Artificial Intelligence-Based Vision Systems for Automated Footwear Disassembly

The increasing consumer preference for shortlifespan products has led to significant environmental challenges in waste management. This process needs to be automated to make the disassembly of footwear easier and thus more widespread; to this scope vision-based systems are paramount. This paper presents the concept of an automated disassembly line for footwear disassembly based on artificial intelligence-based vision systems. The first vision system extracts morphological features, which are useful for identifying and processing the different components of the shoe, while another system extracts the contour of the shoe to be used as a trajectory for the robot in charge of removing the stitching that joins the sole and upper. Experimental results confirm the efficiency of Artificial Intelligence (AI) in extracting morphological features and measuring geometry of the contour crucial for automation.