Multimaterial injection molding (MMIM) enables the integration of distinct polymer properties within a single component, but its tightly coupled thermal–rheological dynamics and interfacial phenomena complicate process control and quality assurance. Concurrently, advances in artificial intelligence (AI) and machine learning (ML) are reshaping manufacturing through predictive modeling, adaptive optimization and data-driven control. This review critically synthesises AI/ML methods and adjacent optimisation and control frameworks relevant to MMIM, distinguishing learning-based predictive models from conventional statistical, surrogate-based, and simulation-assisted approaches. Within this framework, methods such as artificial neural networks, radial basis function models, genetic algorithms, hybrid intelligent optimisation, and sensor-integrated control strategies are mapped against key manufacturing challenges including weld lines, sink marks, short shots, warpage, burn marks, and flow marks. Evidence from 2010–2025 shows that AI can improve dimensional stability, defect prediction and parameter selection, reduce experimental trial-and-error, and support intelligent monitoring. The most promising direction is the integration of AI with in-mold sensing, simulation and closed-loop control for real-time compensation. However, broader adoption requires high-quality datasets, standardized benchmarks, robust cross-material generalization and interpretable models validated in factory conditions. We outline a research agenda to bridge laboratory demonstrations and scalable industrial implementation, positioning AI as an enabler of more robust, adaptive and sustainable MMIM.
Santos et al. (Wed,) studied this question.