TMC-overlaps-and-flipped-vertices

Documentation

MTC Vital Signs leverages TOMTOM and INRIX data to measure Vehicle Hours of Delay for the most congested segments in the 9-county Bay Area region. Our issues faced included many overlapping TMC segments and opposite directional vertices for polylines representing the congested segments.

Define the Problem Statement

Below are two screenshots, one with the line and no offset, and one with the line with the offset applied. They show the line breaking into two directions when the offset is applied. It shows that while the line feature draws as a single line when no offset is applied, the line breaks into multiple segments when the offset is applied, as we were seeing in the interactive map. We think this is attributed to a "To-From"/"From-To" problem. The vertices which make up a polyline draw from start to finish based on the direction of the roadway. ie For a northbound roadway, the first point/vertex in the polyline should be the southernmost, inversely the last vertex should be the northernmost thus the vertices are placed in a sequential order according to the direction of the roadway. This is not the case for this Congested Segments* GEOJSON dataset.

Project Management

• Task on Asana
• Data on Box

Data Sources

MTC Carto tool
Congested Segments* GEOJSON
*Based on previous request "Congested Segments 2017"
Screenshot 1
Screenshot 2

Analysis Parameters

The GEOJSON file was created by a script grabbing the TMC congestion information form both INRIX and TOMTOM data, thus retaining the integrity of that data is important.
There should only be one record per project(/rank of congested segments)
Directionality must be correctly reflected in polyline vertices sequence (because a CSS offset property is used in CARTO for Vital Signs reporting)

Methodology applied to solve problem

Overlapping Congestion Segments

1. Inventory the unique congested segments records
2. Flag the overlapping segments (which look like duplicate records) where multiple TMC segments represent a single congestion segment
3. Merge/Group the like TMC segments so only one record represents one congestion segment
   QC: 153 congestion segments

Flipped Congestion Segments

1. Inventory the segments that reflect incorrect when a CSS offset property is applied via CARTO
2. Isolate the incorrectly reflected segments
3. Edit -> Vertices
4. Reverse Direction
5. Save AND Save Edits once all segments have been flipped

Expected Outcomes

A GEOJSON formatted dataset of only 153 congestion segments (records) with appropriate directional-based vertices.

Results

2016 Time Spent in Highway Congestion