Fuzzy handling of uncertainty in industrial recycling

During the latest years recycling has become more and more important. When the products have reached their end of life they need to be classified for repair or disassembly. The products classified for repair are then manually repaired. For those products classified for disassembly this can be performed either manually or automatically. In order to perform automatic disassembly we need to make measurements, sometimes complex, on the product. To this end, a complex sensor system with flexible sensor signal fusion is required. The fusion algorithm must be fast, flexible, robust, able to handle all kinds of sensors and achieve a high degree of transparency. It is not easy to fulfil all these requirements, which are central to automatized disassembly. Fuzzy logic, however, holds promise to this end. However, there are problems even with fuzzy logic. The theory concerning fuzzy logic deals only with uncertainty for control actions or for measured values but not for the combination of both. When we measure for automatic disassembly there are many sources of uncertainty and concurrently uncertainty in the control actions. Hence a technique for fuzzy handling of uncertainty for the combination is needed. A solution to this problem is proposed in this paper. The proposed method deals with uncertainty in the measured values as well as in the membership functions.