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.

4C H A P T E R 1 Trucks play an essential role in the freight transporta- tion system. According to the 2007 CFS, trucks carried about 85 percent of total tonnage and of total value shipped in the United States. Trucks provide enormous flexibility in the ori- gins and destinations they can serve, the commodities they can carry, and the range of services they can provide. They also provide the key link among most other modes of freight transportation. With dependence upon just-in-time inven- tory practices, forward positioning of supplies and inventory, and continued growth in small-package expedited delivery and e-commerce distribution services, the significance of truck traffic will only grow. Data on truck activity have played an essential role in public- and private-sector decision-making in recent decades. In the public sector, the U.S. Department of Transportation (DOT) has used truck-activity data for analysis of cost allocation, safety issues, proposed investments in new roads and technology, and user fees. The Environmental Protection Agency has used the data to determine per mile vehicle emission estimates, vehicle performance and fuel economy, and fuel conservation practices of the trucking industry. The Bureau of Economic Analysis has used the data as a part of the framework for the national investment and personal consumption expenditures compo- nent of the Gross Domestic Product. Beginning with the pas- sage of the Intermodal Surface Transportation Efficiency Act (ISTEA), continuing with the Transportation Equity Act for the 21st Century (TEA-21) and with the Safe, Accountable, Flexible, Efficient Transportation Equity Act: A Legacy for Users (SAFETEA-LU, 2005), states and metropolitan planning orga- nizations (MPOs) are required to consider freight transporta- tion issues in state and metropolitan transportation plans. Passage of the Moving Ahead for Progress in the 21st Cen- tury Act (MAP-21) created a new national freight policy. The bill also calls for the creation of a National Freight Strategic Plan and the establishment of a Freight Policy Council to be involved in developing it. The national freight policy and the strategic plan are conceived as multimodal endeavors; never- theless, trucking activity on the highways will be an important component. Strategic planning will rely on data since analysis of these data will provide the information needed to inform policy and investment decisions. As a 1992 TRB special report noted “without good data, decisions will be arbitrary, options overlooked, and solutions misguided (TRB, 1992).” MAP-21 also focuses directly on expansion of directions in performance measures and performance management, for which better freight data will be critical. In the private sector, tire manufacturers have used data to calculate the longevity of products and to determine the usage and applications of their products. Heavy machinery manufacturers have used data to track the importance of var- ious parts distribution and service networks. Truck manufac- turers have used data to determine the impact of certain types of equipment on fuel efficiency. Though truck-activity data are increasingly important in both the public and private sectors, the number of sources for Introduction CHAPTER 1 KEY TAKEAWAYS • Significance of truck data for public and private decision-making will only continue to grow. • The number of data sources is dwindling and many have completeness and accuracy problems. • Data are lacking to answer key policy questions of how much freight is moved, what types of freight are moved, and how much road traffic these movements generate. • When addressing freight data limitations, it is important to focus on both data and industry coverage issues.

5 such data has dwindled in recent decades, and many sources have completeness and accuracy problems. Truck-activity data for decision making and performance measurement are criti- cally lacking, and specific strategies for improving the sources of such data are needed now. Gaps in travel data have been enumerated in myriad studies, in congressional testimony, and by noted leaders in transportation policy and planning (Trans- portation Research Board, 2011). Although gaps exist for both passenger and freight data, freight (particularly, truck activity) data have serious gaps that can only be filled by new approaches in obtaining data. Freight data gaps are more costly to fill than passenger data gaps. Reasons for this difference include the complexity of the data, the potentially proprietary nature of its source, and inaccuracies that may be prevalent but time- consuming to detect. Freight data are increasingly important, as well as increasingly incomplete or missing. In the team’s examination of this topic, the researchers focus primarily on publically available sources of data. However, the research team acknowledges that there are some very impor- tant private data sources. For example, R. L. Polk is a provider of vehicle data related to characteristics, transactions (e.g., ownership history), mileage, etc. Such information not only has value in itself but also has use in the production of pub- lic data (i.e., VIUS has used a Polk database as the sampling frame). In addition to isolating freight data gaps, determining appro- priate data architectures, or developing freight system perfor- mance measures, the researchers focus on identifying creative, cost-effective methods for obtaining comprehensive trucking activity data and providing models for their implementation. Key research questions include the following: • What can be learned from the current state of the practice for obtaining and reporting these data in the United States and other countries to inform innovation in strategies for data gathering? • Given the state of the practice, what data issues and limi- tations exist that could be overcome through innovative strategies for data gathering or reporting? • What innovative data gathering or reporting strategies can be applied to overcome current challenges and derive a complete picture of trucking activity? • What are the recommended innovative and improved strategies for obtaining truck-activity data, and what are the risks, barriers, lessons learned, and best practices asso- ciated with their implementation? 1.1 Data Coverage Issues The United States has had many sources of truck-activity statistics. Yet many of these have problems, some are no lon- ger available, and the utility for others is questionable. This research attempts to rectify the situation by identifying and evaluating innovative solutions for improving the gap in truck- ing data. The following key policy questions guide the need for truck-activity data: • How much freight is moved by trucks? • What types of freight are moved by trucks? • How much road traffic is generated by the movement of freight? Answering these questions requires, at a minimum, data and information such as the following: • Vehicle Miles Traveled (VMT): Measure of the extent of motor vehicle operation within a specific geographic area over a given period of time. Estimates of VMT are used extensively for allocating resources, estimating vehicle emissions, computing energy consumption, and assess- ing traffic impact. The Highway Performance Monitoring System (HPMS) is the best current source of this informa- tion for trucks providing national, state, urbanized, small urban, and rural area coverage, often measuring truck- vehicle miles as the residual after considering passenger vehicles. • Tons/Ton-Miles: Total weight of the entire shipment/ shipment weight multiplied by the mileage traveled by the shipment. Ton-miles provides the best single measure of the overall demand for freight transportation services. This is also a measure of the overall level of industrial activity in the economy. A ton-mile estimate is necessary to construct other estimates of transportation system performance, such as energy efficiency and accident, injury, and fatality rates. The CFS data are the best publicly available source to construct trucking ton-miles estimates, but coverage is incomplete among industries. • Value/Value-Miles: Market value of goods shipped multi- plied by the mileage traveled by the shipment. The choice of transportation modes or combinations of modes depends on characteristics such as type and value of commodity, dis- tance, and desired speed and reliability of transportation. This measure can be used for assessing economic trade-offs over distance; it captures increases in activity for goods that tend to have an inverse relationship between technologi- cal improvement and weight. Growth in demand for high- value, time-sensitive goods has driven the forecast growth of trucking activity. Value is available from the CFS but value-miles is not available from any source. • Origin-Destination (O/D) Flows: The start and end points for a particular truck trip. Measuring the sizes of O/D flows is important to many network management applications such as capacity planning, traffic engineering, anomaly detection, and network reliability analysis. Two public sources provide

6information on truck O/D flows—the survey-based CFS and the FAF, which is based on synthetic or modeled infor- mation from the CFS. CFS captures freight flows, although incompletely, at the national level. Private data sources provide more geographically detailed origin-destination flows for some types of trucking flows and commodity movements, but have coverage and transparency prob- lems. As described in NCFRP Report 26, supplemental data collection is required to obtain local detail. • Vehicle Speed: Velocity of a vehicle. Vehicle speed, together with volume, is used as a measure of congestion and as a per- formance metric for roadway maintenance. The availability of speed data is extremely limited and constrained by the number and geographic coverage provided by roadside traf- fic counters, weigh in motion (WIM) stations, aerial photo- graphs, and limited samples of global positioning satellite (GPS)- or onboard diagnostic (OBD)-enabled truck fleets. Private-sector sources, such as INRIX and American Trans- portation Research Institute (ATRI), produce vehicle speed data but with access and coverage limitations. • Transportation Cost: Cost of freight movement by truck. Forecasting costs and evaluating potential responses from the private sector, particularly shippers and carriers, requires access to transportation cost data. Such information is quite important for mode diversion analyses and for assess- ing the transportation economic productivity and impacts on the economy. The collection of transportation cost data, however, was largely discontinued after deregulation. NCFRP Report 22: Freight Data Cost Elements identifies the specific types of direct freight transportation cost data elements required for public investment, policy, and regulatory deci- sion making and describes and assesses very general strate- gies for obtaining the needed cost data elements. Information gaps are further exacerbated when existing data are segmented by commodity type, vehicle type, and vehicle characteristics. To be “comprehensive” and most useful for policy and planning, data should include the following: • Commodity type: Products that an establishment produces, sells, or distributes. Such data, ideally, would use five-digit Standard Classification of Transported Goods (SCTG) codes. • Vehicle type: Based on the vehicle gross weight rating, rang- ing from light duty to medium duty to heavy duty. VIUS standards use eight classes. • Vehicle characteristics: Age, weight, length, fuel type, accel- eration capabilities, fuel efficiency, emissions, and range of operation, among other characteristics, of vehicle. Table 1-1 illustrates these information gaps and further char- acterizes these gaps in terms of weaknesses of quality or usabil- ity. Comprehensive truck-activity data would be represented by 100 percent filled cells in Table 1-1 and the absence of quality or usability issues. • Quality is defined as having “fitness for use” in terms of accuracy (i.e., when reality and what is recorded as data are in agreement), completeness (i.e., sufficient breadth and depth), coverage (i.e., how much of what is recorded is available), or reliability (i.e., consistency over time and with other major datasets). • Usability is defined as having “fitness for use” in terms of timeliness (i.e., at temporal scale that is sufficient for pur- pose or monthly/quarterly, peak/off-peak, etc.) and acces- sibility (i.e., easily retrievable). The information in Table 1-1 answers two questions: 1. Do data exist? If not, “missing” populates the cell. 2. If yes, is data of sufficient quality and usability? If there is a gap in either quality or usability, the gap is identi- fied in the cell. One conclusion from Table 1-1 is that being able to arrive at comprehensive truck-activity data requires filling in many data weaknesses. Given possible biases in determining what is most important to fill, as well as trade-offs in the perceived benefit/cost, the researchers opted for strategies that would fill the greatest number of data needs for the least incremental cost and risk. The team’s feasibility review in Chapter 3 dis- cusses how they arrived at such decisions. 1.2 Industry Coverage Issues This project focuses on truck-activity data—information about the over-the-road transportation of cargo using motor vehicles, such as trucks and tractor-trailers. This study uni- verse covers all trucks and tractor-trailers that move some sort of cargo, as well as those that provide services, regardless of industry as defined by the North American Industry Clas- sification System (NAICS). We use the term “universe” as it is used in statistical sampling to denote the entire aggregation of items from which samples can be drawn. The CFS, which is a comprehensive tool for understanding freight flows in the United States, samples from a universe of business establishments and auxiliary establishments (i.e., warehouses and managing offices) in the 50 states and the District of Columbia. Trucking activity (i.e., for-hire truck, private truck, parcel, or courier) is captured as one of several modes of transport for shipments emanating from sampled establishments. Other modes of transport captured are rail, air, water, and pipeline, all of which are outside the scope of this study. Improving the industry coverage of the CFS would entail expanding the number of sampled industries. By design, the CFS lacks coverage of establishments in forestry, fishing, utilities, construction, transportation, most retail and services

Truck-Activity Measure Segmentation Variable Level of Geography National State Metropolitan Facility VMT Vehicle Characteristicsa Vehicle characteristics data limited; VIUS discontinued. Missing Missing Missing Missing Vehicle Type FHWA VM1 Table using HPMS data; accessibility gaps (microdata) Exists from HPMS but not reported; accessibility gaps Missing Missing Commodity Type Derived from CFS in FAF Derived from CFS in FAF Derived from CFS in FAF Missing Ton Vehicle Characteristics Vehicle characteristics data limited; VIUS discontinued. Missing Missing Missing Missing Vehicle Type Missing Missing Missing WIM data coverage, accessibility issues Commodity Type Derived from CFS in FAF Derived from CFS in FAF Derived from CFS in FAF Missing Ton-Miles Vehicle Characteristics Vehicle characteristics data limited; VIUS discontinued. Missing Missing Missing Missing Vehicle Type Ton-miles available from CFS but not by vehicle type. Missing Missing Missing Missing Commodity Type Derived from CFS in FAF Derived from CFS in FAF Derived from CFS in FAF Missing a Vehicle characteristics include age, weight, length, fuel type, acceleration capabilities, fuel efficiency, emissions, and range of operation, among other characteristics of vehicle; whereas vehicle type is based on the vehicle gross weight rating, ranging from light duty to medium duty to heavy duty. b TMAS refers to the FHWA’s Traffic Monitoring Analysis System. TMC refers to the regional Traffic Management Centers. Value/Value- Miles Vehicle Characteristics Vehicle characteristics data limited; VIUS discontinued. Missing Missing Missing Missing Vehicle Type Missing Missing Missing Missing Commodity Type Value, but not value-miles available from CFS. Derived from CFS in FAF; availability issues Derived from CFS in FAF; availability issues Derived from CFS in FAF; availability issues Missing O/D Flows Vehicle Characteristics Vehicle characteristics data limited; VIUS discontinued. Derived from CFS in FAF Derived from CFS in FAF Derived from CFS in FAF and from MPO surveys; completeness, accessibility issues Derived from TMAS, TMC; coverage, completeness, accessibility issuesb Vehicle Type Commodity Type Value, but not value-miles available from CFS. Vehicle Speed Vehicle Characteristics Vehicle characteristics data limited; VIUS discontinued. Private sources; coverage, completeness, accessibility issues Private sources; coverage, completeness, accessibility issues Derived from TMAS, TMC; coverage, completeness, accessibility issues Derived from TMAS, TMC; coverage, completeness, accessibility issues Vehicle Type Commodity Type Value, but not value-miles available from CFS. Missing Missing Missing Missing Truck Transportation Costs Vehicle Characteristics Vehicle characteristics data limited; VIUS discontinued. Private sources; coverage, accessibility issues Missing Missing Missing Vehicle Type Services Annual Survey and private sources; coverage, accessibility issues Missing Missing Missing Commodity Type Value, but not value-miles available from CFS. Table 1-1. Quality and usability assessment of existing truck-activity data.

8industries, farms and government-owned entities, and for- eign-based business importing to the United States. In reviewing the ways in which the trucking industry universe has been presented in prior NCFRP reports, the team identified several different universe definitions. Not all NCFRP reports describe the trucking universe. Those that did were focused on the objectives of their particular study (e.g., performance measures, freight trucks as part of freight system, association with FAF truck definition, urban truck movements, revenue generation by firm type). This led to various universe definitions, noted as follows: • Truckload (TL), less-than-truckload (LTL), private truck- ing (NCFRP Project 2); • TL, LTL, bulk, general freight (NCFRP Project 3); • TL, LTL, specialized carriers, third-party logistics, other spe- cialized (bulk liquid, flatbed carriers) (NCFRP Project 4); • Interregional: TL, LTL, private trucks; intraregional ship- ments (i.e., parcel, delivery trucks to commercial/home); intraregional movements to support construction, utilities (i.e. telephone, contractors); fleet allocations and patrols that operate on fixed route (i.e., garbage truck, mail) (NCFRP Project 6); • LTL; intermodal; package and courier (NCFRP Project 11); • TL vs. LTL; single vs. combination trucks (NCFRP Proj- ect 12); • Long-haul trucks with both O-D outside urban area; long-haul trucks with pickup/delivery in region; truck drayage—the short-haul movements of intermodal con- tainers moving to and from railroad intermodal yards and marine container ports; local trucks moving goods among facilities on pickup and delivery runs within the region; construction vehicles (e.g. cement mixers, dump trucks, construction cranes); utility and other residential service vehicles (e.g., refuse trucks); trucks delivering freight with special requirements; package services (NCFRP Project 15); • Private fleet truck operations; for-hire TL companies; for- hire LTL trucking companies (includes parcel, express); drayage/cartage companies; brokerage/third-party logis- tics companies; specialized trucking companies (NCFRP Project19); • For-hire: TL, LTL, parcel delivery, drayage; private: long-haul, interplant, direct store delivery (NCFRP Project 27); and • Universal vehicle classification: Classes 1-2: primarily pick- ups, vans, and SUVs used for personal transportation or light service and delivery purposes; Class 3: includes the largest pickups and vans ordinarily used for personal trans- portation and the smallest trucks routinely used to carry goods, supplies, and equipment; Classes 4-6: medium- duty trucks, step vans, flat beds, small dump trucks, and other trucks used to move freight; Classes 7-8: heavy-duty trucks, both straight trucks and tractors for use with semi- trailers, predominantly used to move freight and very heavy service vehicles, such as concrete pumpers, cranes, and drilling equipment (NCFRP Project 29). For creating a more complete picture of truck activity, the universe must be as expansive as possible. For this reason, the team’s proposed trucking classification system organizes the trucking industry into the following four types of truck- ing activity. 1. For-hire carrier, 2. Owner-operator, 3. Shipper-owned trucking, and 4. Construction/utility/services. The fourth type, construction/utility/services, while impor- tant to clearly delineate the trucking industry universe, is not relevant to commodity-related trucking activity. Therefore, this category is not used further in this report. However, the other trucking activity categories are used in Chapters 3 and 4 to gauge the potential effectiveness of proposed strategies. This report also presents two types of geography—long distance/interstate and local.