An enhanced multiobjective inventory routing model to meet sustainable goals for assembly supply network under uncertainty

This study proposes an enhanced multiobjective assembly inventory routing model to assess the impact of sustainability practices on supply network performance goals. The model considers supplier incentives, supply risk, carbon emission penalty, heterogeneous fleet configuration, and demand uncertainty. Specifically, the model has economic (e.g., minimizing network cost), environmental (e.g., minimizing emission penalty), and social (e.g., maximizing supplier incentives) goals. The model incorporates a supply risk reduction policy. The novelty of this study lies in its simultaneous consideration of diverse factors in an inventory-routing context. Data sets from the existing literature are used to validate the model across various problem sizes. A modified hybrid non-dominated sorting genetic algorithm-II (HNSGA-II) is proposed to determine Pareto solutions and compare them with those derived from a speed-constrained multiobjective particle swarm optimization algorithm (SMPSO). HNSGA-II outperforms SMPSO in several critical performance metrics. The study further explores the impact of incentive schemes, low-risk supplier prioritization, and fleet configurations on sustainability performance. This study demonstrates a factorial experimentation-based sensitivity analysis on four objectives. The findings reveal that a heterogeneous fleet configuration can reduce emission penalties. However, this can result in increased network costs. A combination of low- and medium-duty vehicles is also recommended to attain economic and environmental efficiency. Service-level-based supplier incentives are found to enhance supply reliability and reduce shortages. However, this can elevate network costs and emissions. In scenarios of high demand variability, supplier incentives can ensure reliability. Conversely, cost and emission reduction can be prioritized over maximizing supplier incentives in high-supply risk scenarios.