Progress Monitoring of Construction Projects Using Neural Networks Pattern Recognition

Abstract

The traditional monitoring practice involves collecting actual quantity data, and comparing against a benchmark which represents the relevant planned data. The encountered well-known problem in monitoring is that the quality of the collected data is often subjected to great variation due to the variation in the reporting skills as well as the willingness to record data accurately. A potential technique to circumvent this problem is to conduct the comparison against multiple possible outcomes rather than a single-valued benchmark. The main objective of this research is to utilize the Pattern Recognition (PR) techniques to classify the work planned at specified cut-off dates during the planning stage and use the classification to monitor and evaluate the progress during the construction stage. The PR technique generalizes a virtual benchmark to represent the whole project based on multiple possible outcomes generated at a given cut-off date. The generalization feature offers a potential tool to overcome the problem of variation in the quality of data collected. Patterns are constructed to encode work of the project at different cut-off dates. Finally, the PR concept and technique proved its robustness to monitor and evaluate progress of construction projects based on the technique of Critical Path Method (CPM).