Geographic Information System-Based Framework for Estimating and Visualizing Unit Prices of Highway Work Items

Abstract

Historical bid items from previous projects are an important source of data for project cost estimation. State highway agencies have widely used this data source for budget planning of highway projects. Current practices on using the data are, however, not efficient and fail to provide a reliable estimate. The estimate is simply based on the mean value, which is then manually adjusted by professionals using their knowledge and experience. A highway project is unique, and its price is highly dependent on various factors including location and the type of construction activities. A geospatial analytic method for project cost estimates would be highly useful for automated assessment of these influencing factors. The state of the art has applied interpolation methods to location cost-adjustment factors to adjust the total costs of two similar projects in two different cities. However, existing methods are mostly beneficial to conceptual cost estimation without considering variances between two projects in the same city and various effects of location on different work activities. This study contributes to the body of knowledge by proposing a geographic information system-based framework that leverages historical bid data for unit-price estimation and visualization with consideration of the effects of project-specific location on different bid items. Apart from applying established spatial interpolation methods to unit-price estimation, various strategies such as the use of quantity in interpolation models are proposed to improve the accuracy of the estimates. Temporal changes in unit prices and relationships between quantities and unit prices are also explored.