- Long-Term Bridge Performance (LTBP) Portal (Data Extraction)
- LTBP Deterioration Model Algorithm
- Cost Analysis Model
The Bridge Portal is more than just a database. It not only contains bridge performance-related data mined from existing sources (National Bridge Inventory, State Highway Agency bridge element-level data, national weather data, traffic data, weigh-in-motion data, bridge maintenance data, and other data sources), but it also serves as a central repository for all field data collected through the LTBP Program. Additionally, the Bridge Portal also functions as a research and decisionmaking tool by implementing bridge life cycle and deterioration modeling using both mined data sources as well as LTBP-acquired field data to allow users to investigate bridge performance on many different levels.
Figure 1. Screen capture. LTBP Bridge Portal home page.
The LTBP Bridge Portal is the new bridge analysis site with performance and experience unmatched by any other national bridge inventory search and analysis website. The site contains the complete National Bridge Inventory (NBI) database; national bridge element data is being added as it becomes available. Additionally, bridge owners are contributing bridge condition data to the site, including element-level data that was collected prior to the American Association of State Highway and Transportation Officials (AASHTO) establishing the bridge element inspection and reporting requirement.
As the LTBP Program collects data, it will be added to the database for analysis. Data from the successive field data collection efforts for the Mid-Atlantic bridge clusters will be the first LTBP data sets submitted to the bridge portal database. Data include visual inspection, nondestructive evaluation (NDE) results, and results of material testing as they become available.
The LTBP data collection protocols ensure continuity of data types and collection methods across different bridges or on the same bridge in different years. The quality control requirements applied to the portal data result in research-quality data that is collected on a uniform and repeatable basis. From the Bridge Portal Homepage, you can access the Simple Search, Advanced Search, Global Filters, and Settings. The setting tab (figure 2) allows the user to define the search attributes and define the results to be selected. The simple search allows you to use a set of filters for quick searches. The advanced search allows you to choose from a list of filters, providing greater control over the search.
Figure 2. Screen capture. LTBP Bridge Portal Settings tab.
Users can view a summary of the search results (figure 3), with information and statistics in a table or in a dynamic spreadsheet with sortable content and individual bridge information below. The search results can be exported to formats such as Excel, PDF, and KML. Searches can be executed on a population of bridges or users can access in depth data on a single bridge.
Figure 3. Screen capture. Sample summary of search results in LTBP Bridge Portal.
On the results page (figure 4), users can see the information grid and sort, reorder, add columns, or filter to see specific information. Users can save and share searches with the community, or access globally shared searches.
Figure 4. Screen capture. Sample LTBP Bridge Portal results page.
Users can view the bridges from their search on a Google Maps interface with region selection tools and customizable marker sizing and coloring, then export as an image or PDF.
Bridge deterioration models are an important component in decisionmaking with respect to allocation of current and future funding for maintenance and repair (M&R) of bridges. A primary drawback to the current in-service deterioration modeling approach is that it is not data-driven and can result in inaccurate deterioration forecasts. The LTBP Program endeavors to improve deterioration forecasting by developing a data-driven approach to deterioration modeling that uses both legacy data and field data collected through LTBP Program efforts. The LTBP deterioration-learning algorithm is both data-driven and extremely flexible.
The LTBP deterioration-learning algorithm improves on the traditional approach first by expanding the choices of model functional form (MFF). Instead of a single MFF, the LTBP learning algorithm allows multiple MFFs thereby allowing the learning process to "choose" the MFF or combination of MFFs that actually agree with the data. The ability to propose multiple MFFs allows the flexibly to incorporate different opinions. When multiple MFFs are used, each MFF is assigned an initial weight (the sum of all weights is 100 percent). The individual weights assigned to a particular MFF represent the likelihood of a specific MFF agreeing with the input data. Through the learning process, the individual weights are updated with each new round of data points.
This deterioration modeling approach proposed by the LTBP Program not only provides the user with tremendous flexibility but also will improve forecasting accuracy through a truly data-driven approach. It will be deployed within the next version of LTBP Bridge Portal, expected in June 2017.
The LTBP Program initiated the development of the theoretical cost analysis model (CAM) that draws from the comprehensive data collection effort and the development of advanced deterioration model(s). As with the advanced deterioration model, CAM will be probabilistic and supported by the LTBP data collection program.
At this stage of model development, the program is working with States to obtain maintenance action and cost data to further develop and validate the LTBP CAM. Based on the data, inclusive of visual inspection, NDE, material testing, maintenance action, and cost data, the CAM will be a learning model like the deterioration model.
The objective of the model is to support cost-analysis or data-driven repair planning and budgeting.