ASSESSMENT OF CONCRETE BATCH PLANT PRODUCTIVITY ANDDELAYS USING ARTIFICIAL NEURAL NETWORK METHODOLOGY

摘要

Assessing productivity, cost, and delays is essential to manage any constructionoperation. This paper focuses on assessing the above-mentioned items for concrete batch plant (CBP)operations using Artificial Neural Network (ANN) methodology. Data were collected to assess cycle time,delays, cost of delays, cost of delivery, productivity, and price/cy for the CBP. Two ANN models weredesignated to represent the CBP process considering many CBP variables. Three-layer, feed forward, andfully connected ANNs were trained with an architecture of twelve input neurons, four output neurons, anddifferent hidden layer neurons. Input variables include distance, concrete type, and truckload. Outputvariables include cycle time assessment, cost of delays, delivery cost, productivity, and price per concreteunit. The ANN outputs have been validated to show the ANN's robustness in assessing the CBP outputvariables. The average validity percent for the ANN outputs is 95.84%.