Food Engineering
Robotics

Is there a robot in your line’s future?

Robots aren’t all things to all people, but in the right place, they can be a big time and labor saver.

August 7, 2013
The Schneider robotic vertical carton loader

From a machine builder’s point of view, robots can solve many processing/packaging line problems, but at the same time, there are certain applications where conventional equipment performs just as well—if not better. For example, a candy bar packaging line has been making the same bars for years with few changes to the product or the packaging line. Conventional equipment has wrapped the bars and filled the cases for years—so why reinvent the wheel?

But with today’s fast-paced and short product lifecycles, candy, snack and nutritional bars may be here today and gone tomorrow; their sizes and shapes change, their labels change according to the art director’s new design or changing allergen information, and the number and variety of bars going into cases change as well. So why use robotics?

“We use robots in packaging for multiple reasons,” says Terry Zarnowski, director of sales and marketing at Schneider Packaging Equipment Company, Inc. “For us, the robot is nothing more than a highly flexible, repeatable and adaptable tool we use to get the job done. For our customers, it means they are getting what they want—flexibility and high reliability with fast changeover times.”

These abilities add up to an increase in overall equipment effectiveness (OEE), which improves a processor’s bottom line, adds Zarnowski. “Before we used robots, we had built a conventional [servos, air cylinders, etc.] case packer for packing a product made in multiple sizes and packaged in multiple formats—literally hundreds of SKUs.” Built for a copacker, the machine used conventional automation to manipulate the product and get it into the correct orientation and pattern to load into the case, making the machine large and complex, with multiple change parts and manual adjustment points that could take a half-shift to change over. Since all the conventional hardware has been replaced with robots, the copacker has seen a big improvement in OEE, primarily due to streamlining changeovers to less than 15 minutes and making changeovers accurate and repeatable, reports Zarnowski. The copacker can now take on more short-run orders for more customers.

Variability in product is a challenge, but robots combined with vision systems have been stepping up to the plate for many years, according to Mike Wagner, global OEM business manager/packaging industry, Rockwell Automation. Wagner points to a pork belly slicing application installed at Hormel many years ago. People used to do this operation manually with a U-shaped knife, which was not only dangerous because of possible cuts from the blade, but also because of carpal tunnel injuries. Now, an eight-axis servo/robotic system, teamed with a vision system, accurately keeps the knife in the fat layer—but close to the meat—making perfect lean cuts and increasing production. At the same time, worker safety issues have decreased.

While pick-and-place robots and pallet stacking robots work well in some applications, they’re not always problem-solvers for not-so-obvious reasons. This is where careful consideration of applications comes into play. In many cases, robots take up less space and can decrease the number of conveyors needed, but not always. Obviously not a food application, a bullet cartoning application at an arms manufacturer went out to bid as a robotic system and a conventional cartoning system, according to Wagner. The number of robots needed to pick up randomly placed bullets on a conveyor and place them into a box at the rate of 1,000 per minute would have occupied up to 4,500 sq. ft., whereas a dedicated cartoning machine (which won the bid) used only 650 sq. ft. of floor space.

With a dedicated line where only one candy bar size is run, and a processor produces large quantities of a single product, it’s often easier to use traditional conveying technology for feeding flow wrappers, says Paul Garms, Bosch Packaging Technology product marketing manager. “We still install product distribution conveyors in many different bar applications for feeding wrappers simply because they can achieve higher speeds than a robot for feeding a flow wrapper, especially when there isn’t much product variation.”

Similarly, KP Foods at Stockton-on-Tees, UK, a manufacturer of Potato Crisp snack foods, considered a typical four-axis robot palletizing solution to handle one box every four seconds. This would have required up to eight robots to keep up with the cartons. Rather than dedicate a large amount of floor space for these robots, Güdel installed two FP-5 standard three-axis portal gantry systems with a custom-designed gripper system to handle the range of 16 different boxes and load new empty pallets into the system. The installation saved valuable floor space for other activities.

Dos and Don'ts for successful integration projects

 

Do:

1. Look beyond direct labor savings to understand true ROI. Consider lifecycle costs, changeover times, accommodation of future projects, uptime, etc. compared to top alternatives.

2. Investigate your choice of an integrator thoroughly to verify it can do your project successfully. A good place to find an integrator is Robotic Industries Association (RIA).The RIA certifies integrators that have trained personnel, a proven track record and the ability to provide a solution that is safe and efficient.

3. Do not over-spec a line relative to throughput. Unrealistic expectations (e.g., double or triple the actual throughput) can lead to complex and expensive solutions. Be realistic in your throughput expectations.

4. Understand the possible exceptions that can occur in the process and how the robotic system should respond (e.g., fully automatic recovery, manual intervention, continuation from current state or reset and start over).

5. Be sure expectations are set and documented.

6. Robotics allows greater flexibility in equipment design; be sure to evaluate the solution from each bidding vendor since the solutions are not the same.

7. Have a face-to-face meeting with the bidder to review in detail its solution.

8. Build ownership among your stakeholders (with clear communication of objectives, training, early involvement in design reviews, participation in run-off/acceptance testing, etc.)

 

Don’t:

1. Buy based on the price of the equipment itself. Cheaper does not mean better or faster return on investment.

2. Assume the robot is the always the best solution; sometimes the answer may be a conventional solution that will meet your goals at a lower cost.

3. Assume you know the right solution and demand the integrator use it. Allow the integrator to be creative and use its experience to develop what it believes is the best solution based on your detailed requirements.

4. Be too ambitious with your first project. Walk, then run.

Slippery, hard-to-pick-up items such as sliced meat and bacon are becoming good applications for robot handling. A new system developed by AEW Delford Systems has the ability to pick up sliced meats and deposit them onto trays, ready to go to the thermoformer for packing. Operating with a vision system to recognize the positioning and orientation of portions, the robot portion loading system (RPLS) has a specially designed end effector that makes it possible to gently grip the meat and transfer it to the tray.

 

Fringe benefits

In the RPLS described above, no human comes into contact with the meat from slicing to wrapping/packaging, eliminating one area of contamination. At the same time, the machine’s vision system has the ability to identify incorrectly sized or oddly shaped portions, rejecting them if necessary or grading them into various quality cuts.

“Our systems typically include vision, and we are now doing more and more product/package inspection,” says Craig Souser, JLS Automation CEO. “In addition, we have customers telling us rework has gone down as the robot is gentler to the product than the operators were.” Souser also reports processors are more interested in gaining statistics on reject rate, nature of rejects, etc., Vision systems can accumulate and report that data. Food safety is another big driver, with proper sanitation as a given; robots don’t get head colds or forget to wash their hands, and they don’t get tired.

Cost savings are paramount to processors, according to Bob Rochelle, food & packaging industry specialist at Staubli Robotics. With robotics-based packaging systems, traceability is easier in the event of a recall, and the systems allow changes to be made on the fly such as weighing, labeling and packaging. Accuracy is improved in all these steps because humans aren’t involved. Robotics can make the job of sanitation a little easier, as well, because there are typically fewer surfaces to clean, especially when robotics reduces the need for excess conveyor equipment that must be washed. In addition, robots generally help reduce energy consumption, says Dick Motley, FANUC Robotics senior account manager. As with sanitation, it’s reasonable to assume that because robots can reduce the amount of required conveyor footage, fewer conveyors equal less energy used. Plus, robots are generally more tolerant of exceptions and have a faster recovery than non-robotic systems. Speaking of flexibility, robots are easier to adapt to future products.

Motley lists several applications in the food industry for which robots are suitable. These include palletizing (cases, bundles, cans, plastic and glass containers, pails), stretch wrapping, case packing, carton loading, form-fill-seal load and unload, blister machine load and unload, pouch handling and packaging, bagging, processed meat handling, cheese processing, frozen food handling and palletizing, produce handling/sorting/packaging, product assembly, product inspection and sorting, dairy cattle processing and pork carcass handling.

 

Jumping the hurdles

There are several hurdles to getting started with robotics; some are perceived, and some are connected with the physical plant, electrical connections and communications (networks). But fear of the unknown remains the largest hurdle, says Souser. Many prospective users perceive robotics to be so complicated they can’t support it. The other looming issue is ROI; it can be difficult to get the 12-month payback some processors want with automation.

“Robophobia” is what Zarnowski calls it: The processor hasn’t used robotics before and is afraid of adopting the technology. “Often they believe they may have inadequately trained operators and technicians to keep the lines running when, in fact, if the integrator has done a good job of integrating the robot, it rapidly becomes [the processor’s] friend due to the simplicity of operation and maintenance.” One plant that had never used robotics before was very resistant, so a naming contest was held before the equipment’s arrival. Plant personnel named the two robots Lucy and Ricardo. That line became the most sought-after to work on and is now the best operating line in the plant.

“Often it comes down to a matter of cost, as there is a perception that a robotic solution is more expensive,” says Zarnowski. However, when processors are provided with and analyze the value proposition, the decision is usually a no-brainer, especially when the added flexibility and adaptability can extend the life of the existing capital investment, adds Zarnowski.

Besides the perceived obstacles, real-world plant considerations need to be considered. Motley lists a few:

  • Floor space—is it adequate?
  • Facility issues—foundation/anchoring, power, compressed air, etc.
  • Personnel safeguarding, especially if the workforce is unfamiliar with robotics
  • Employee training and ownership
  • Interface to upstream and downstream equipment—may be old/outdated
  • Product quality—especially when switching from a manual process, processors often discover their people are compensating /masking upstream process problems
  • Exception handling—how should the system behave when things inevitably go wrong, e.g., jams, out-of-specification product, etc.?
  • Selecting the right integrator—what company has the applicable expertise and resources?

Of these considerations, floor space is critical. “In many cases, there is minimal floor space to accommodate the requirements for safety fencing,” says Staubli’s Rochelle. This is due in large part to the quantity of people working in close proximity to the product and each other in today’s processing lines.

This problem may resolve itself to some extent when robots are integrated into the overall system. Rochelle points to a recent production line where robots were installed over an existing line. Eliminating the people and having the robots overhead actually reduced the floor space required. “Quite often with the use of a robot, you can eliminate space because the robot is capable of performing multiple tasks, which removes the need to move products from workstation to workstation. This reduces the amount of floor space required,” explains Rochelle.

One way to avoid any unpleasant surprises is to run a simulation software program before committing to an integrated packaging line. Usually this software can show all the what-ifs before integrating an old line or building a new one; it can also help processors plan for new products after the line is installed. “We have a software platform called Gemini, which provides both visualization and simulation capabilities for our delta-style robots,” says Bosch Packaging Technology's Garms. The software can be used to plan lines and show how a line will work with changing product specifications. The software was designed so processors can use it themselves; it also serves as a 3-D HMI.

 

Simplifying conveyor layout

Can robot technology simplify conveyor layout? “Yes. Robot technology has the ability to do multiple tasks,” says Bill Natsch, Intelligrated director, robot integration. “With conventional palletizers, you might need multiple machines to handle multiple production lines, while one robot can do the same task on two production lines.”

“Certainly in a palletizing application, the need for accumulation is significantly reduced [with robotics] vs. conventional alternatives, from whole layers or whole unit loads done to just a few cases needed for the next robotic cycle,” says Motley. Also, a robot’s ability to sort at the point of palletizing can reduce or eliminate significant investment in sortation conveyors.

“Using vision to track a robot and designing the system for ‘agility’ allow you to eliminate a number of guides and devices that might normally be present,” adds Souser. “We have customers tell us, ‘It looks like something is missing, but I know it’s not.’ A simple system is easier to clean, maintain and change over.”

One project for the defunct Hostess Corporation used two robots working side-by-side to put layers of Twinkies into cases, according to Rockwell’s Wagner. The Twinkies were collated and organized into the pack pattern, done outside of the robot. Once the Twinkies got into position on the in-feed, the robots took over. One picked up a layer and put it into the box; the second robot put a piece of wax paper between the layers. Then the first robot would put in another layer. The two robots worked with very human-like behavior in the same 3-D space without collision. They could do any pack pattern—just by changing the end effector, says Wagner. It was possible to move quickly from a four-up pack to a 16-pack to a 32-pack carton or case. The real magic, according to Wagner, was not so much the pair of robots, but the tooling—the end effectors.

 

Fingers and eyes

According to Wagner, the more weight an end effector has, the less efficient it is, and the bigger the robot must be to handle the combined weight and inertia of itself and the product. Truly, there is a point of diminishing returns where the weight of the end effector takes away the agility of the robot to move the product efficiently.

End-of-arm tooling is becoming lighter through the use of composites. And in some cases, end effectors have been created on 3-D printers, according to Rockwell’s Wagner. For example, one particular OEM has developed its own end effectors using CAD and 3-D printers. The benefit, according to Wagner, is that a tool can be designed, printed and air freighted all in a day’s time.

“End effectors are critical to the success of the robot, especially when you’re talking about picking raw or unpacked products for the primary feeding of machines,” states Garms. A number of advances have been made recently, especially in modeling airflow, allowing more fragile  and crumbly products to be picked than in the past. This also is due to better vacuum systems that can remove the crumbs and debris from the products.

In primary packaging, cleanliness is another issue as the tooling contacts the product. Souser reports vacuum cups keep being improved, and JLS Automation has been developing its own vacuum generators specifically for direct product handling.

FANUC has successfully integrated what it calls intelligent capabilities into its robots, says Zarnowski. Besides 3-D vision capabilities, force and accelerometer feedback allows robots to be aware of the products they handle. “Force feedback is used to enable gentle product handling, weight checking, etc.,” says Zarnowski. “FANUC also uses accelerometers to automatically ‘tune’ a robot to a specific application to improve throughput and power consumption.”

“Gripper technology is limited only by the imagination and expertise of various system integrators and end-users,” says Motley. “Standard vacuum, mechanical and pneumatic components can be complemented by application-specific details to make a reliable custom handling solution for a wide range of products.”

While there are applications where robotics function without vision (e.g., they pick up packages in the same position every time), integrated vision not only guides the robot, it identifies defective product, allowing the robot to respond appropriately, often avoiding jams and downtime, says Motley. In addition, as an inspection tool, it can confirm basic quality attributes. Recent advances in vision include color evaluation, hyper-spectral inspection (using light wavelengths outside the visible spectrum) and 3-D spatial analysis and guidance.

“Vision systems are faster, more accurate, offer more algorithm choices and are easier to set up and use,” says Zarnowski. “We use FANUC robots, which have vision integrated within the controller, eliminating the need to ‘glue’ a vision system from manufacturer A to manufacturer B, as well as eliminating integration headaches and having to go to different suppliers for support.” Vision systems are becoming a standard offering by most robot suppliers today.

 

Putting it all together

Until 2005, one of Hershey’s chocolate manufacturing plants was a hand-palletized operation, which had low palletizing rates and high manual labor costs. Alex Diaz, packaging systems engineer, describes the situation: “We had three to four people palletizing per line, per shift, three shifts per day. We’re talking about 12 to 15 people palletizing at once, depending on how complicated the patterns were.” Employees also did a lot of heavy lifting, with some boxes weighing more than 50 lbs.

With Hershey’s increasingly wide variety of product SKUs and packaging configurations came demands for more flexible, faster palletizing. Hershey worked with Intelligrated to develop an automated palletizing system that could handle its low- to mid-velocity SKUs, ranging from one to 40 cases per minute. All this had to fit within the plant’s 7000-sq.-ft. palletizing area.

Robotic palletizing enabled Hershey to send three different lines to one robot and palletize in three positions at once. While this reduced the amount of required conveyance, it posed a problem at first, in that the robots couldn’t keep up with the production rates of Hershey’s higher-velocity SKUs.

A partnership between Intelligrated and Motoman, Inc. facilitated the move to robotic palletizing for the first time at the plant. Major components for the new system included an Intelligrated/Alvey 780 case palletizer with slip-sheet dispenser, three Motoman robotic arms with UniGripper intelligent vacuum tools, several conveyors, a laser-positioned transfer car, motor-operated turntable, air-operated pushers and an ITW Muller Octopus stretch wrapper. Since the completion of the project in 2006, Hershey has seen increases in palletizing speed and efficiency, and has no major problems with the equipment or the system.

 

For more information:

Terry Zarnowski, Schneider Packaging Equipment Co., 315-676-3035, tzarnowski@schneiderequip.com

Mike Wagner, Rockwell Automation, 414-382-2000, mwagner@ra.rockwell.com

Craig Souser, JLS Automation, 717-505-3800, csouser@jlsautomation.com

Bob Rochelle, Staubli Robotics, 864-433-1980, b.rochelle@staubli.com

Paul Garms, Bosch Packaging Technology, 715-246-6511, sales.packaging-nrd@bosch.com

Dick Motley, FANUC Robotics, 248-377-7000, dick.motley@fanucamerica.com

Bill Natsch, Intelligrated, 866-936-7300, bill.natschintelligrated.com


Resources:

“Hybrid Palletizing Solution Hits Sweet Spot for Hershey,” Intelligrated, www.intelligrated.com, 2010.

“Using Robots to Solve End-of-Line Issues in the Food Industry,” FANUC Robotics, www.fanucrobotics.com/robotics-articles/Food_Robots_End_of_line_Solutions.aspx.

 Robotic Industries Association website, www.robotics.org