ITS in WZ: ARAMPS and Other Initiatives!

SLIDE 1

ITS in WZ
ARAMPS and Other Initiatives!

Jawad Paracha PE, PTOE
MD State Highway Administration

SLIDE 2

Outline

  • Research
  • ARAMPS Demo
    (An Automated Real-Time Travel Time Prediction System)
  • Where are we heading!
SLIDE 3

Ongoing Research

  • ARAMPS
  • Dynamic late merge system (Phase-II)
  • License Plate Recognition System (Phase-II)
  • Variable Speed Limit System (Phase-II)
SLIDE 4

ARAMPS Demonstration Site

Demo period: 6/4/2006 to 8/4/2006

A road map showing highway 70 and the crossing of Maryland 27, Maryland 94, Maryland 97, Maryland 32, and highways 40, 29, and 695 from left to right.

  10 detectors       5 signs         25 miles

SLIDE 5

Introduction of the ARAMPS System

An image shows how the ARAMPS system works.

SLIDE 6

System Flowchart

A system flowchart that provides the process that occurs when predicting travel time. It starts with Detector Location Identification Module, goes to Real-time detector data at time t, Travel Time Estimation Module, Database of Historical Travel Times, and Travel Time Prediction Module. It goes back to Real-time detector data at time t with the formula of t=t+1. From there, it goes to Database of Traffic Data, to Incident Detection Module, to Missing Data Estimation Module, and to Travel Time Prediction Module. The cycle continues.

SLIDE 7

ARAMPS Website (Current Prediction)

A screen capture of the Current Prediction section of the ARAMPS Website

SLIDE 8

ARAMPS Website (Historic Data)

A screen capture of the Historic Data section of the ARAMPS Website

SLIDE 9

Evaluation Summary

Traffic Condition Average Absolute Error Actual Travel Time
Free Flow Less than
1 minute		 18 – 22 minutes
Moderate Congestion Less than
3 minutes		 22 – 30 minutes
Heavy Congestion Less than
3 minutes		 30 – 45 minutes
SLIDE 10

Potential Issue: Missing Data

A graphic explains the potential Short-term (communication) and Long Term (Device Failure) issues

SLIDE 11

Potential Issue: Missing Data

A graphic illustrates the missing data from several spots on the highway 70 due to the reliability issue of the wireless network

Missing data and/or delayed data occur at least one time per hour due to the reliability issue of the wireless network

SLIDE 12

Missing Data Estimation

  • Scenario 1: Only 1 detector has data missing
    Importance factor of each detector can be determined from the following table
     Det ID 1 2 3 4 5 6 7 8 9 10
     MP L L M H M M H M M H
     EP L L L L M M M M M H
    OP L L L L L L L L L L

    MP = Morning Peak Hours             L = Low Importance
    EP = Evening Peak Hours             M = Moderate Importance
    OP = Off-peak Hours                     H = High Importance
  • Scenario 2: More than 1 detector has data missing
    Additional to above rules, the importance factor of one detector will be changed to high if its neighboring high-importance detector is experiencing missing data
SLIDE 13

Missing Data Estimation (cont’d)

  • Determine if missing data can be estimated or the system has to suspend involved signs
  • Importance factor vs. current traffic conditions 

      Low Importance Moderate Importance High Importance
    Congestion-free  Estimate Estimate Estimate
    Moderate Congested Estimate Estimate Suspend
    Heavy Congestion Estimate Suspend Suspend
SLIDE 14

Potential Issue -Incident Scenarios

  • An incident detection model is running in real time (for detecting major accidents only) 

    Major incidents were successfully identified
    Near 695, 9:18PM on 6/4/06
    Near 695, 8:26AM on 6/7/2006
    Near 695, 8:28AM on 6/20/2006
    Near 695, 8:30AM on 6/30/2006
SLIDE 15

Potential Issue -Incident Scenarios

A graphic showing the traffic situation on highway 70 heading towards highay 596 with all electronic signs displaying EXPECT CONGESTN NEAR 695 or EXPECT CONGESTN AHEAD

SLIDE 16

Potential Issue -Incident Scenarios

A graphic showing the traffic situation on highway 70 heading towards highway 695 with all electronic signs displaying STAY ALERT

SLIDE 17

System Monitors

  • Monitoring the system component failures: 
    • Detectors
    • Web site status

       A screen capture of the web site where the live traffic situation on highway 70 heading towards highway 695 is displayed 
    • FTP status
    • Database status
    • VMS sign status 
  • System administrator notified
SLIDE 18

Conclusions

  • 10 detectors for a 25-mile freeway segment
  • Reliable estimation/prediction of TT under different traffic conditions
  • Improved system reliability by integrating the missing data estimation and incident detection modules
SLIDE 19

Where are we heading!

  • Statewide ITS in WZ contract
  • Performance based specs
  • Performance Mgmt (Final Rule)
  • Automated Enforcement
SLIDE 20

End of Presentation

Thank you!

Categories: |