Existing Sales and Operations Planning (S&OP) models typically address production, pricing, and risk- management decisions in isolation, limiting their effectiveness under commodity price uncertainty. This paper presents a stochastic mixed-integer linear programming framework that integrates decisions across multiple levels—strategic (network design, supplier selection), tactical (multi-stage production, workforce allocation, inventory positioning), and financial (dynamic pricing with LME pass-through, commodity hedging)—within a unified optimization model for aluminum extrusion manufacturing. This multi-level integration enables coordinated responses that exploit cross-functional trade-offs unavailable to sequential approaches: building inventory during low-cost periods, adjusting prices dynamically with London Metal Exchange movements, and calibrating hedge positions to anticipated demand. Computational experiments demonstrate that the integrated approach improves expected profit by 14%, reduces margin volatility by 62%, and eliminates margin violations under extreme scenarios including pandemic disruptions and commodity price crashes.

Master Production Scheduling (MPS) remains a cornerstone of manufacturing planning, yet its effectiveness is often undermined by uncertainty arising from demand, supply, and production-side disruptions. Existing approaches typically address these uncertainties in isolation, limiting their robustness in dynamic industrial environments. This paper proposes a Modified Separated Estimation and Optimization (MSEO) framework that extends the SEO paradigm by explicitly incorporating production-side uncertainties such as capacity limitations, crew unavailability, machine breakdowns, and product changeovers. The framework integrates machine learning–based probabilistic forecasting with stochastic optimization in a rolling-horizon setting, enabling adaptive and resilient production schedules. By uniting forecasting and scheduling into a single decision-support system, the proposed framework provides a data-driven approach for uncertainty-aware MPS. The proposed MILP model was validated through a realistic numerical case study involving three products over five planning periods under multiple disruption conditions. The model achieved an optimal objective function value of $5398.60 with a computation time of 0.0121 seconds, demonstrating its ability to effectively adapt production decisions under disruption scenarios while minimizing profit loss. The proposed framework advances MPS research by enhancing operational resilience and scheduling robustness in modern manufacturing systems.

In contemporary supply chains, supplier selection has been extensively studied using Multi-Criteria Decision-Making approaches, while the evaluation of buyers from the supplier’s perspective remains comparatively underexplored. This study proposes an expert-driven multi-criteria decision support framework for bidirectional buyer–supplier evaluation that formalizes both supplier and buyer assessment within a unified structure. Industry experts from FMCG sector were surveyed to identify and weight evaluation criteria using direct weighting and Analytic Hierarchy Process (AHP) pairwise comparisons. The framework integrates AHP, TOPSIS, and VIKOR to comparatively evaluate ranking behavior under different organizational contexts. In addition, the model distinguishes between binary qualification criteria and weighted performance criteria to better reflect real industrial decision practices. The proposed framework is implemented within an interactive decision-support system and evaluated through multiple comparative

The proliferation of industrial information technology has made digital transformation and intelligent manufacturing central to modern manufacturing, with digital twin technology enabling real-time optimization and dynamic decision-making. Meanwhile, e-commerce growth has driven fragmented, small-scale, multi-batch uncertainly demand, forcing enterprises to balance delivery time windows, production-line flexibility, and sequence-dependent setups costs, thereby substantially complicating scheduling. To address this, we formulate a mixed-integer programming model for unrelated parallel machine scheduling aimed at maximizing total profit, under constraints including delivery time windows, sequence-dependent setup times, product heterogeneity, and machines capacity differences and so on. On the demand side, deep reinforcement learning is integrated for order forecasting, embedding demand into production planning to balance loads and optimize resources, thereby mitigating stockout and overstock risks and furnishing data-driven planning foundations for digital twin systems. On the solution side, an improved adaptive large-scale neighborhood search (ALNS) algorithm is designed based on the tabu search escape mechanism, with destroy and repair operators tailored to the model's characteristics to enhance efficient real-time solving for complex scheduling problems. Computational experiments validate the effectiveness of the proposed approach, demonstrating improved delivery reliability and capacity utilization, reduced costs and decision-making opacity, and a scientific basis for intelligent scheduling in manufacturing enterprises.