PCI Data Collection

Step by Step instructions on how to generate the PCI data collection base data for the HPMS submittal.

1. Cross-section Data Checks

Since a primary driver to the PCI segmentation is the lane counts, we must perform a rather intensive check on the DDOT cross-section data to ensure minor gaps and slivers do not cause the segmentation data to become unnecessarily fragmented.  Here's the general workflow for the checkCrossSectionData.py script, which drives the inital segmentation process.

2. Refresh the LRS Cross-Section data

This step involves creating the cross-section union data, which is essentially a summary of the roadway cross-section data.  The tool we use for this is Transcend's Segment Analyzer and a template for this process has already been created.

• Go to DDOT's Segment Analyzer (DDOT Internal only)

• Use Template called CrosssectionUnion_mval

• After process is complete, a new table is created in the DDOT LRS database called cross_section_all_mval.  If the table already exists in the database, then it will be overwritten by this process.

3. Process PCI Data

How we chop up the routes.  We do Large and Small slightly differently

Preparatory Steps for All Segmentations

1. Dissolve the Functional Class by type
2. Make copy of Cross Section Union and update through lane counts (LaneSum)

Large Roadway Segmentation

• SegmentationKey = RouteID_Geometry/FeaturePart_SegmentOrder
• Functional Class Types: 
  • Interstates
  • Other Freeway and Expressways
• Incorporates Functional Class and Lane Count for Segmentation
• Does not incorporate the Blocks or Subblocks. 

Basic Steps

1. Filter the Functional Class Dissolve by the above types.
2. Dissolve by Cross-section lane count. 
3. Union #1 and #2 above to create the base segmentation
4. Take the Union in #3 and segment each geometry part by .10mi using breakupSegsByDistance function.
5. Create .10mi segments until done with the FC/LaneCount chunk. 
6. Evaluate 'Change' mileage and process ending event before proceeding to next chunk. (see red info box below)

Small and Local Roadway Segmentation

• SegmentationKey = BlockKey_Geometry/FeaturePart_SegmentOrder
• Functional Class Types: 
  
  • Principal Arterials (Small)
  • Minor Arterials (Small)
  • Collectors (Small)
  • Locals (Local)
• Incorporates RH.Blocks along with Functional Class for Segmentation.

Basic Steps

1. Filter the Functional Class Dissolve by the above types.
2. Iterate through the Functional Class Dissolve and find underlying RH.Blocks using the getBlockIDListfromRouteMs function.
   1. Iterate through each block found and add to Small/Local segmentation.

Cross-Section Overlay

Cross-section data plays heavily in this process.  Basically any width or lane type/directionality information is derived from these data.  Once the FC is segmented for large and small routes (above), the next step is performed.

• Directionality summary (DIRSUM) is updated on the cross_section_all table copy.
• Input segments are rendered into midpoints, which are more clean when it comes to LRS overlay.
• The midpoint PCI segments are overlayed against the cross_section_all table copy.
• Cross-section analysis is performed to provide information on a wide array of items:
  • Number of Through Lanes (inbound/outbound)
  • Number of Parking Lanes (inbound/outbound)
  • Number of Buffered/Barrier features
  • Width (in ft) of Through Lanes
  • Width (in ft) of Parking Lanes
  • Width (in ft) of Buffered/Barriers
  • Bike Lanes (various)
  • General Directionality of the Segment (DIRSUM)
    • OB - Outbound
    • IB - Inbound
    • BD - Both Directions

Complete Attribute Overlay and Update

The above process sort of takes the segments through the ringer and a few key attributes get washed out after the cross-section part.  This last part re-adds the missing data by doing a join to the original table prior to the cross-section.

Additionally, LRS information is overlayed and also added after that:

• Ownership
• HPMSSection
• NHS

2020 Prep Meeting/Discussion with ARA