Author(s)

P. María-Sánchez

Abstract

In Mexico there is an urgent necessity to develop new ways for evaluating environmental risks, mainly based on new modelling techniques. This necessity is a critical problem for the Mexican government, due to the high development of the urban areas in Puebla. In addition, there is an absence of knowledge on this topic within many government agencies. The aim of this paper is to develop a neural network approach to assess the impact of environmental risks in construction projects in Puebla (Mexico). The objectives are: to create a humanintuition approach to advise government agencies towards the impacts of environmental risks, to store knowledge about risks in a single tool, to forecast the possible values of risks for developing appropriate contingence measures, to develop a flexible tool for use as an expert’s opinion for similar future projects and to examine the feasibility of neural networks for pattern recognition. The data used for creating, training and testing the neural network was obtained from private contractors who are constantly involved with environmental risks. It was possible to demonstrate with the results, the pattern recognition of this neural approach, whilst testing the network with unknown data. In conclusion, this new approach for evaluating environmental risks in construction projects is an alternative tool that can be simulated for obtaining the probable risks impacts for a specific project. The methodology has a potential in modelling environmental risks, providing valuable outcomes for project managers working in government agencies. Keywords: environmental risks, neural networks, project managers, construction management.

Keywords

environmental risks, neural networks, project managers, construction management.