AICC Members On AI (CONT’D FROM PAGE 1)
For Landaal Packaging, the starting point was to estab- lish structure around data management to organize the massive amounts of information coming in on a daily basis. “There’s a ton of data servicing our core businesses, so we needed to figure out how to frame it, who should be in charge of it, and how to launch results in a structured way,” shared Landaal.
AI can be significant,” explained Gopakumar. “So it makes sense to focus first on the areas of the business with the greatest growth potential. For us, that was aftermarket. There’s a lot of opportunity for improvement, and if every customer touchpoint is supported by the latest technolo-
David Wiens, Founder & CEO, BPS AI Software & AI & MIS Manager, Danhil Container, moderated a panel of AICC members that featured, from left, Davo Laninga, Corrugated Product Specialist at ePS; Mark Landaal, Vice President of Sales and Marketing at Landaal Packaging; Jeremy Stuckey, Commercial Sales and Marketing Manager at Saica Pack; and Gokul Gopakumar, Vice Presi- dent Technology and Business Development at SUN Automation.
gy, there’s a two-fold benefit: our internal team saves time, and the customer gains a better experience.” Laninga, an 18-year industry veteran, described a sim- ilar approach. His team began by embedding AI into their ERP and scheduling platforms. “We started with sched- uling,” he said. “AI components track orders through the plant, analyze scheduled products, and learn from them. So the next time the project gets scheduled, we’re smarter about it.” One of their first practical projects was corrugator roll stock usage analysis. By running historical data through AI models, ePS was able to predict roll stock demand more accurately. The improvements were clear, and even small changes added up quickly. “A little amount of effort, say ten minutes, can translate to thousands of dollars in sav- ings,” Laninga noted. ePS has also been piloting conversational AI tools with- in its ERP software. Operators can provide input like, “I’m not sure how to do this task,” and get instant guidance pulled from documentation and training resources. Lanin- ga believes this access to the help system will be partic- ularly valuable for onboarding new employees who may not have years of experience. That challenge of retaining tribal knowledge was exact- ly what led Saica Pack to explore AI. As a global company with its U.S. operation still in its early stages, the compa- ny faced the challenge of losing critical expertise when employees left. “When you lose a tenured employee, you also lose their knowledge and expertise, and it’s becoming harder to replace,” said Stuckey. He and his team identi- fied knowledge gaps by consulting department managers across the organization, including customer service, de- sign, and business development, then using AI to fill those holes to support training and onboarding processes.
To start, the company formed an internal AI taskforce in April 2025, comprised of executives, leaders from op- erations and HR, and a technical lead, to create a detailed roadmap for execution. “The important part is to have a disciplined approach and take it one step at a time,” Lan- daal said. The Challenges With AI, there’s no ‘one size fits all’ approach. Regard- less of a company’s entry to this technology, the panel- ists agreed that challenges should be expected. The big change in going down this path is changing identity, shares Gopakumar. It’s a bit alien, and probably going to be a bit slow and bumpy. It’s a change of mindset all the way through. Aside from cost being a strong consideration for adopt- ing any new technology, hurdles for AI adoption in cor- rugated lie in the accuracy of data used to educate the system, the ability to manage organizational change, and the responsibility to safeguard sensitive information. Data Quality Laninga noted that predictive systems are only as ef- fective as the information they receive. “If you don’t have quantifiable, verifiable data, you may wind up trying to solve problems that don’t even exist,” he warned. Gopakumar elaborated, sharing a common challenge associated with downtime reporting. He noted that when a machine goes down, the operator’s priority is to get pro- duction running again, not to log the exact number of min- utes lost. “By the time they go back to record it, the details may be wrong or forgotten,” he explained. “For a person, a five-minute discrepancy might not matter, but for a ma- chine trying to understand and learn performance, it can throw off the entire calculation.”
CONTINUED ON PAGE 30
28
www.boardconvertingnews.com
August 25, 2025
Made with FlippingBook - Online catalogs