Estimating Labor Production Rates for Industrial Construction Activities

Abstract

This paper discusses an approach based on artificial neural networks that enables an estimator to produce accurate labor production rates (labor/unit) for industrial construction tasks such as welding and pipe installation. The paper first reviews factors that were found to affect labor production rates on industrial construction tasks, current estimating practices and their limitations, and the process followed in collecting historical production rates. An artificial neural network model is then described. The model is composed of a two-stage artificial neural network, which is used to predict an efficiency multiplier (an index) based on input factors identified by the user. The multiplier is then used to adjust an average production rate given in man-hours/unit for use on a specific project. Estimates of production rates from the new approach are compared to the existing estimating practices and conclusions are presented.