TravelTime Based Performance Monitoring

1
Travel-Time Based Performance Monitoring

• L.A. Griffin
• Orlando-Orange County Expressway Authority
• September 23, 2008

2
Agenda

• Objectives
• Data Sources
• How We Did It
• Report Samples

3
Objectives

• Leverage archived travel time data to monitor
  expressway operations
• Focus on outcome measures
• Monitor entire OOCEA system
• Monthly reporting
• Create tools to generate reports
• Meaningful to operations/engineering staff while
  also useful for Executive management

4
Source Data Travel Time System

• 111 Data Collection Sensors
• Read transponders to determine travel times
• Automated Vehicle Identification (AVI) Technology
• At interchanges, system boundaries, DMS locations

5
Data Server Travel Time Software Feeds OOCEA
DMS, FDOT 511
6
Data Server Travel Time Software Query Archived
Data
7
How Much Data Do We Deal With?

• 143 unique links
• Link ID, Travel Time, Link Speed, Time Slice
• 1440 timeslices / day
• 1 Month Travel Time Data
• 6,383,520 records in a 31 day month
• 750 MB text file (.csv)

8
Define Reporting Segments
9
Define Reporting Segments
10
Tools

• Data Consolidator (Custom)
• Graphical Reports (Excel Based)

11
Report Examples
12
Report Examples
13
6/3/2008
14
(No Transcript)
15
(No Transcript)
16
So What Happened?
17
Monthly Reports

• 18 pages total
• (4) Reference 2 Link Maps/2 Key Sheets
• (1) Monthly Summary Table
• (1) Highest Congestion Table
• (12) Travel Time Analysis for each Reporting Link

18
Our Experience

• Plan on an iterative process to select the
  measures and the report format
• Use mock ups to prototype and refine the reports
• Need efficient tools to translate large data sets
  into meaningful reports
• Play to strengths of your tools
• Data manipulation more suited to custom apps.
• Excel good for visuals.do not need to reinvent

19
Our Experience

• Where possible, use visuals to communicate the
  meaning of your measures
• Define thresholds based on your operational
  objectives
• Travel Time Index allows you to compare segments
  of different lengths
• Planning Time more easily understood than 95th
  percentile reliable travel time
• Goal is to turn data into information

20
Thank You!