Artificial intelligence-based approaches to sequencing and balancing problems in disassembly lines: A systematic literature review

In accordance with the principles of sustainable manufacturing and the circular economy, the efficient recovery of end-of-life (EOL) products has become increasingly critical. Disassembly systems, in this context, face two primary optimization challenges: the Disassembly Line Balancing Problem (DLBP) and the Disassembly Sequencing Problem (DSP). Due to their multi-constrained and multi-objective nature, these problems are classified as NP-hard, posing significant challenges to conventional solution methods. This study provides a systematic review of 233 publications, including 223 research articles and 10 review papers, addressing artificial intelligence (AI)-based approaches for DSP (2000–2025) and DLBP (2004–2025). The studies are systematically analyzed and categorized according to multiple criteria, such as product type, constraints, objectives, line configurations, disassembly level, disassembly processes, data requirements, and AI applications. The results indicate that AI-based hybrid and machine learning methods offer more flexible, effective, and practically feasible solutions compared to traditional deterministic models. This review provides a comprehensive comparative analysis of existing research and outlines a methodological roadmap for future studies, thereby highlighting prevailing gaps in the literature.