This critical review examines the transformative role of machine learning (ML) in revolutionizing thermoplastic recycling across mechanical, chemical, and biological pathways. As global plastic waste challenges intensify, sophisticated ML approaches are emerging as powerful tools to overcome traditional recycling limitations. Recent technological breakthroughs are systematically analyzed that leverage ML to optimize sorting precision, process efficiency, and quality assurance in recycled thermoplastics. The review presents a detailed analysis of feature engineering strategies that have proven most effective across diverse recycling applications. By identifying current implementation barriers and unexplored opportunities, a forward-looking research agenda is established for ML integration that can accelerate progress toward a truly circular thermoplastic economy. This interdisciplinary perspective bridges materials science, computer science, and sustainability to provide actionable insights for researchers and industry practitioners.
Albuquerque et al. (Thu,) studied this question.