The Use of Neural Network Classifiers in Solid Waste Recycling

The massive adoption, in these last years, of specialized equipment or complex processing architectures specifically developed to separate different solid waste materials, resulting from the selective collection of solid urban waste, equipment or manufactured goods dismantling at the end of their life cycle, more an more requiring control systems able to “qualify” the products during the processing. Such a goal, when implemented “on-line”, is usually realized in two steps. The attributes (physical, chemical, morphological, morphometrical, textural, etc.) of the materials resulting from processing are detected and numerically modeled. The resulting feature vector is then “handled” by a software architecture performing the required recognition/classification procedure and defining the quality of the investigated products. From the results further “feed-back” or “feed-forward” control strategies can be applied in order to improve equipment or processing architectures performances. In this paper are analyzed and described the advantages and the problems encountered by the authors when neural network (NN) based architectures have been adopted to define “artificial intelligence software unit” able to perform the recognition, at industrial recycling processing plant level, of several solid waste materials starting from their preliminary optical recognition