AI Before the Cut – Smarter Design for Corrugated

November 5, 2025

When it comes to AI advancements, the corrugated industry spends enormous energy talking about predictive maintenance, scheduling algorithms, and smart sensors on the plant floor. These technologies matter, but they come with long timelines, expensive integrations, and data challenges.

Meanwhile, the biggest cost in our industry often gets less attention. Choosing board combinations is still driven by experience, McKee’s formula, and conservative lab testing. That means boxes are frequently overengineered “just to be safe,” burning through fiber, margin, and sustainability goals.

The next wave of quick wins may not come from another shop floor upgrade but from AI applied earlier in the process: at the design stage, before the first sheet is ever cut. Here, AI is showing real potential to optimize board recipes, propose novel configurations, accelerate customer sampling, and even test designs virtually.

Engineering the Right Recipe

Ask any designer: picking the right liner and medium combination is equal parts science and gut feel. But what if AI could eliminate the guesswork?

Recent research suggests it can. A 2025 study from Kasetsart University trained a backpropagation neural network on 630 commercial box samples with 17 design and material inputs. The model predicted box compression test (BCT) results with an R² of 0.982, far outperforming McKee’s formula at 0.737. That is not a minor statistical improvement. It means AI can pinpoint the lightest paper grades that still deliver required strength.

The business case is obvious. Instead of overspecifying to avoid failures, converters could use data-driven predictions to specify just enough paper. For a high-volume plant, even a 2%–3% reduction in fiber use translates into significant savings and sustainability gains.

Beyond Standard Flutes

Most converters select from a familiar playbook: B, C, and E flute single-wall or a handful of double-wall options. But AI has no allegiance to the catalog. Algorithms can explore hybrid or unconventional configurations such as pairing recycled liners with virgin outers, mixing microflutes with larger flutes, or even proposing entirely new flute geometries.

Academic work from Poznań University of Life Sciences has already shown that AI can evaluate variations in flute angle, curvature, and wall layering to predict strength and cost trade-offs. Instead of defaulting to 200# C flute, a future AI co-pilot might recommend 175 gsm recycled medium paired with 200 gsm kraft outer liner and an F flute micro layer for stacking strength at lower cost.

The point isn’t that AI replaces engineers. It’s that it broadens the design space, giving humans new material recipes that might never be considered otherwise.

These advances at the corrugator level highlight how AI is reshaping material choices and board construction. But design innovation doesn’t stop at the paper recipe. Once the board itself is specified, converters still face the challenge of turning those materials into box designs quickly and cost-effectively for customers. That’s where AI is beginning to transform the design office through automating drawings, accelerating sampling, and reducing the time it takes to get from concept to customer approval.

Faster Sampling and Customer Drawings

Every converter knows the pain of customer sampling. A prospect asks for a new shipper, and the design team spends days drafting dielines, cutting prototypes, and iterating until the buyer signs off. AI is attacking that bottleneck.

Tools such as Sourceful and Pacdora now generate dielines and 3D mockups automatically from basic dimensions. Sourceful even markets its AI as “your personal packaging designer,” delivering print-ready dielines by combining AI-generated concepts with expert human refinement.

Applied to corrugated, this means a customer request could trigger an AI agent that drafts the drawing in minutes and hands off to the team for finalization. Designers still review and refine, but they’re no longer redrawing rectangles all day. The payoff is obvious: shorter sales cycles, faster customer response, and design teams focused on high-value creativity instead of repetitive CAD work.

Virtual Testing Without the Waste

Design is only half the battle. Testing strength and durability is just as imperative for a plant. Traditionally, every iteration requires physical samples, compression tests, and sometimes drop or vibration cycles. That’s time, labor, and wasted board.

AI is making those cycles virtual. Studies show neural networks can predict edge crush test and BCT results for boxes with cutouts, hand holes, and perforations with R² values around 0.97. In other words, the AI “knows” how structural quirks will affect strength.

AptarGroup demonstrated the value in practice. Partnering with Monolith AI, the company built a self-learning model to predict bottle stability in seconds, slashing the number of prototypes required. The same principle applies to corrugated. An AI tool trained on past compression and drop tests could estimate stacking strength, or likely failure modes, of a new box before anyone sets up the CAD table.

That doesn’t eliminate the need for final lab validation, but it drastically reduces the number of prototypes and accelerates time to market.

Quick Wins vs. Long-Term Vision

The beauty of design-side AI is that practical benefits are available now:

Predict paper requirements more precisely.
Generate dielines and 3D previews instantly.
Reduce prototype waste through virtual testing.

These don’t require full-scale plant integration or years of collecting sensor data and training models. They are modular tools a converter can begin piloting today.

Longer-term, we may see generative AI proposing entire board configurations, novel flute profiles, or hybrid double-walls engineered for specific loads. Think of it as a board-grade advisor that sits alongside your design team, suggesting optimized recipes in real time.

Smarter Boxes, Not Just Smarter Machines

AI in corrugated doesn’t have to mean costly sensors or complex scheduling models. Sometimes the most immediate gains come upstream.

By adopting AI for board optimization, dieline generation, and virtual testing, converters can save paper, speed sampling, and cut development costs while improving responsiveness to customers.

AI isn’t set to replace skilled designers anytime soon, but it is already beginning to equip them with a better arsenal. For an industry facing margin pressures and rising material costs, that’s a technology shift worth embracing.

Don’t just think about smarter machines. Think about smarter boxes, designed with AI before the cut.