Details:
-
Creators:
-
Corporate Creators:
-
Corporate Contributors:
-
Subject/TRT Terms:
-
Publication/ Report Number:
-
Resource Type:
-
Geographical Coverage:
-
Edition:Final Report
-
Corporate Publisher:
-
Abstract:This research project focuses on deteriorating roadway asset conditions, emphasizing the challenges encountered by the roadway maintenance division of the Washington State Department of Transportation (WSDOT). The project goal is to develop algorithms for prediction models that will forecast the levels of service (LOS) performance conditions of six important highway assets: culvert maintenance, barrier maintenance, traffic signal systems, ditches, slope repairs, and shoulder maintenance. These algorithms are based on a data-driven approach. The algorithms provide a step-by-step process to develop prediction models. The models can be used to forecast LOS performance conditions and trends under various funding levels, allowing them to set performance targets that align with available funds and asset maintenance priorities, potentially preventing expensive reactive maintenance. Data collection included direct collection from WSDOT and two-phase questionnaire surveys to document factors impacting LOS performance conditions. Statistical analyses such as the Relative Importance Index (RII), Kolmogorov-Smirnov and Shapiro-Wilk normality tests, and Mann-Whitney U tests were employed to determine critical factors for each of the six assets. The project identifies the top five highly ranked factors for each asset, which are utilized during model development. Based on the dataset collected, a future study employing Machine Learning approach is recommended to develop prediction models for the assets. Prediction models serve as a tool for forecasting asset conditions, calculating base funds required for each asset, and optimizing resource allocation. Through the project outcomes, states will be able to improve asset management decision-making, resulting in safer and more environmentally friendly roads.
-
Format:
-
Funding:
-
Collection(s):
-
Main Document Checksum:
-
Download URL:
-
File Type: