Engineering performance improvement using a hybrid artificial intelligence system

Publisher Summary Industrial construction projects have been experiencing serious problems such as cost overrun and schedule delay. While seeking for the causes of the problems, people have come to realize the impacts of engineering activities on a successful project implementation. Improving engineering performance can lead to better project outcomes. This chapter proposes a generic model, which integrates genetic algorithms with artificial neural networks, for modeling engineering performance measurement, and improvement on industrial construction projects. Due to its robust and efficient search ability in complex situations, genetic algorithms are employed to search for solutions improving engineering performance with the searching criteria, and fitness function, being the neural networks that have been trained to predict performance for given engineering inputs. The chapter finally presents the research introducing the idea of self-comparison evaluation, which compares a project's actual engineering performance to its possible better performance searched by genetic algorithms in order to make suggestions for possible engineering performance improvement.