Perspective of self-learning robotics for disassembly automation

Abstract

Increasing attention has been paid to remanufacturing which plays an important role in environmental protection and circular economy. Disassembly is a key operation in remanufacturing, repair, and recycling. Several robotic disassembly developments have shown that the use of robots in disassembly is feasible; however, the programming of robots is usually complex, schedule-based, and time-consuming. Recent research about self-learning robotics and human-robot collaboration have created an opportunity for schedule-free robotics, in which various machine learning and deep learning techniques have been developed. This paper attempts to review the development of self-learning robots with applications in robotic disassembly and remanufacturing. Key algorithms, designs, control methods, and future research directions have been highlighted and analysed. This review paper serves as a useful resource for researchers in the areas of robotics, smart remanufacturing, and disassembly automation.