A close relationship exists between neural and fuzzy systems since they both work with degrees of imprecision in a space that is not defined by sharp deterministic boundaries. Fuzzy neural technologies can be fused into a unified methodology known as fuzzy neural networks. A neural network with the proposed architecture maps a fuzzy input vector to fuzzy output. A learning algorithm is derived from the fuzzy actual output and the fuzzy target output. The fuzzy neural networks have been applied to PCP diagnosis, Concrete mix design and for the design of industrial roofs. Copyright (C) 1996 Elsevier Science Ltd