Stochastic Collaborative Supply Chain Optimization With Fair Hierarchical Gain-Sharing
Horizontal supply chain collaboration (HSCC) offers substantial cost savings and carbon footprint reductions, yet commercial adoption remains hindered by non-scalable gain-sharing mechanisms and operational disruption risks. This study presents a scalable two-stage hierarchical framework for post-Pareto selection, integrating a two-stage stochastic mixed-integer linear program (MILP, Level 1) with an Equal Profit Method linear program (EPML, Level 2) to optimize multi-objective routing and equitable profit allocation across multi-partner logistics networks. Evaluated on a medium-scale benchmark network (15 partners, 8 demand nodes, and 50 stochastic scenarios), the framework achieves an alliance-wide cost reduction of 18.42% and a carbon reduction of 14.10% relative to non-cooperative baselines. At Level 2, the EPML engine executes in 0.32 s for 15 partners and 1.87 s for 30 partners, overcoming the exponential computational bottlenecks of traditional cooperative game theory methods while theoretically guaranteeing expected individual rationality $\left(E\left[\pi_i\right] \geq \pi_i^0\right)$, where $E\left[\pi_i\right]$ denotes partner $i$'s expected collaborative profit and $\pi_i^0$ denotes that partner's standalone baseline profit—provided that the selected physical solution satisfies the net-synergy feasibility condition $(\sum_{i \in I} E\left[\pi_i\right] \geq \sum_{i \in I} \pi_i^0)$, where $I$ is the set of alliance partners. The framework also achieves empirically verified core stability, with $e_{\max} \leq 0$, where $e_{\max}$ denotes the maximum excess across the tested subcoalition. Sensitivity analysis demonstrates robust customer fulfillment (99.59% service level) under high shortfall penalties $\left(p_j^{\text {pen}} \geq 2.5 \times p_j^{\text {sell}}\right)$ and a marginal abatement cost (MAC) of \$2,318.68 per metric ton of carbon dioxide equivalent (tCO$_2$e) at executive compromise configurations. To facilitate industry adoption, we translate these algorithmic outputs into a practical 4-step managerial decision protocol balancing ex-ante contractual rules, ex-post financial settlements, and alliance-level greenhouse gas (GHG) reporting boundaries. The framework effectively resolves the trade-off between allocation fairness and computational scalability in stochastic environments, providing supply chain decision-makers with a practical, mathematically rigorous, and computationally efficient post-Pareto decision-support tool to establish stable, equitable, and resilient horizontal logistics coalitions.
Authors
- Asma Sbaï (ORCID: https://orcid.org/0000-0003-0566-5004)
- Malika El Kyal
- Hanan Ouhader (ORCID: https://orcid.org/0000-0001-9694-9242)
Publication Details
- Journal
- Journal of Engineering Management and Systems Engineering
- Published
- 2026-11-19
- DOI
- https://doi.org/10.56578/jemse050407
- Primary Topic
- Supply Chain and Inventory Management
- Type
- article
- Field-Weighted Citation Impact
- 0.00