Machine Learning in Industrial Remanufacturing: A Five-Year Review of Applications

Abstract

Remanufacturing is one of the cornerstones in advancing circular economy objectives and sustainable manufacturing operations by extending product lifecycles and minimizing resource use. However, the process is labour-intensive and often economically unviable. This structured literature review explores the recent applications of machine learning in the industrial remanufacturing process, with a focus on inspection and disassembly stages where machine learning has been employed the most in journal articles published between 2020 and 2025. The findings reveal that convolutional neural networks are extensively used for automating visual inspections, improving defect detection precision, and reducing throughput times. An array of reinforcement learning techniques shows great promise in optimizing disassembly planning and control tasks, particularly in human-robot collaborative systems. These technologies collectively enhance productivity, reduce dependence on manual labour, and improve the overall cost-efficiency of the industrial remanufacturing process. While the findings outline considerable promise, the lack of real-industry cases limits practical validation.