A neural network adaptation on neutrosophic triplets for robotic assembly line optimization in smart manufacturing

Abstract

The decision-making process in smart manufacturing often involves complex, multi-criteria scenarios characterized by uncertainty and conflicting objectives. Traditional decision-making approaches face inherent limitations in managing indeterminacy, ensuring robustness, and addressing computational complexity, which compromise their reliability in dynamic manufacturing environments. This study introduces an innovative framework that integrates the VIKOR method, neural networks, and Neutrosophic Triplets (NTs) to address these challenges. The proposed approach is specifically designed to optimize robotic assembly line configurations by balancing key objectives such as cost, operational efficiency, and sustainability. VIKOR's compromise solution methodology is leveraged to evaluate trade-offs between group utility and individual regret, while Neutrosophic Triplets enhance the management of indeterminate information. Neural networks provide scalability and adaptability, enabling dynamic ranking refinement and reducing computational overhead. Additionally, a ranking strategy based on occurrence pattern analysis ensures robust and reliable decision-making outcomes. Validated through a case study on robotic assembly line optimization in a smart manufacturing environment, the framework demonstrates its effectiveness in improving productivity, adaptability, and sustainability. These results position the smart VIKOR method as a powerful and scalable solution for addressing the complexities of modern manufacturing systems.