Performance Measurement of Transportation Systems: Summary of the Fourth International Conference (2013)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

39 BREAKOUT SESSION 3-B Transforming Experiences From Data to Measures, Measures to Information, and Information to Decisions with Data Fusion and Visualization Robert M. Winick, Motion Maps, LLC (Moderator) William Bachman, GeoStats Yoshihide Sekimoto, University of Tokyo Patricia S. Hu, Bureau of Transportation Statistics, U.S. Department of Transportation terabytes to deCisions: the inCreasing Potential of loCation-based analytiCs for transPortation Planning William Bachman discussed how location-based data can be used to improve transportation planning. His presentation highlighted existing commercially available databases and described their potential applications. • Location-based data are widely available today. Some areas are beginning to use location-based data in a broad range of transportation planning applications. There are numerous opportunities to expand the use of location-based data in transportation planning, perfor- mance measurement, and decision making. • INRIX, AirSage, and TomTom Traffic Stats are examples of commercially available sources of location- based data. – INRIX, which publishes an annual report on congestion in urban areas throughout the country, provides a rich source of data on historical traffic pat- terns. INRIX data are now being used in the mobility reports published by the Texas A&M Transportation Institute and in before-and-after studies, bottleneck analyses, and support for congestion-monitoring programs. The INRIX database provides for flexible aggregation of travel time and speed. – AirSage origin–destination (O-D) data provide information on movement patterns and long-distance travel patterns. The data are used in support of travel demand models. – TomTom Traffic Stats provides speed and travel- time data based on routes and segments. It boasts online query capabilities, O-D data, and a wide range of Global Positioning System data from sponsored surveys. • Although these data are intrinsically attractive, there are still questions concerning their use in transpor- tation planning, performance measurement, and travel forecasting models. The data requirements of the vari- ous topic areas associated with transportation planning are different. For example, the data needs associated with system performance evaluation include separating data on nonrecurrent and recurrent congestion, focus- ing on arterials and intersections, and capturing speed and volume data. Transportation development models and maintenance models require data on baseline speed by link and class level, O-D travel times, bottlenecks, and travel behavior. The data needs related to the effec- tiveness of transportation improvement include before- and-after delay data, facility measures, and regional measures. Also needed are data on real-time speeds and travel time, arterials and intersections, atypical events, intelligent transportation systems, and traffic operations. • Historical travel speed data are available now and are being used by some metropolitan planning organi- zations and state departments of transportation (DOTs) in network performance studies. The use of data from cell phones and other sources is being included in recent requests for proposals from metropolitan planning orga- nizations and other agencies. • Location-based data can also capture atypical events and provide information on the movement of freight and

40 performance measurement of transportation systems goods. These data can also be used in activity-based mod- eling and travel blending. Still other applications include community planning, analyzing livability and mobility, and safety planning. • There are different levels of data needs and fixed data collection standards and methods. For example, decision makers in state DOTs are interested in return on investment, while division managers may be inter- ested in regional percentage of delay reduction. A project manager would focus on measures such as percentage of delay reduction, stopped time delay, and percentage of travel time change. A consultant team would possibly be interested in lengthy formatted reports, methods, sched- ules, and analysis. Task lead engineers and analysts are focused on maps, charts, tables, formatted reports, and raw and processed data. • In conclusion, location-based data solutions are increasing the ability of planners to accurately under- stand existing conditions. Scientific methods are still valid with these data sources. The market value of the data is not clear, however. Questions on data quality and bias exist, but should diminish with frequency of use and acceptance. transforming exPerienCes in transPortation systems in JaPan Yoshihide Sekimoto discussed the following projects underway in Japan for collecting traffic and pedestrian data: a new initiative by Japan’s Ministry of Land, Infra- structure, Transport and Tourism (MLIT) for collecting and analyzing road traffic data, the People Flow Proj- ect (PFLOW), and a geospatial information exchange involving the national government and local govern- ments. Sekimoto’s presentation covered the following points: • The MLIT project focuses on new methods for collecting and analyzing traffic data. It is intended to address some of the issues associated with current data collection techniques. • Japan conducts a road traffic census once every 5 years. Human observers positioned at 24,000 locations throughout the country for 1 day conduct a traffic vol- ume survey. The survey is an efficient, low-cost method of data collection. The data for the particular day in the fall are used as the average annual road traffic data. This approach is not adequate for detailed analysis. Because the daily distribution of traffic volume fluctuates widely, determining recurrent and nonrecurrent congestion on the basis of only 1 day of observation is difficult. • In the future, traffic volumes will be obtained by observation of key road sections 24 hours a day, 365 days a year. The constant observation data will be used to estimate the traffic volume of other road sections. To date, the collection of travel speed data from probe vehicles has been limited (Figure 6). In the future, how- ever, collection of probe information from vehicles on a 24-hour basis should be possible. Upward holiday speed (national routes directly managed) 7:00 in October 2009Upward weekday speed (national routes directly managed) NA Less than 10 km/h Less than 20 km/h Less than 30 km/h Less than 40 km/h Less than 50 km/h 50 km/h or more Travel speed (Unit: km/h) FIGURE 6 Nationwide travel speed in Japan, October 2009.

41transforming experiences • New result indices are being introduced to evalu- ate the smoothness of traffic. One of these focuses on a quantitative evaluation of time loss caused by traffic congestion. The time loss is the extra time it takes to make a trip under congested conditions. The time-loss rate is the ratio of time loss to car utilization hours. In fiscal 2009, car use on national routes throughout Japan was approximately 13.3 billion hours. The number of hours required was approximately 8.3 billion, so some 5 billion hours were lost as a result of congestion. • The PFLOW project addresses the increasing need for time-based location information for large segments of the population. The recent earthquake in Japan is one example of this need. Having real-time data may help prevent or reduce the impact of secondary disasters in complex urban settings. Real-time data may also be used to help manage special events involving large groups of people. • The PFLOW project uses large-scale, real-world data to reconstruct macroscopic people flow. Person trip data are available from the personal travel survey con- ducted by the Tokyo Metropolitan Region Transporta- tion Planning Commission. O-D information can also be obtained. Data from these and other sources are used in the spatiotemporal modeling of people flow. • The PFLOW data set is being used in numerous research projects that cover a wide range of areas, includ- ing transportation, spatiotemporal analysis, risk analy- sis, personal information, security, the environment, and marketing. • The experimental geospatial information exchange consortium was initiated in FY 2009–2010. The national government and local governments provide data to the consortium exchange platform. The data users, which include various companies and organizations, can access the data via the exchange program. There are 125 con- sortium members, of which 85 are private companies. Other members include foundations, universities, asso- ciations, nonprofit organizations, national organizations, and local governments. The exchange platform contains a wide range of data from different sources. Data fusion is supporting the logistics response to the recent earthquake and tsunami. enhanCing PerformanCe measurement and deCision making With visualization tools Patricia S. Hu discussed visualization tools for enhanc- ing the presentation of data used in performance mea- surement. She provided examples from the Oak Ridge National Laboratory (ORNL), the Bureau of Transpor- tation Statistics, the U.S. DOT, the National Geospatial- Intelligence Agency, and other groups. Hu’s presentation covered the following points: • A visualization tool is software that provides visual displays and representations. One of the key benefits of visualization tools is the ability to integrate many differ- ent layers of data and analytical capabilities. Visualiza- tion tools can range from the relatively simple to the very complex, depending on use and the desired analysis and display capabilities. The size of the spatial coverage can add to the complexity of the analysis. • Visualization tools integrate layers of data. Users select the layers of data needed to analyze a specific ques- tion or issue. Visualization tools are also scalable and adaptable; additional layers can be added or removed as needed. Visualization tools can be used for the following purposes: – To communicate large amounts of data, system performance, and other measures; – To present the results of simulation models and analyses; and – To plan scenarios for special events and emer- gency responses. • One application at ORNL focused on displaying freight data. Data from the Freight Analysis Frame- work were used to examine freight flows in a corridor. Another application examined the travel needs of older individuals. • Several visualization tools were demonstrated, including tools on the websites of the Bureau of Trans- portation Statistics and the U.S. DOT and the Bomb Card developed by ORNL. This application can be used by first responders to determine the evacuation distance and take-cover distance related to differ- ent types of bomb threats. The number of buildings and employees on the ORNL campus that would be affected can be identified. Another visualization tool can be used to evaluate different routes for the ship- ment of hazardous materials. The routes displayed depend on the criteria selected (e.g., shortest distance, avoiding major population centers, avoiding environ- mentally sensitive areas). • Several lessons have been learned from the expe- rience to date in developing and using some of these visualization tools. Collaborating and working with stakeholders is critical. Identifying data needs, data availability, and data compatibility is also important, as is fusing the data. Additional collaboration with stake- holders as the tool is being developed is needed to ensure that it meets the intended purposes and the capabilities of the users. Ongoing refinements and enhancements are common after the tool is developed.