A neural network application to subcontractor rating in construction firms

In construction companies, subcontractor rating, due to its intrinsic ambiguity and difficult formalization, is a particularly complex task, usually accomplished by management experience and intuition. In this paper, a neural network is proposed to support management in subcontractor rating. Neural networks, being able to learn directly by examples the managers' logic, are suitable to support the solution of this type of problem. By an application case related to an assembly operation in a construction site, the neural network implementation and the related managerial and technical innovations are investigated.