Date of Graduation


Document Type


Degree Name

Master of Science in Civil Engineering (MSCE)

Degree Level



Civil Engineering


Suman Mitra

Committee Member

Sarah Hernandez

Second Committee Member

Lakeshmi K Sasidharan


Forecasting;Machine Learning;Neural Networks;Regression Models;State Transportation Agency;Time Series


A decline in the number of construction engineers and inspectors available at State Transportation Agencies (STAs) to manage the ever-increasing lane miles has emphasized the importance of workforce planning in this sector. One of the crucial aspects of workforce planning involves forecasting the required workforce for any industry or agency. This thesis developed machine learning models to estimate the person-hour requirements of STAs at the agency and project levels. The Arkansas Department of Transportation (ARDOT) was used as a case study, using its employee data between 2012 and 2021. At the project level, machine learning regressors ranging from linear, tree ensembles, kernel-based, and neural network-based models were developed. At the agency level, a classic time series modeling approach, as well as neural networks-based models, were developed to forecast the monthly person-hour requirements of the agency. Parametric and non-parametric tests were employed in comparing the models across both levels. The results indicated a high performance from the random forest regressor, a tree ensemble with bagging, which recorded an average R-squared value of 0.91. The one-dimensional convolutional neural network model was the most effective model for forecasting the monthly person requirements at the agency level. It recorded an average RMSE of 4,500 person-hours monthly over short-range forecasting and an average of 5,000 person-hours monthly over long-range forecasting. These findings underscore the capability of machine learning models to provide more accurate workforce demand forecasts for STAs and the construction industry. This enhanced accuracy in workforce planning will contribute to improved resource allocation and management.