Guide to Truck Activity Data for Emissions Modeling (2019)

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.

7 Table 2.1 provides a typology of sources on vehicle fleet and activity data that could poten- tially provide truck-specific data. Each data source is associated with the MOVES activity and fleet data inputs that it could support. Some of the data sources must be used in conjunction with other sources to provide the complete information needed for MOVES inputs (e.g., travel demand models may provide data for three or four time periods, but not all 24 hours of the day). There also may be practical limitations to the acquisition and/or use of some of the listed data sources (e.g., high cost, limited coverage, or data owner restrictions on distribution). Table 2.2 identifies whether and how each of the data sources in Table 2.1 is discussed in this guide. Table 2.2 also references other resources providing guidance on the data source. Activi- ties already fully covered in Volumes 1 and 2 of NCHRP Web-Only Document 210 (Porter et al., 2014a, 2014b) are not covered in depth in this guide. There has been rapid growth in recent years in the availability of truck telematics data. Table 2.3 discusses specific telematics data sources in some detail, including data sources presented in this guide and data sources investigated but not used. These sources can potentially be used to gener- ate MOVES inputs for truck speed distributions or drive cycles, temporal and road type distribu- tions, and off-network data (starts, soak time, and idling). However, the ability to generate inputs specific to a local study area will depend upon the available, resource-constrained sample size and any confidentiality restrictions on geoprocessing. While most sources differentiate between medium- and heavy-duty trucks, a further breakdown of data by more detailed vehicle types is rarely available. Also, activity patterns can vary by vocation and geography, and the instrumented or sampled vehicles may not be representative of the truck fleet as a whole. S E C T I O N 2 Truck Fleet and Activity Data Sources

8 Guide to Truck Activity Data for Emissions Modeling Data Source V eh ic le P o p u la ti o n A g e D is tr ib u ti o n s V M T b y V eh ic le T yp e V M T b y R o ad T yp e an d /o r T im e o f D ay S p ee d D is tr ib u ti o n s O p er at in g M o d e / V S P D is tr ib u ti o n s S ta rt s an d H o te lli n g Traffic Operations Data ATR—classification counts ITS—detectors, video, Bluetooth Traffic Simulation Models Traffic simulation models Truck Flow Data and Forecasting Models Freight flows—FAF, TRANSEARCH Travel demand forecasting models Commercial Telematics Data Aggregated GPS and cellphone data ECU data Registration and Inspection Data State registrations Federal motor carrier—MCMIS Program registrations—IRP, SmartWay Other Field Data Collection License plate survey Facility databases (e.g., truck stops, port terminal gate data) Field surveys (observational, intercept) Field measurements (instrumented vehicles) ATR = automatic traffic recorder; ECU = engine control unit; FAF = Freight Analysis Framework; GPS = global positioning system; IRP = International Registration Plan; ITS = intelligent transportation systems; MCMIS = Motor Carrier Management Information System. Table 2.1. Truck fleet and activity data sources and MOVES inputs.

Truck Fleet and Activity Data Sources 9 Data Source Coverage Traffic Operations Data ATR—classification counts Volumes 1 and 2 of NCHRP Web-Only Document 210a provide information on use of ATR data to develop VMT and temporal distributions by vehicle/source type. ITS—detectors, video, Bluetooth Volumes 1 and 2 of NCHRP Web-Only Document 210 provide information on using ITS data to develop speed distributions for all vehicles. Some sources (e.g., video) could potentially be used to develop truck-specific distributions, but methods are not explored in this guide. Traffic Simulation Models Traffic simulation models Volumes 1 and 2 of NCHRP Web-Only Document 210 provide information and a tool for aggregating simulation model output (vehicle trajectories) into OMDs suitable for MOVES input. The procedure would be the same for trucks as it is for other vehicle types and is not repeated here. Truck Flow Data and Forecasting Models Freight flows—FAF, TRANSEARCH Volumes 1 and 2 of NCHRP Web-Only Document 210 describe how freight flow data can be used to estimate long-haul truck volumes. Travel demand forecasting models Volumes 1 and 2 of NCHRP Web-Only Document 210 provide information on the use of travel demand forecasting models to develop speed and VMT information, as well as to support VMT distributions by hour and road type. This information will be available from some models separately for trucks and cars and also possibly for medium- versus heavy-duty trucks. Methods for improving truck traffic forecasts are covered in other resources and are not treated in this guide. Commercial Telematics Data Aggregated GPS and cellphone data The following uses of GPS/cellphone data are described: Section 4.0—speed inputs, Section 6.0—start inputs, and Section 7.0—hotelling inputs. Case Studies #3, #4, #5, and #6 explore these uses in more detail. ECU data Sections 6.0 and 7.0 describe how ECU data combined with GPS data can provide better information on starts and hotelling, respectively. Case Studies #3, #4, and #6 explore these uses in more detail. Registration and Inspection Data State registrations Volumes 1 and 2 of NCHRP Web-Only Document 210 provide information on the use of state registration data to develop age distributions and source type populations by vehicle/source type. Federal motor carrier—MCMIS Section 3.0 and Case Study #2 describe the use of MCMIS for developing data on local age distributions. Other Field Data Collection License plate survey Section 3.0 describes the use of license plate field surveys for developing data on local age distributions. Facility databases (e.g., truck stops, port terminal gate data) Section 7.0 and Case Study #4 describe how facility databases might be used in conjunction with GPS data to improve estimates of off-network activity. Field surveys (observational, intercept) Section 7.0 discusses methods to gather field data to support hotelling inputs for MOVES and, in general, how field studies may still be needed to capture more global data that instrumented vehicle samples cannot capture. Truck stop databases Section 7.0 discusses how commercially available truck stop databases can be used in conjunction with field studies to generate broad estimates of truck hotelling activity within an area. aPorter et al., 2014a, 2014b. Table 2.2. Coverage of truck data sources in this guide.

10 Guide to Truck Activity Data for Emissions Modeling Data set Description Outcome Sources Acquired/Used California Air Resources Board (CARB)— instrumented trucks A set of 100 instrumented trucks collected by the University of California (UC)—Riverside for CARB for research purposes. The data include engine on/off data and a cross-section of truck types. The NCHRP Project 08-101 research team obtained permission from CARB to use the data for NCHRP research purposes and has evaluated OMDs. Information is provided in Case Study #1. National Renewable Energy Laboratory (NREL), Fleet DNA Drive cycles collected from nearly 500 trucks nationwide, publicly available (not georeferenced). Drayage truck OMDs were compared against OMDs from Port of Los Angeles/Long Beach trucks (CARB data set) and other sources. Information is provided in Case Study #1. StreetLight Collects GPS and cellphone data for light-, medium-, and heavy-duty vehicles; performs custom analysis. The NCHRP Project 08-101 team purchased processed data from three counties to investigate the use of the data to look at start patterns. Information is provided in Section 6.0 and Case Study #3. Vnomics Collects truck activity and ECU data for instrumented fleets. The NCHRP Project 08-101 team purchased raw, anonymized data for both a sample of 300 trucks nationwide and a sample of 100 trucks each in three counties. Various items related to starts, soak times, and idling were evaluated. Information is provided in Sections 6.0 and 7.0 and Case Studies #4 and #6. Sources Not Acquired American Transportation Research Institute GPS and cellphone data—speeds and trip patterns from trucks nationwide. This source was not acquired due to lack of engine on/off data. It could potentially be applied with the same limitations presented in Case Study #3 for the StreetLight data. Verizon On-board diagnostic (OBD) port data collected for insurance purposes. Raw data are not made publicly available. The NCHRP Project 08-101 team was unable to complete negotiations for acquisition of custom- processed data. Transport Canada— Canadian Vehicle Use Survey Includes second-by-second GPS and ECU information for about 3,000 trucks. Raw data are not made publicly available. The NCHRP Project 08-101 team was unable to complete negotiations for acquisition of custom- processed data. California Department of Transportation (Caltrans) Truck Survey Freight truck activity and use data in California, collected via interviews, surveys, and GPS. This survey, modeled after the National Vehicle Inventory and Use Survey (VIUS), was in the data collection stage at the time this research was conducted (2016–2017). U.S. VIUS Physical and operational characteristics of a sample of U.S. trucks. VIUS was last conducted in 2002; a new VIUS is awaiting funding as of March 2019. Table 2.3. Truck telematics data sources investigated.