Solving the cutting stock problem: driving over $20M in profit

HIGHLIGHTS:

Problem:

A manufacturing company was handling the cutting stock problem locally at plants, focusing only on minimizing waste. This narrow approach conflicted with global supply chain goals and excluded machine constraints from strategic planning, leading to suboptimal results across the network.

Solution:

SimWell used anyLogistix for cutting stock optimization. By combining bill of materials modeling, ERP data pipelines, and Python-generated cutting combinations, anyLogistix evaluated thousands of feasible production patterns to balance demand fulfillment, waste reduction, and cost efficiency.

Results:

• Delivered over $20 million in annual profit improvements.
• Achieved a 5% average waste reduction across the network.
• Enabled data-driven strategic decisions with a digital twin supported by cost-to-serve visibility.

INTRODUCTION: FROM AN OPERATIONAL PROBLEM TO A PROFIT-MAXIMIZING TOOL

SimWell, a global consulting firm specializing in supply chain optimization, previously helped a North American manufacturer achieve a 3,700% ROI with a freight planning tool. Building on the success of that innovative project, SimWell continued exploring new ways to further enhance supply chain optimization.

In the next phase, the consulting company worked with the client to address a long-standing challenge: the cutting stock problem. Typically managed at the plant level, this problem involves cutting large rolls of material into smaller sizes while minimizing waste.

The objective was to integrate the cutting stock problem into a broader supply chain optimization model. By embedding production constraints into a digital twin built in anyLogistix, the project aimed to align operational realities with strategic planning.

PROBLEM: A NARROW APPROACH TO THE CUTTING STOCK PROBLEM

In most organizations, cutting stock is addressed at individual plants and machines, focusing only on minimizing waste. This narrow approach leads to:

• Conflicts between local cutting schedules and global supply chain goals.
• The exclusion of physical constraints (machine widths, cutter counts, grade compatibility) from higher-level planning.
• Missed opportunities to connect operational waste reduction with network-wide profit maximization.

In this project, the client operated multiple plants across North America and managed thousands of Stock Keeping Units (SKUs). This complexity made traditional approaches to cutting stock optimization unscalable, and local solutions often resulted in globally suboptimal outcomes.

The client’s network of factories and specific constraints at each (click to enlarge)

SOLUTION: BALANCING DEMAND FULFILLMENT, WASTE MINIMIZATION, AND SUPPLY CHAIN COSTS

Previously, SimWell identified the digital twin development as the next stage in advancing their client’s supply chain strategy. So this time SimWell built a digital twin in anyLogistix, supported by Python-based ETL pipelines connected to the client’s ERP system.

The cutting stock optimization process started with understanding how parent rolls are cut on each machine, considering width limits, cutter counts, and grade compatibility. The bill of materials (BOM) structure can be described as follows:

• Large parent rolls are split into main products and by-products.
• By-products can include other demanded products.
• Any remaining portion is treated as waste (with no matching demand).

A cutting stock scenario: dividing a large parent roll into smaller product sizes, waste is shown in black (click to enlarge)

To capture operational realities, thousands of feasible cutting combinations were generated through Python-based ETL processes. These combinations were then loaded into anyLogistix.

Solving the cutting stock problem: production and BOM setup in anyLogistix (click to enlarge)

anyLogistix evaluated all possible BOMs and chose the combinations that best balanced demand fulfillment, waste minimization, and supply chain costs. Because cutting stock optimization was the focus of the network optimization experiment, the model provided solutions that were both strategically optimal and operationally feasible.

The digital twin also supported scenario analysis, enabling the client to test various strategies and assess their financial implications. For example:

• Demand shifts across customer segments.
• Facility shutdowns or relocations.
• Customer portfolio adjustments to focus on the most profitable accounts.

With cost-to-serve visibility for every SKU, plant, and customer, the client could make confident decisions about network design and customer prioritization.

RESULTS: OVER $20 MILLION IN ADDED PROFIT

The project on cutting stock optimization delivered measurable impact:

• Annual profit increased by over $20 million compared to baseline scenarios.
• A 5% average waste reduction across the network, achieved by adding waste into the profit-maximization objective.
• Smarter strategic decisions enabled by complete cost-to-serve transparency and scenario planning.
• A repeatable and scalable process, adaptable to changing demand and operational conditions.

By solving the cutting stock problem at scale, SimWell demonstrated how anyLogistix empowers companies to merge operational constraints with strategic goals, driving both efficiency and profitability.

This case study was presented by Ershad Jahagirdar and Jean-Daniel Mathieu from SimWell at the anyLogistix Conference 2025.

The slides are available as a PDF.