|<< Previous||Contents||Next >>|
Evaluation of Highway Performance Measures for a Multi-Study Corridor - A Pilot Study
2.0 Data Aquisition and Analysis Plan
This section presents the plan for acquiring pavement and bridge data from the states in the I-95 Corridor Coalition.
2.1 Scope of Data Collection Effort
The Federal Highway Administration (FHWA) and the project team identified the following three states along the I-95 Corridor for the analysis of highway performance measures: Delaware, Maryland, and Virginia. Cambridge Systematics, Inc. (CS) contacted the states and all agreed to provide their pavement and bridge data. We have completed the process of gathering data from the states.
Table 2.1 displays the available sources of pavement and bridge data from the three states that CS has collected.
|State||HPMS||PMS Data||NBI||Pontis BMS Data||Other BMS Data|
The following paragraphs describe each data source. At the national level, the pavement data are available through the Highway Performance Monitoring System (HPMS) and the bridge data through the National Bridge Inventory (NBI). Most agencies collect additional pavement and bridge data that go beyond Federal requirements in database-driven systems such as a pavement management system (PMS) and/or a bridge management system (BMS).
- HPMS. HPMS includes inventory, performance, and condition data for the nation's highways. States are required to collect and submit these data to the FHWA annually. Present Serviceability Rating (PSR) and International Roughness Index (IRI) are the only pavement condition measures reported in HPMS. HPMS does not include pavement distress data such as rutting, cracking, or faulting. However, many states do collect these additional data and employ an overall pavement condition index that is agency-specific. The HPMS Reassessment 2010+ effort will add a number of pavement distress measures, and standardize data collection and reporting requirements.
- PMS. Most states have developed custom pavement management systems that they use to store inventory, condition, traffic, work history, and cost data. Since there are no standards on how these data are stored, the PMS specification varies widely from state to state.
- NBI. NBI includes the inventory, condition, and functional data for the nation's highway bridges. States are required to inspect bridges biannually and submit these data to the FHWA every year. More frequent inspections are performed if warranted based on condition. The NBI dataset contains condition ratings for deck, superstructure, substructure, channel/channel protection, and culvert.
- BMS. Pontis includes element-level inspection data as well as NBI data. Element-level data provide more detailed bridge condition information. Many states use AASHTO's Guide for Commonly Recognized (CoRe) Structural Elements for element definitions and condition state language, but states may create their own custom elements and/or modify the CoRe element definitions. Virginia and Delaware collect element-level bridge data using Pontis. They also have added agency-specific data items to the Pontis database. Maryland has its own BMS that stores element-level data.
In order to obtain a comprehensive view of the data collected by states, CS held additional discussions with pavement and bridge personnel in each state. The goal of these discussions was to more fully understand the process that each state uses to gather the data (e.g., what technologies are used, under what circumstances are inspections performed, how are data validated) and what constraints govern the use of these data.
For example, the completeness and accuracy of pavement data gathered as part of a visual inspection will differ from data gathered using automated electronic recorders. Users of the data need to understand these limitations in order to properly interpret the data that is available. This information is included in the Data Collection Report found Section 4.2 of this document.
2.2 Data Collection Methodology
In order to collect the pavement and bridge data necessary for this study, CS contacted key personnel in the candidate states and requested that they provide information from their pavement and bridge management systems, respectively. Specifically, CS requested a complete set of condition data, including element-level information for bridges and all measured or computed values for pavement, for all bridges and roads covered by these systems. If states felt that the effort to provide these data would be burdensome, CS requested, at a minimum, pavement and bridge data for the I-95 corridor.
Table 2.2 summarizes the information received and the persons that provided this information. These persons also served as the main points of contact to clarify data already received and to obtain additional information regarding the inspection and quality assurance processes used by each state agency.
|State||Data Received||Contact Person|
|Delaware||Bridge Data - a Pontis database containing NBI and element-level condition information for all bridges in the state.||Douglas E. Finney, P.E.
Bridge Management Engineer
|Pavement Data - a Microsoft Excel spreadsheet containing IRI and overall pavement condition (OPC) values for the Delaware road network.||Kim Johnson
|Maryland||Bridge Data - a Microsoft Excel spreadsheet containing element condition data for the state-owned bridges on I-95 only. State-owned bridges make up approximately one-half of the I-95 bridges. The other bridges are owned by the Maryland Turnpike Authority.||Robert J. Healy
Deputy Director, Office of Structures
Maryland State Highway Administration
|Pavement Data - a Microsoft Excel spreadsheet containing IRI, rutting, cracking and friction values and indexes for I-95 only.||Mark F. Wolcott, P.E.
Deputy Director for Material Engineering
Maryland State Highway Administration
|Virginia||Bridge Data - a Pontis database containing NBI and element-level condition information for all bridges in the state.||Anwar S. Ahmad, P.E.
Assistant Division Administrator
|Pavement Data - a Microsoft Access database containing data on transverse and longitudinal cracking, alligator cracking, clustered cracking, patching, potholes, delaminations, bleeding, rut depth, load distress ratings, shoulder conditions, and other values for the Virginia interstates.||Tanveer Chowdhury, P.E.
Assistant Division Administrator
Maintenance Division, VDOT
2.3 Potential Issues with Data
Even though HPMS and NBI reporting requirements provide a consistent stan-dard for collecting pavement and bridge data, in reality the level of detail for these data varies significantly from state to state. Furthermore, it is much harder to compare state-specific data items that are outside of HPMS and NBI due to a lack of national standard and guidelines. States may have distinct agency policies, such as different thresholds for identifying deficiencies.
- Many states use an overall pavement condition index to represent average pavement condition, but there is no specific standard for calculating this measure.
- States often use automated road profiling instruments that measure road characteristics and calculate surface smoothness. Due to different equipment used by the states, the pavement condition data being collected are not based on the same specification.
- HPMS sample data are not collected for all interstate highway sections.
- The number of pavement values collected varies widely by state, which limited the number of state-to-state comparisons that were made.
- Visual inspection of pavement and bridges can be subjective, and the inspector training is different from state to state.
- Comparing health index across states can be problematic due to a lack of consistency in element definitions. States may modify the CoRe element definitions and add their own custom elements.
Addressing issues such as these was a main thrust of this project.
2.4 ICAT Integration
The goal of Task 2 was to integrate data from states into ICAT. This process, discussed in Section 3, included a more complete analysis of the data available from each state. Part of the integration with ICAT involved resolving discrepancies between the ICAT network and the routes and other location information included in the bridge and pavement data collected from participating states. At a minimum, this required that CS ensure each route in ICAT had a unique identifier across all states and that there was sufficient information in the pavement and bridge data to reproduce this identifier. This allowed pavement and bridge data to be linearly referenced against the ICAT road network.
Once a route system was built from the ICAT road network, dynamic segmentation techniques were used to categorize and overlay the pavement and bridge data. These overlays were published as standalone files that were incorporated into the web-based ICAT (WebCAT) interface as part of Task 4, which is discussed in Section 5. Users can select from and display categorized performance measures as well as displaying raw data for any bridge or pavement section.
For Task 3, CS established candidate performance measures that summarize the detailed pavement and bridge data. CS also performed any necessary data calculations and transformations to compute these measures using the available data. The following is a list of statistical functions that were used to evaluate the quality, consistency, and completeness of the pavement and bridge data.
For individual measures, CS:
- Calculated the minimum, maximum, mean, median, variance and number of data points;
- Plotted the probability and cumulative distribution functions;
- Performed calculations separately with each pavement section or bridge having equal weight and weighted using deck area for bridges and section length for pavements; and
- Performed calculations separately for each state and combined across states where the same measure is collected in multiple states.
For reasonable combinations of measures, CS:
- Calculated the correlation coefficient for scalar measures that appear to be correlated;
- Calculated the correlation coefficient of the rank for categorical measures that appear to be correlated; and
- Performed these calculations separately for each state as well as combined where the same measure is collected in multiple states.
For bridges, the following combinations were considered:
- Health Index versus Sufficiency Rating; and
- Deck/superstructure/substructure condition index versus condition ratings (converted into ranks).
For pavement, the only measure common across all three states is IRI. It was not possible to compare rutting and cracking between Maryland and Virginia because these states record these values using different measures (e.g., index values versus linear feet of cracking).
CS also considered other possible indicators that could be created if additional information were available. For example, values such as load capacity (per unit surface area) and remaining service life could be constructed from information on pavement composition and/or maintenance history. In Section 4.4, CS suggested future areas of research for consideration by FHWA. In Task 4, CS developed the means to display bridge and pavement data and performance measures within the WebCAT web application. This application uses Geographic Information System (GIS) processes to display the network of roads in I-95 member states. WebCAT also is capable of displaying information associated with these roads.
CS reviewed information collected previously by the I-95 Corridor Coalition for incorporation into ICAT. The existing database includes geospatial data, HPMS data, and NBI data. We leveraged this previous work when integrating the data collected for this project with ICAT. As part of incorporating the new pavement and bridge data received from Virginia, Maryland, and Delaware, CS extended the ICAT tables to include not only the superset of all the data elements received from these states, but additional performance measures that were calculated from these elements.
|<< Previous||Contents||Next >>|