摘要: |
Following recent state and federal legislation related to the use of performance and risk-based asset management strategies to inform transportation investments, there is an increased need to update methods and models utilized within existing North Carolina Department of Transportation (NCDOT) asset management practices to ensure reliable and optimal use of these tools. Specifically, the National Bridge and Tunnel Inventory and Inspection Standards (U.S.C. Section 144) were revised by the Moving Ahead for Progress in the 21st Century (MAP-21) legislation to mandate the use of "a data-driven, risk-based approach and cost-effective strategy for systematic preventative maintenance, replacement, and rehabilitation of highway bridges and tunnels to ensure safety and extended service life." A vital component of such a data-driven, risk-based approach to asset management to ensure that cost-effective strategies are identified are the prediction models for accurately anticipating the costs of bridge replacement, rehabilitation, and preservation actions. The research needs statement recognizes that the current techniques used for bridge replacement cost estimation, while often sufficiently reliable when applied to typical bridge replacement projects, are often significantly inaccurate and unreliable when applied to bridges at the high- and low-ends of the cost scales. This evidences a serious and potentially costly shortcoming in the current approach used for financial forecasting and project prioritization and selection. The inaccuracies are most likely the result of inadequately accounting for unique local project factors in the historical data analysis or a failure to adequately forecast construction cost trends that may significantly influence the project costs. The proposed scope of work for this research effort will provide research on best practices for cost-based estimation, strategies to incorporate the latest construction cost indexes into the analysis of historical data, and methods to forecast the impact of cost trends within estimates. Historical cost data sourced from the NCDOT HiCAMS system will provide the granularity necessary to enable the research team to examine the potential factors most significantly contributing to inaccuracies in bridge replacement cost estimates relative to actual costs. Strategies to better incorporate or condition these variables that drive the inaccuracies in the current estimates will be developed to deliver improved cost estimating algorithms. Furthermore, incorporation of construction cost forecasting models will be investigated. To characterize the performance and research impact delivered by these revised models, improvements in cost estimation accuracy will be assessed using both predictions on historical data and documented new bridge replacement costs that are reported over the course of the research effort. This research directly supports data-driven and performance-based asset management initiatives and complements recent and concurrent research providing updates and improvements to the NCDOT Bridge Management System (BMS). |