Review on Aging of Biodegradable Polymers: Experimental Approaches and Emerging Role of Machine and Deep Learning

Abstract Biodegradable polymers are currently the focus of intense research due to their biodegradability and potential to replace conventional plastics. However, they still face several limitations: high cost, limited large-scale production, and faster decline of its properties over time, which compromises its performance and durability. Their complex molecular structures, combined with the strong influence of physical, chemical, and environmental factors, make their performance difficult to predict with traditional methods. These complexities highlight the need for advanced modeling approaches capable of capturing nonlinear and high-dimensional behaviors. Machine learning (ML) and deep learning (DL) offer promising tools in this context. Their application to biodegradable polymers can improve predictions of aging behavior and lifetime under different environmental conditions, guide the design of biodegradable polymers, identify optimal synthesis and processing parameters for extended service life and detect complex patterns in large datasets. This review discusses experimental techniques commonly used to evaluate the aging of biodegradable polymers and examines recent applications of ML and DL in this field. It also links experimental data with ML/DL models and outlines the key challenges and future perspectives for integrating these approaches into biodegradable polymer research.