pitch has to accommodate the width of the carton plus a few inches extra. If a trade customer like a club store comes along and wants bigger cartons, all of a sudden that cartoner has to have a pitch of 18 inches. If the conveyor is running at a constant, maximum speed, the wider pitch will slow down the cartoner, which will increase the fill time for any smaller cartons.
Robotics, while initially as expensive or more so than servos on machinery, has the potential for maintaining both high throughput and flexibility, observers say. There are two ongoing developments that are making robots more versatile and able to add value at multiple points.
Vision systems are getting better, both in terms of what they’re able to recognize and how they coordinate with other equipment. And collaborative robots, or “cobots,” are designed with limitations on their range and speed of motion that allow them to work safely near employees.
Other emerging technologies have the potential to bring another dimension to processing and packaging. One such is what might be called beltless conveyors. These are material handling systems that rely on autonomously powered parts to deliver a payload.
Rockwell offers such machine components, which Mulder refers to “independent carts,” on its iTrak material handling system. They use windings of the type found inside electric motors, but flattened out, to create an electromagnetic field that interacts with a track, propelling the part. The result is a machine that can vary the timing of its material handling instantly, speeding up and slowing down as needed.
A similar development comes from Beckhoff. Its XTS and XPlanar transport systems carry payloads with tiles that “float,” suspended by magnetism. They travel along a physical track with XTS and with more freedom with XPlanar.
Beltless conveyors have the potential to alleviate one of the thorniest problems food companies face during and, especially, after the pandemic: maintaining social distancing. For instance, if animal carcasses moved down a cutting line independently, instead of being attached to a belt or hanging from a chain, they could in theory be sped up from station to station, then slowed or stopped as needed.
More pandemics coming?
Whether and how to establish permanent distancing is one of the things food companies will have to decide as the pandemic recedes. One of the biggest factors in long-range planning is whether those plans should include another pandemic.
In a survey by AIB International, 78% of food and beverage executives said they are actively preparing for a future global pandemic, with 30% expecting another one within the next four years and 50% expecting one within the next decade.
“I do believe that companies will all be talking about and considering appropriate responses to disruptions in the industry that include a future pandemic,” says Bank of America’s Maddox. “From conversations with our customers over the years, this has already been going on and will continue to.
"What is different now is that we have a real example to base it off of and consider in future planning," he continues. "I’m anticipating higher capital expenditures over the next few years as companies continue to invest in machinery that will serve multiple purposes.”
Even where distancing is not possible, automation—especially later-generation technology—still has the potential to save labor. Artificial intelligence can crunch and compare dozens of variables in a process to correlate them with deviances that lead to product flaws—in effect, getting to the root of a problem much quicker than humans can. The Industrial Internet of Things (IIoT) opens up new vistas for equipment monitoring and maintenance, allowing real-time vigilance with less human intervention.
The pandemic, in other words, is giving processors an extra incentive to go farther along the path they’ve already taken.
Says Neurala’s Versace: “What we have seen from manufacturers at this point is a tendency to inject additional automation everywhere it’s possible, and be sure that if another event occurs, they have their own vaccination, if you will.”