Pattern Generation for a Sustainable One-Dimensional Cutting Stock Problem with Limited Usage of Unique Standard Lengths and Unique Patterns to Minimize Overall Trim Losses and Cost

As we progress into an era plagued by the climate crisis, where the survival of distributed-but-interconnected planetary ecosystems comes closer to risky tipping points forcing eternal oblivion, intelligent approaches to identify and mitigate drivers of such catastrophic change must be set forth and implemented swiftly. Effective waste management in this regard is a very important problem. A major chunk of the industrial wastes is generated during manufacturing processes. The first step towards reducing these is scrap optimization with the next phase leveraging unavoidable scraps for further utilization. Addressing a part of this problem, we focus on methods to reduce trim losses implementing sustainability in manufacturing. A deterministic heuristic approach is suggested to develop good cutting patterns to the classical one-dimensional cutting stock problem with multiple standard lengths, focusing on minimizing wastage. We assume any extra final cuts, above a threshold cut-length, may be stored in inventory for satisfying future requirements. We follow a step-wise optimization procedure of sequentially minimizing waste (W), cost (C), unique standard (S) length usages, and unique pattern (P) usages for a WCSP multi-level optimization idea; each minimization step being connected to the next problem as a constraint. We generate synthetic datasets (from an industrial problem, with altered data) and perform computational tests demonstrating the effectiveness of our proposed method to solve this problem.