Machine Learning Fusion Based Technique for Predicting the Concrete Pouring Production Rate Based on Traffic and Supply Chain Parameters

Most construction materials are supplied by out-sourced suppliers and are transferred via road transportation. Concrete is the most used construction material in the world and demand for concrete is ever increasing. In addition, a wide range of crews and machineries are involved in concrete based constructions tasks. As a result, being able to accurately estimate the concrete pouring task will potentially have both cost and time saving effects. Moreover, due to space limitations as well as technical obligations, fresh concrete is mixed at a Ready Mixed Concrete (RMC) depot and then hauled by trucks to constructions sites. Therefore, to predict the concrete pouring duration managers must consider both traffic and supply parameters. In this paper, a data structure is presented to cover these parameters and Machine Learner Fusion-Regression (MLF12R) is used to predict the production rate of concrete pouring tasks. A field database that covers a month of deliveries across a metropolitan area was gathered for evaluating the proposed method. The dataset includes over 2600 deliveries to 507 different locations. Finally, the MLF-R was tested with the proposed dataset and the results compared with ANN-Gaussian, ANN-Sigmoid and Adaboost.R2 (ANN-Gaussian) which are trained with the exact training sets. The results show that MLF-R obtained the least RMSE in comparison with other methods, and also acquired the least standard deviation of RMSE and correlation coefficient with the stability of this approach.

• Supplemental Notes:
  • This paper was sponsored by TRB committee AFH50 Portland Cement Concrete Pavement Construction.
• Corporate Authors:

  Transportation Research Board

  500 Fifth Street, NW
  Washington, DC  United States  20001
• Authors:
  • Maghrebi, Mojtaba
  • Shamsoddini, Ali
  • Waller, S Travis
• Conference:
  • Transportation Research Board 94th Annual Meeting
  • Location: Washington DC, United States
  • Date: 2015-1-11 to 2015-1-15
• Date: 2015

Language

• English

Media Info

• Media Type: Digital/other
• Features: Figures; References; Tables;
• Pagination: 15p
• Monograph Title: TRB 94th Annual Meeting Compendium of Papers

Subject/Index Terms

• TRT Terms: Concrete; Data fusion; Machine learning; Neural networks; Pouring; Productivity; Time duration
• Geographic Terms: Adelaide (Australia)
• Subject Areas: Data and Information Technology; Freight Transportation; I72: Traffic and Transport Planning;

Filing Info

• Accession Number: 01555296
• Record Type: Publication
• Report/Paper Numbers: 15-5928
• Files: TRIS, TRB, ATRI
• Created Date: Feb 26 2015 10:05AM