Beware of accuracy and privacy issues in using truck transponder data for developing real-time traveler information applications.

An Oregon Department of Transportation Experience

August 2009
Oregon; United States

Background (Show)

Lesson Learned

The truck transponder data, archived from operations at weigh-in-motion (WIM) stations, were processed and filtered through data matching algorithms to generate real-time estimates of corridor level travel times. The first algorithm matched transponder equipped trucks in a time window between upstream and downstream stations on each corridor. The second algorithm filtered these matches for through trucks. The resulting data represented corridor level travel times for trucks over a two year period (2007-2008). To determine if these data could be used to represent travel times for passenger cars, field observations were made using probe vehicles to compare truck travel time performance with car travel time performance.

Findings from the study offer the following lessons learned:

Consider using truck transponder data to monitor various freight corridor performance measures. For freight corridors, promising freight performance measures that can be evaluated using the transponder linked WIM data include: average travel time on key corridors; ton-miles on each corridor by various temporal considerations; overweight vehicles on corridors by temporal variation; empty vehicles; seasonal variability in loading, routes, and volumes; percent trucks with tags on each corridor; potentially estimating an origin-destination matrix; and average weight for various configurations.

Beware of limitations in using truck transponder data to develop any real-time traveler information application. Transponder tag matching at successive weigh stations can be used to estimate long-term freight corridor performance. However, caution should be used when developing applications for real-time traveler information systems. The following challenges were noted.
  • The market penetration of transponder tags in freight vehicles is generally low and varies by location. The analysis indicated that on primary links there were enough trucks with tags (both numbers and frequency) to establish travel times between stations (assuming data issues can be resolved). However, on secondary links not enough trucks were available to generate real-time data.
  • The distance between weigh stations (reader locations) was relatively long. The report suggested that sensor spacing of 100 miles or less is reasonable. However, shorter spacing may be required around typical areas of delay and places where routes diverge. Additional sensors to read transponders could be installed to improve the accuracy and decrease the latency of travel time estimates. The hardware and labor for additional transponder readers were estimated to cost $9,000 each, not including the cost of integrating these sensors with the current WIM system data.
  • Trucks may stop for fuel, rest, and deliveries. More sophisticated algorithmic approaches could be applied to filter out trucks that made a stop between stations.
  • Comparisons to passenger cars may be difficult since truck travel speeds are different.
  • Recent advances in alternative traffic monitoring technologies such as cell phone, navigation devices, vehicle-to-vehicle, vehicle–to-infrastructure, and Media Access Control (MAC) address matching may be more suitable for providing real-time traveler information.
Address privacy issues in any public release of truck transponder data. Typically, state policy protects the identity of any individual truck or carrier associated with a transponder. Where performance analysis is required, but privacy must be maintained, database archives can be sanitized by linking data to fictitious names and numbers.

Performance measurement can help agencies achieve planned objectives in terms of improving mobility, reliability, safety, and reducing costs and environmental impacts. Overall, the findings of Oregon’s research indicate that truck transponder matching algorithms can support long-term monitoring of freight corridor performance. Applications for real-time traveler information, however, are limited when data quality is poor (e.g., incorrect time stamps). When data quality is good, the research indicates that average speeds and standard deviations can be calculated to detect corridor delays, especially weather based.

Lesson Comments

No comments posted to date

Comment on this Lesson

To comment on this lesson, fill in the information below and click on submit. An asterisk (*) indicates a required field. Your name and email address, if provided, will not be posted, but are to contact you, if needed to clarify your comments.


Developing Corridor-Level Truck Travel Time Estimates and Other Freight Performance Measures From Archived ITS Data

Author: Christopher M. Monsere, Michael Wolfe, Heba Alawakiel, and Max Stephens

Published By: Oregon Department of Transportation

Source Date: August 2009

URL: http://www.oregon.gov/ODOT/TD/TP_RES/docs/Reports/2009/Truck_Travel_Time.pdf

Lesson Contacts

Lesson Analyst:

Mike Mercer


Average User Rating

0 ( ratings)

Rate this Lesson

(click stars to rate)


Lesson of the Month for November, 2009 !

Lesson ID: 2009-00497