Dipartimento di Ingegneria informatica, automatica e gestionale

Waste management is one of the key areas where circular models should be promoted, as it plays a crucial role in minimizing environmental impact and conserving resources. Effective material identification and classification are essential for optimizing recycling processes and selecting the appropriate production equipment. Proper sorting of materials enhances both the efficiency and sustainability of recycling systems. The proposed study explores the potential of using a cost-effective strategy based on hyperspectral imaging (HSI) to classify space waste products, an emerging challenge in waste management. Specifically, it investigates the use of HSI sensors operating in the near-infrared range to detect and identify materials for sorting and classification. Analyses are focused on textile and plastic materials. The results show promising potential for further research, suggesting that the HSI approach is capable of effectively identifying and classifying various categories of materials. The predicted images achieve exceptional sensitivity and specificity, ranging from 0.989 to 1.000 and 0.995 to 1.000, respectively. Using cost-effective, non-invasive HSI technology could offer a significant improvement over traditional methods of waste classification, particularly in the challenging context of space operations. The implications of this work identify how technology enables the development of circular models geared toward sustainable development hence proper classification and distinction of materials as they allow for better material recovery and end-of-life management, ultimately contributing to more efficient recycling, waste valorization, and sustainable development practices.