Making Trucks Count: Innovative Strategies for Obtaining Comprehensive Truck Activity Data (2014)

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.

1 The objective of NCFRP Project 39: Making Trucks Count is to suggest innovative approaches to obtain and make publicly available comprehensive truck-activity data. To fulfill this objec- tive, the research team reviewed the current state of truck-activity data and assessed data issues and limitations to determine critical gaps. Based on this review, the team proposed strategies for obtaining more comprehensive truck-activity data and conducted feasibility reviews of these techniques. After feedback from the project oversight panel, the research team then provided in-depth implementation strategies for the three most promising strat- egies. The research methods for this study included literature reviews, dataset evaluations, and key informant interviews. Implementing a national freight policy, as required by the Moving Ahead for Progress in the 21st Century Act (MAP-21), requires strategic planning and data. MAP-21 also focuses directly on expanding the use and application of performance measures and performance management, for which better freight data will be critical. Although such planning and data must be multimodal, documenting trucking activity will be most important: trucking comprises about 70 percent of total tonnage and value shipped in the United States, and is growing (U.S. Bureau of the Census Commodity Flow Survey, 2007). Effective policy making will require information to answer such questions as the following: 1. How much freight is moved by trucks? The answer to this question is helpful not only for planning purposes but also as a leading indicator of the economic health of the country. 2. What types of freight are moved by trucks? Vehicle type, weight, and origin and destination vary by type of freight, which also has differing impacts on the transportation system, air quality, and land use. 3. How much road traffic is generated by the movement of freight? Data on truck-vehicle miles can help analyze the relationship between the total volume of freight and the amount of vehicle traffic required to carry it, as well as how this traffic contributes to congestion and environmental degradation. Answering these questions requires comprehensive, longitudinal statistics that are not available. Currently, there is no public dataset with comprehensive, longitudinal statistics on highway-truck freight activity. In part, this is because trucking activity is highly fragmented, with hundreds of thousands of businesses, varying enormously in size, and shipping myr- iad goods over innumerable routes. Yet, given its scale and complexity, it is impossible to analyze and understand highway-truck freight activity in the absence of large statistical databases. S U M M A R Y Making Trucks Count: Innovative Strategies for Obtaining Comprehensive Truck Activity Data

2Information Gaps Early in the study, the team identified the data elements that were most important for freight policy and planning needs, assessed their availability and accessibility, and used this information to begin determining potential innovative strategies for ameliorating current coverage limitations. The data elements were as follows: • Vehicle Miles Traveled (VMT): Measure of the extent of motor vehicle operation within a specific geographic area over a given period of time. • Tons/Ton-Miles: Total weight of the entire shipment/shipment weight multiplied by the mileage traveled by the shipment. • Value/Value-Miles: Market value of shipments multiplied by the mileage traveled by the shipment. • Origin-Destination (O/D) Flows: The start and end points for a particular truck trip. • Vehicle Speed: Velocity of a vehicle. • Transportation Cost: Cost of freight movement by truck. The research team focused not only on these specific research elements, but also on how comprehensive current sources were in covering these variables by commodity type, vehicle type, vehicle characteristics, and different levels of geography. The team’s initial assessment considered the range of ways for obtaining truck-activity variables: surveys, administrative records, synthetic or amalgamated datasets, roadway operations data sources, and modeling approaches. The key findings of this assessment follow: • Because of the patchwork nature of the truck-activity data infrastructure, users must piece together information from many sources to answer critical policy questions. • The Commodity Flow Survey (CFS) is the basis for understanding freight in America. For data on tons and ton-miles of goods transported, all current data sources depend on the CFS. This survey is conducted only every 5 years and does not cover all shipments by truck. • Usability is a significant issue for many current data sources, which require a great deal of technical knowledge to use and visualize the data. • There is no method for verifying modeled truck-activity data, such as the Freight Analysis Framework (FAF) or Transearch. • Transparency in data sources and auditing procedures is a problem for nearly all the reviewed data sources. The only source that comes close to being sufficiently transparent is CFS. All other sources have problems ranging from private access to data, to unclear or obscured processing procedures, to lack of standardized collections, auditing, and unifor- mity of data collection periods. • Two ways to offer an improved, comprehensive source are (1) expanding an existing source to contain more needed measures than currently available or (2) a method, such as the FAF, for integrating and synthesizing data and yielding more of the needed measures. • A new innovative source for comprehensive truck-activity data needs to focus not only on data elements but also on statistical rigor, quality assurance, and accessibility. Innovative Strategies In identifying possible new approaches, the research team focused on improving existing approaches rather than creating completely new ones. Hence, the team characterizes the strategies addressed in this study as “relative” innovations—not “disruptive” ones. Given the institutional and political climates in which such improvements must be made, this seemed

3 to be a reasonable and realistic approach. These are approaches that could yield meaningful results in 5 to 7 years. The researchers considered innovation as a process of evolution rather than a revolution in methods for obtaining data. Such evolution offers new ways of obtain- ing data—either more efficiently, more effectively, or both. In the short term, the proposed strategies may provide opportunities for analyses that inspire a new perspective or for the collection of data that break down traditional institutional silos. The team identified 11 innovative strategies that could overcome limitations of current data sources. These strategies fell into one of four categories: (1) survey approaches, (2) pas- sively collected data, (3) administrative data, and (4) modeling approaches. The researchers assessed the feasibility of these strategies. This included assessing the ability to collect the data as well as technical, institutional, operational, geographic, and financial issues. A star rating system was used to compare approaches. The research team identified three approaches that were most likely to be implemented successfully, and developed detailed implementation scenarios for them. These were 1. Using GPS traces to understand trucking activities. Trucks equipped with GPS, or drivers carrying GPS-enabled devices (e.g., smart phones), create traces for the movement of each truck. Amassing this data using cloud technologies would create an innovative dataset on origins and destinations by time of day. Appending multiple attributes such as commodity or ownership can create many additional dimensions of truck activity. 2. Re-conceptualized Vehicle Inventory and Use Survey (VIUS). The TIUS/VIUS (Truck Inventory and Use Survey/Vehicle Inventory and Use Survey) series, which was initiated in 1963 and conducted as part of the economic census every 5 years starting in 1967, ended in 2002 due to financial constraints. This survey was central to national and state-level measures of the scope and character of the trucking industry. This strategy would entail reestablishing the survey system and consider opportunities or requirements for modifying and expanding approach and content. 3. Agent-based models for freight transportation. Agent-based modeling (ABM) is a new approach in freight transportation modeling. A main characteristic of it is modeling “firms” as the decision-making units. It seeks to improve typical four-stage state-of-practice models through better representation of firm behavior and their logistics and supply chain decisions. NCFRP Project 39 research addresses innovative strategies for obtaining truck-activity data. Such strategies typically include data “collection” or “capture.” This strategy deals with data “production,” that is, data that are the outcome of model simulations. The first two programs could be operational within 5 years. The third is on a longer term development schedule. Report Organization This report is organized as follows. Chapter 1 introduces the research need and objectives. Chapter 2 provides the results of the assessment of current primary data sources. Chapter 3 introduces the menu of innovative strategies compiled in this study and provides a feasibil- ity assessment for each. Chapter 4 presents three implementation scenarios for the strate- gies considered most feasible for obtaining truck-activity data. Chapter 5 summarizes the research needs and provides conclusions.