International Journal of Transportation Vol.5, No.3 (2017), pp.45-56 http://dx.doi.org/10.14257/ijt.2017.5.3.04 ISSN: 2287-7940 IJT Copyright ⓒ 2017 SERSC Australia Visualization Tools for Traffic Bottleneck Analysis Mark L. Franz 1* , Michael L. Pack 1 , Drew Lund 1 and Catherine Plaisant 1,2 1 Center for Advanced Transportation Technology Laboratory 2 Human-Computer Interaction Lab, University of Maryland, College Park *mfranz1@umd.edu, packml@umd.edu, alund1@umd.edu, plaisant@cs.umd.edu Abstract Traffic bottlenecks are a major source of highway user delay. Therefore, identifying bottlenecks and estimating their impacts is critical for making effective and efficient congestion management decisions. While traditional automated traffic recorders (ATR) provide valuable measurements of traffic flow characteristics, their network coverage is inherently limited. However, increased percentages of vehicles and travelers equipped with global positioning systems (GPS) have enabled transportation analysts to collect continuous and wide spread speed and travel time information from probe vehicles, thus allowing for comprehensive monitoring and analysis of the highway network. This paper presents the development and sample application of novel online bottleneck analysis and visualization tool to assist transportation professionals in identifying bottlenecks, and understanding their impacts using innovative data visualizations based on vehicle probe data analysis. The interactive visualizations include 1) maps illustrating the segments impacted by a given bottleneck(s), 2) timelines displaying the evolution of congestion over one or multiple days, as well as incidents occurring during congestion, 3) charts showing the spatial and temporal impacts of a given bottleneck, and 4) tables summarizing the impacts and rankings of each bottleneck. These visualizations can be used to provide both planning and operational congestion management decision support. Keywords: Traffic Congestion, Data Visualization for Decision Support, Traffic Operations 1. Background and Objectives Traffic bottlenecks plague many of the world’s urban areas and are a major source of delay, wasted fuel [1], secondary crashes [2], and increased emissions [3]. With limited resources for congestion mitigation, highway agencies must strategically allocate funds to bottlenecks susceptible to improvement by engineering intervention. State-of-practice is for transportation analysts to run traffic simulation models or to use engineering judgment to allocate limited resources for congestion mitigation. Simulation takes significant time and effort from highly trained experts to calibrate each bottleneck location to ensure realistic representation of real-world conditions. On the other hand, engineering judgment is subject to potential observational biases regarding which bottlenecks are more severe than others. To address this issue, we developed an intuitive web-based tool for identifying, analyzing, comparing and visualizing traffic bottlenecks using real-time vehicle probe data. This tool is part of an existing University of Maryland online-suite of probe data tools called the Probe Data Analytics (PDA) Suite that is part of the Regional Integrated Transportation Information System (RITIS). The tool allows real time analysis as well as retrospective analysis to identify and rank bottlenecks for a user-defined region and time period. When this data is combined with additional data sources such as event/incident * Corresponding Author