Axium Foods, a mid-sized manufacturer of salty snacks, had to evolve quickly when it found out that its high-speed packaging equipment wasnāt flexible enough to meet new customer demands.
The company offers co-packing, contract manufacturing and private label services in addition to producing its own tortilla chip brands. And, while theyāve diversified their offerings over the years, the challenge came from finding new ways to deliver uniquely packaged products to consumers.
The retail business changes fast. And, while itās hard to predict what consumers will want in the future by walking through a grocery store, Axiumās private label customers could anticipate the need for mixed products and a variety of carton types for use on shelves and displays in the store.
That variety meant flexibility on the packaging line, something Axium didnāt have at the time.
āHigh speed packaging is set up to do one spot with one product in one case. It is difficult to deal with retail that has multiple products in a package,ā said Jerry Stokely, former president and current board member at Axium Foods. āWe needed to adapt to pack two or three different products in one carton, but that didnāt exist.ā
Stokely, who was speaking to attendees at PMMI Media Groupās Automation Conference & Expo this week in Chicago, said that he couldnāt find any case packer that could meet these needs. āNot only did they not have what I needed, but they werenāt interested in listening to what I wanted.ā So, Stokely made a decision that was based on the companyās successful manufacturing history that spans almost 60 years. āI decided we needed to make our own.ā
The company employs about 150 people at its South Beloit, IL facility in which there is a lot of existing automation that was built in-house to run the plant 24 hours a day between three shifts.
āWe had a history of manufacturing our own custom equipment. We also had the financial resources, a supportive management [team] and highly skilled controls engineers,ā Stokely said. And, they had adequate in-house fabrication and access to outsourced fabricators for doing off-site development.
The need for a high degree of flexibility underlined everything they did as the team moved forward designing a case packing line. It had to accommodate a lot of changeovers and specialized short runs. In addition, it had to fit in the existing footprint, be operator friendly, mechanical friendly, intuitive and reliable, and include a six-axis robot.
A six-axis robot would provide scalable payloads, a definable reach envelope, multitasking and subtle and complex motion. āBut the traditional robots have proprietary programming languages, a long learning curve, large footprint and a large investment,ā Stokely said. A collaborative robot (cobot), on the other hand, āhas an intuitive programming language, a short learning curve, is cage-free and is a modest investment.ā
So a cobot it was. Stokely chose theĀ Universal RobotsĀ UR5 with 5 kgs payload and an 850mm reach as snack food is light, so payload was not as important as reach. While designing the cobot into the workflow the company spent a lot of time figuring out how to build a system capable of running itself, even during a changeover.
A key learning during the trial and error phase was that to run at medium-to-high speed, the robot had to function as a component of the system. And, every movement had to be broken into a sub-function with every step perfected along the way to create a total system. And, it wasnāt easy. āWe built a catalog of what didnāt work,ā Stokely said.
The other key learnings: Control systems outside of the robot are critical. And as the Internet of Things moves forward, interoperability could be an issue as, right now, a lot of control systems and plant floor equipment canāt talk to each other. The other key learning came from end effectors. Whatever you are picking up or grabbing, thereās an end effector for it, and itās easy to lose time searching for the right one, Stokely explained. āYou have to educate the fabricator on your business needs and clear their mind of preconceptions.ā
Lastly, focus on the humans in the process. It may seem that the guy who programmed the system would be the best person to train others, but thatās not always the case. āFrankly, that was a major failure for us,ā Stokely said. āFind a person who is a natural teacher and give them permission to ask stupid questions.ā
The result of all this work is that Axium is now a fully-functional and reliable multi-tasking case packer. And the use of cobots was what enabled them to change the status quo.
āWe are a relatively small snack food company that built our own robotic case packer because there wasnāt anything else out there,ā Stokely said. āIn the process, we had a lifetime of learning and can see the potential to go beyond the basics.ā

