Using wireless location technology to estimate origin destination matrices: A feasibility study

Brent Michael Tucker, Purdue University

Abstract

An origin-destination (OD) matrix is an important element of a travel demand model. It is also a difficult set of data to acquire. In recent years, new technologies have been deployed to monitor traffic in real time. One way that may replace old travel pattern data collection methods is through the use of wireless location technologies (WLT). WLT includes cell phones, GPS units, and any other devices that can be wirelessly located on a network. Several studies have used WLT devices to determine travel time on road segments. This study tests the feasibility of using wireless location technology to estimate origin destination matrices. The data requirements for estimating origin destination matrices are different from the data requirements to estimate travel time. The study focuses on the data requirements for penetration rate (percent of vehicles that have detectable WLT devices in the study area) as it affects the quality of the estimated OD matrix. Other related issues, such as location error, ping interval (how often the cell phone is located), privacy issues, and noise reduction (detected cell phones that are not making a trip) are also discussed. In the analysis for a 67-zone study area, it was determined that a penetration rate of 50% resulted in an OD matrix that was statistically similar to the “real” OD matrix. However, when the OD matrices estimated from a range of penetration rates were used in a travel demand model, the points of greatest congestion changed very little, even for lower penetration rates. This indicates that the use of wireless location technology is feasible for the estimation of OD matrices, subject to the other issues discussed in the study.

Degree

M.S.

Advisors

Fricker, Purdue University.

Subject Area

Civil engineering|Transportation

COinS