tional improvements in real time and in retrospect. At a broader level, archiving arterial performance data over time would allow planners to monitor more than just the freeway portions of the net- work, which is useful for congestion monitoring and regional transportation planning. One reason that arterial performance systems are not as mature as those for freeways is that interrupted flow facilities are more diffi- cult to monitor. Travel time along an arterial can be broken into two distinct components: the time to traverse the links between traffic signal influence areas and the time to traverse the signalized inter- section itself. When unaffected by queues, traffic between nodes can be treated like that on an uninterrupted flow facility and point-based speed measurements can be extrapolated over segments of influence (e.g., from loop detectors). Travel time measurements through inter- sections require more detailed measurements about queue formation and dissipation, as well as the status of the traffic signal settings. Given these challenges, recent research has been directed toward improving techniques to monitor arterial performance (1–11). Some research has measured each component directly (3–5), and some has modeled relationships between traffic flow and signal control settings (6, 7 ). Another approach addresses the data collection problem by predicting arterial corridor travel time based on count and lane occupancy data from corridor signal loop detectors, green times, cycle lengths, and offsets (9). The use of individual vehicle actuation data rather than arbitrary temporally aggregated data (e.g., 30-s data) for performance measurement at isolated signals was examined (10), and the effect of fusing these data with signal timing data was tested (11). Although these approaches are very promising, high-resolution data are not currently available from most signal systems. This paper builds on past research in arterial performance based on signal detection as well as past research of transit vehicle geolo- cation systems as a source of probe data for real-time arterial mon- itoring and travel time reporting (12–14). Transit and other fleet vehicles are increasingly equipped with automatic vehicle location (AVL) systems and other Global Positioning System–like devices. In Portland the 700 buses in TriMet’s system are “smart,” in that their bus dispatch system (BDS) monitors and archives their exact location, schedule status (on-time, late, or early), passenger load and individual passenger movements, and other important perfor- mance parameters. From a historical perspective, these data can be processed to provide a comprehensive picture of current arte- rial speed across all arterials that contain bus routes. As an illus- tration of this, Figure 1 shows a map of the Portland metro area with bus speeds (indicated by the color code shown on the figure) Prototype for Data Fusion Using Stationary and Mobile Data Sources for Improved Arterial Performance Measurement Mathew Berkow, Christopher M. Monsere, Peter Koonce, Robert L. Bertini, and Michael Wolfe Arterial performance measurement is a critical issue for transportation system management, traveler information, and real-time situation-aware routing. In many urban areas current information on freeway conditions is available, appropriately given the large amount of travel that occurs on these facilities. However, because nearly 40% of the vehicle miles traveled in the United States occur on arterials, there is a need to pro- vide similar information that can be used not only by travelers but also by traffic engineers and managers. Because many arterials are equipped with actuated traffic signals, the use of already installed sensors has been explored as one source of traffic volume, occupancy, or speed data to inform an arterial performance system. Coupled with this, there is a potential to exploit the availability of mobile probe geolocation data from sources such as automatic vehicle location systems for fleets of buses or taxis, or from cellular phone or other Global Positioning System–type devices. To demonstrate the potential value of fusing data from fixed and mobile surveillance systems toward improved arterial performance reporting, this paper describes the results of a case study from Portland, Oregon, that extracted improved arterial performance measures by combining data from traffic signal system detectors and from buses act- ing as probe vehicles. In particular, graphical techniques are developed that trace the boundaries of the congested regime in time and space along an arterial corridor. The paper includes recommendations for expanding the techniques to other corridors, using higher resolution, real-time transit location data, and online implementation of an arterial travel time information system. Most urban areas in the United States offer real-time freeway speed, travel time, and incident information, but there is a need to expand traffic management and traveler information to the arterial system because it handles nearly 40% of the nation’s vehicle miles traveled. It is clear that for many travelers an arterial performance measure- ment system would be useful for trip planning, en route information, and dynamic routing. In addition, traffic engineers and transporta- tion system managers would benefit by being able to continuously evaluate the effectiveness of signal timing plans and other opera- M. Berkow, Alta Planning + Design, 711 Southeast Grand Avenue, Portland, OR 97214. P. Koonce, Kittelson & Associates, Inc., 610 Southwest Alder Street, Suite 700, Portland, OR 97205. C. M. Monsere, R. L. Bertini, and M. Wolfe, Department of Civil and Environmental Engineering, Portland State University, P.O. Box 751, Portland, OR 97207. Corresponding author: R. L. Bertini, bertini@pdx.edu. Transportation Research Record: Journal of the Transportation Research Board, No. 2099, Transportation Research Board of the National Academies, Washington, D.C., 2009, pp. 102–112. DOI: 10.3141/2099-12 102