IMPROVING THE ACCURACY OF TRAVEL TIME ESTIMATES USING ARCHIVED ITS DATA Sirisha M. Kothuri*, Kristin Tufte, Soyoung Ahn and Robert L. Bertini Department of Civil and Environmental Engineering Portland State University P.O. Box 751, Portland, Oregon 97207-0751 Phone: 503-725-4285 Fax: 503-725-5950 Email: skothuri@cecs.pdx.edu ABSTRACT Advanced Traveler Information Systems (ATIS) are integral components of Intelligent Transportation Systems (ITS). These systems are aimed at providing transportation system users with pre-trip or en-route travel information, thereby allowing travelers to make informed decisions about their travel choices. Most urban travel occurs on freeways; hence, the accurate prediction of travel time over freeway networks has become necessary. In order to develop benchmarks for improved travel time estimates in the Portland, Oregon (USA) metropolitan region, the objective of this paper is to evaluate several travel time algorithms using archived loop detector data and ground truth (probe vehicle) data. KEYWORDS traveler information, travel time, archived data INTRODUCTION Improving the accuracy of travel time estimates is essential, especially when used for providing real time estimates to the public via variable message signs (VMS), as shown in Figure 1. Travel time estimates can be generated using different technologies such as inductive loop detectors, video detection, license plate matching, cellular phone matching and various other technologies [1]. The most popular method is to generate travel time estimates using data from inductive loop detectors. Figure 1: Sample VMS Display Travel times for different freeway links (corridors) were estimated using two algorithms, the Midpoint and Coifman algorithms, and compared to two ground truth data sets. The algorithms were used to estimate travel time using archived loop detector data from PORTAL – the Portland Oregon Regional Transportation Archive Listing, an Archived Data User