Accounting for the impact of learning and scheduled overtime: Developing dynamic models to estimate and control labor time and labor cost.

摘要

Labor cost is a major part of construction cost, and it is important for contractors to accurately estimate and control labor cost on their construction projects. Current methods of estimating labor productivity generally do not consider the quantity of work and resulting number of repetitions of a task nor the impact of overtime, when it occurs. Instead, current labor estimating and control practice assumes a linear relationship between quantity of work and time and cost of labor. But this is not true in practice. As the workers continue on an assigned task, they develop site-specific familiarity that allows them to perform that task more efficiently, and their productivity therefore increases. This is called the learning effect.; Some projects require workers to work scheduled overtime, which increases worker fatigue under which work is less productive. This is called the overtime effect. Learning and overtime interact resulting in complex changes in rates of productivity. Due to these variations of the productivity rate in a construction task, non-linear dynamic time estimating and control models are more appropriate than linear static time models. The objective of this dissertation is to develop models that can include non-linear dynamic effects, specifically labor time and cost estimating and control models for use with single tasks and complex tasks.; The body of this dissertation is separated into three parts: learning, overtime, and labor time and cost estimating and control. Chapter 2 discusses the unit learning curve and the cumulative learning curve models with attention to four areas: fitting and predicting accuracy, time estimating models with processes of obtaining learning curve factors from construction sites, unit sizes, and combination of subtasks. Chapter 3 presents improved Business Roundtable (BRT) information, new overtime functions, and estimating models based on an overtime Table that is developed from the improved BRT information. Chapters 4, 5, and 6 use the time estimating models from Chapters 2 and 3 to include the impacts of learning and/or overtime on labor time and cost estimating and control. In total, this dissertation develops and demonstrates methods by which managers can better estimate and control projects by including the impacts of learning and overtime.