Abstract
A large number of problems faces the installation of pile (drilled shaft) foundations: unseen subsurface obstacles, lack of contractor experience, site planning, etc. These problems make it difficult for the estimator to assess the pile construction productivity and cost. Several techniques might be good candidates for this assessment problem. A fundamental question arises: which technique is the most appropriate to solve this assessment problem? This study focuses on answering this fundamental research question. Data were collected through designed questionnaires, site interviews, and telephone calls to experts in different construction companies. Four different techniques were listed as candidates to solve this problem: deterministic, simulation, multiple regression, and artificial neural network (ANN). They were categorized into two groups: process oriented techniques, deterministic and simulation; and data oriented techniques (DOT), regression and ANN. All techniques were used to assess productivity and cost of pile construction. Their results were compared to determine the closest assessment to real world practice. Research results show that the DOT techniques were the most appropriate whereas they had the lowest deviation from real world practice.
Original language | English |
---|---|
Pages (from-to) | 490-499 |
Number of pages | 10 |
Journal | Journal of Construction Engineering and Management |
Volume | 130 |
Issue number | 4 |
DOIs | |
Publication status | Published - 1 Jul 2004 |
Externally published | Yes |
Keywords
- Construction industry
- Data analysis
- Neural networks
- Piles
- Predictions
- Productivity
- Simulation
- Statistical analysis
ASJC Scopus subject areas
- Civil and Structural Engineering
- Building and Construction
- Industrial relations
- Strategy and Management