Highway Performance Monitoring System Field Manual

Chapter 5 Guidance on Special Topics

5.1 Overview

The purpose of this chapter is to provide additional guidance on the reporting of Functional System, and Traffic data. This information is a supplement to the data item requirements discussed in Chapter 4.

5.2 Functional Classification System Descriptions and Groupings

FHWA focuses scarce national resources on the most important roads and highways in the Nation for condition and performance improvement purposes. This practice has been in existence on a national level since the 1960s. In order to ensure that the State Departments of Transportation (DOTs) have information needed to support this effort, FHWA disseminates Functional Classification guidance documentation on a periodic basis. The most recent comprehensive Functional Classification Guidance Document was published in 2013 and can be accessed online at: https://www.fhwa.dot.gov/policyinformation/hpms/hfcccp.cfm. This document should be used by the State DOTs as guidance for the purpose of updating and maintaining their respective FC Systems.

Functional classification is the process by which streets and highways are grouped into classes, or systems according to several factors that contribute to the overall importance of a given roadway to a region or area. All streets and highways are grouped into one of seven classes, depending on the character of the roadway and the degree of land access that they allow. The seven functional classes are represented by a one-digit code and are used to represent a specific classification of road regardless of whether it is located in an urban or rural area. These classifications are as follows:

The Role of Urban Areas in Roadway Functional Classification

The U.S. Census-based Urban Area (UA) Boundaries are also an important part of the FC system. The aforementioned FC codes and Census-defined UA Boundary codes (see full list of valid UA codes in Appendix I) must be reported for all Federal-aid roadways. Although an urban or rural designation does not dictate the functional classification of a roadway, it may inform classification designation. Because road usage and design typically adapt to changes in urban growth patterns, the urban boundary modification process commensurate with the Decennial Census is a good time to assess functional classification throughout the State.

Traditionally, the Census Bureau releases new Urban Area Boundaries two years after the initial Decennial Census as a byproduct of that effort. Since these boundaries are developed primarily through automated methods, they are often coarse and irregular, generally not reflective of transportation facilitates. While a State may choose to use the unadjusted original Census boundaries as part of the overall FC program, it is advisable to adjust these polygons to efficiently account for the highway system. FHWA guidance for procedures and best practices regarding Functional Classification and Urban Boundary delineation can be found in the Highway Functional Classification Criteria, Concepts and Procedures, 2013 Edition document.

See Chapter 4, Sec. 4.4 for specifications and requirements pertaining to the reporting of the ‘Functional System’ and ‘Urban Code’ data items. Spatial Analysis should be used by the States to relate the FC code to the UA code for HPMS reporting purposes.

5.3 Traffic Monitoring Procedures

Introduction

Traffic monitoring data are a key component of the HPMS. They are some of the most analyzed and used data elements and must be of high quality to accurately represent conditions in all States. Traffic data are used for a variety of work program objectives which include the following:

• Apportionment of various Federal-aid highway and safety-related funds;
• Analysis and presentation of data for the Status of the Nation’s Highways, Bridges, and Transit: Conditions and Performance report to Congress;
• Freight analysis including corridor, commodity, economic productivity size and weight, and others
• Clean Air Act travel data requirements for non-attainment areas;
• U.S. DOT performance measures such as vehicle crash rates, pavement condition, and congestion;
• Various FHWA national performance measures
• Analysis of vehicle types for freight movements and safety; and
• Technical support to policy and legislation development and implementation.

Traffic monitoring data are also key inputs for the development and maintenance of the HPMS data set. Traffic data drive the HPMS sample stratification and selection process by assigning roadway sections into volume groups and for statistical analysis to develop the sample panel as further discussed in Chapter 6. The validity of the entire HPMS sample panel and the development of the sample expansion procedure depends on the proper maintenance of a comprehensive traffic monitoring program.

A State traffic monitoring program that is developed following the guidance contained in the Traffic Monitoring Guide (TMG) will provide data that meets the needs of HPMS. The AASHTO Guidelines for Traffic Data Programs (AASHTO Guide) provides another reference for developing and maintaining a State Traffic Data Program. Since HPMS is a key driver for State’s traffic monitoring programs, States should use a combination of guidance from the TMG, HPMS Field Manual, and other sources such as the AASHTO Guide to develop their traffic program.

The traffic data reported in HPMS must be the same data the State uses for their own purposes as contained in their traffic monitoring system. Using the same data provides assurance that it was collected and processed following the State’s traffic monitoring program and not processed independently for HPMS. If the same data are used, then products from the HPMS data submittal are approximately the same as the State’s traffic data products such as VMT. In summary, the specific travel data needs for HPMS can be accommodated with minor adjustments and implementation of good practices as presented in the TMG and in the AASHTO Traffic Data Guidelines.

This section provides specific guidance for traffic monitoring procedures to meet the HPMS requirements and builds on the recommendations provided in the TMG. It is important to recognize that this Field Manual refers to traffic data in several sections: Sections 3.3, 4.3, etc. Stakeholders involved in collecting, analyzing and reporting on traffic data for HPMS should refer to this section as well as other references to traffic monitoring throughout the Field Manual. Stakeholders are encouraged to get familiar with FHWA’s Traffic Monitoring Guide to establish a comprehensive traffic monitoring program.

This section is presented in three parts:

1. General, high level requirements for the traffic monitoring program,
2. Volume Group Assignments for HPMS, and
3. More detailed traffic monitoring program guidelines.

State Traffic Monitoring Data Program

General Requirements

State maintenance of a comprehensive traffic monitoring data program to provide quality, timely, and complete traffic volume and vehicle classification data is important for meeting HPMS requirements. This section describes the fundamental macro-level requirements of a State Traffic Data Program for HPMS. Specific guidance is contained in the TMG and readers are encouraged to refer to the TMG for more detail.

While traffic data are collected at points on the highway system, HPMS is oriented toward roadway sections. So an initial step is for the State to segment their roadways into sections with consistent traffic. As highways evolve and traffic patterns change, these traffic monitoring sections may need to be revised. An advantage of the new HPMS data model is that States may submit section-level data for these traffic monitoring sections independent of other HPMS section-specific data.

Count Cycles and Coverage

A State should have minimum count cycles and coverage as follows:

Minimum 3-year count cycle – The State’s traffic monitoring program shall cover all NHS and Principal Arterial System (PAS) roadway sections (i.e., Interstates, Other Freeways and Expressways, and Other Principal Arterials) on a three-year cycle or better; at least one-third of these roadway sections should be counted each year. The remaining two-thirds counts must be estimated based on a documented process in accordance with the TMG and the Field Manual. The State shall cover all roads on these systems, not just State-owned roads, so data provided by MPOs, cities, or counties should be included in the count cycle.

Minimum 6-year count cycle – The State shall also have a traffic count program on a six-year cycle or better for all non-NHS lower functional system roadway sections (i.e., minor arterials, major collectors, and urban minor collectors). Traffic data for ramps, as defined in Chapter 4, are also to be collected on a six-year cycle or better.

All traffic data for HPMS shall be based on a minimum of 48 hours of continuous monitoring for volume and vehicle classification, which is referred to as short term monitoring. States are permitted to perform counting durations shorter than 48 hours for roadway functional classes Arterial and Interstate. For functional classes of collector and local roadways, if a State has a duration of monitoring that is less than 48 hours, they must be able to demonstrate no loss in quality of data based on documented statistical analysis provided to FHWA’s Office of Highway Policy Information via FHWA’s Division Office located in their respective States.

The program should provide for a sufficient number of Continuous Count Station (CCS) volume and continuous vehicle classification (CVC) stations to permit factoring of short term counts for estimates of annual average daily traffic (AADT). If there are insufficient CCSs for statistical accuracy in a factor group, use of statewide factors is encouraged. Hour of day, day of week, monthly (seasonal), axle correction, and annual adjustment factors are the only factors to be used as necessary to keep all AADTs current to the year for which they are being reported.

The HPMS traffic data needs should be conveyed to the traffic monitoring office within the State in a timely manner that allow enough time to develop and schedule the State’s comprehensive traffic monitoring program. Areas of the State selected for counting in a program year should be selected on a random basis. Highways with high variability should be counted more often than those with low variability, and highways with high traffic volume should be counted more extensively than those with low volume. To make the most of available resources, an area traffic count plan may consider using cluster count techniques whereby several counts are taken in the same general area. Counts scheduled and obtained under other programs may be incorporated into the count plan to avoid duplication of monitoring sites.

Sources of Data

Continuous Count Stations provide 24/7 monitoring of existing traffic conditions around the State. Travel on freeways, expressways and other multilane facilities can be monitored by route. Travel can also be monitored by area through statewide, or MPO freeway management or travel surveillance programs, which are often referred to as Intelligent Transportation System (ITS) deployments. Other highway functional systems, both State and off-State, can be monitored by geographic area, such as by county or highway district. Traffic information in a comprehensive count program should be compiled from all available sources -- State, MPO, ITS, city, and county.

Coordination and cooperation with local governments to implement a comprehensive count program is highly desirable; however, the State ultimately maintains responsibility for ensuring that these data meet minimum collection and quality requirements. To meet these responsibilities, the State should have a comprehensive quality assurance program that includes data collection, the conversion of traffic counts into current year AADT values, routine equipment testing provisions, and routine traffic count calibration procedures.

The following list of standards from ASTM International provides detailed guidance on traffic monitoring techniques and technologies:

E17.52 Traffic Monitoring Committee

• E 1318-09 Standard Specification for Highway Weigh-in-Motion (WIM) Systems with User Requirements and Test Methods
• E 1957-04 Standard Practice for Using Pneumatic Tubing for Roadway Traffic Counters and Classifiers
• E 2300-09 Standard Specification for Highway Traffic Monitoring Devices
• E 2415-05 Standard Practice for Installing Piezoelectric Highway Traffic Sensors
• E 2467-05 Standard Practice for Developing Axle Count Adjustment Factors
• E 2532-09 Standard Test Methods for Evaluating Performance of Highway Traffic Monitoring Devices
• E 2561-07a Standard Practice for the Installation of Inductive Loop Detectors
• E 2259-03a Standard Guide for Archiving and Retrieving ITS-Generated Data
• E 2665-08 Standard Specification for Archiving ITS-Generated Traffic Monitoring Data
• E2667-09 Standard Practice for Intersection Turning Movement Traffic Data
• E2759-10 Standard Practice for Highway Traffic Monitoring Truth-in-Data

AADT Calculation

The development of roadway section-based AADT estimates from traffic monitoring data using continuous or short term volume, vehicle classification, or truck weight data must include the use of adjustment factors if the data does not cover all months. The AADT estimates reported to the HPMS for all roadway sections not counted during the current year must be updated to current year AADT estimates by use of annual (growth) adjustment factors.

The rounding of AADTs is acceptable for HPMS following the scheme recommended by the AASHTO Guide but is not encouraged unless it is common practice for the State to round all traffic data in their traffic monitoring database and the practice is applied to all traffic data consistently. This applies to the reporting of volume and vehicle classification data. Rounding should be performed after all adjustments to the raw count have been made and should not be performed when calculating percent single unit and combination trucks. Low volume counts must not be rounded to report zero as a volume or as a percent since this will not accurately represent the presence of minimal volumes and will also show no change in trends. Zeros should only be reported when the actual count is zero.

Work performed in 2015 by Battelle Memorial Institute through a FHWA led Pool Funded research effort and reported in Assessing Roadway Traffic Count Duration and Frequency Impacts on Annual Average Daily Traffic Estimation (Krile, et. al.), FHWA-PL-16-008, has shown that there are two limitations with the traditional AASHTO method. One limitation is that the above equation uses only complete days of data. This means that the loss of one hour of data due to errors in the data collection process results in the loss of a full day of data from the AADT computation, reducing the accuracy of the resulting AADT estimate. The second limitation is that the averaging process used in the AASHTO method produces a small amount of bias in the resulting AADT estimate by slightly under-valuing both weekday traffic and traffic occurring in months with 31 days in comparison to months with fewer days.

As a result, FHWA is proposing an alternative modified formulation for computing AADT. This computation is performed in two steps. The first step computes monthly average daily traffic from the available hourly (or other temporal period) count records. The formula will work equally well with any temporal interval data, such as the 5-minute or 1-minute data frequently recorded by ITS-based traffic management systems. The second step then computes AADT from the twelve available monthly values. These two mathematical steps are as follows:

and

Where:

AADT = average annual daily traffic

MADT_m = monthly average daily traffic for month m

VOL_ihjm= total traffic volume for ith occurrence of the hth hour of day within jth day of week during the mth month

i = occurrence of a particular hour of day within a particular day of the week in a particular month (i=1, n_hjm) for which traffic volume is available

h = hour of the day (h=1,2, 24) – or other temporal interval

j = day of the week (j=1,2, 7)

m = month (m=1, 12)

n_hjm = the number of times the hth hour of day within the jth day of week during the mth month has available traffic volume (n_hjm ranges from 1 to 5 depending on hour of day, day of week, month, and data availability)

w_jm = the weighting for the number of times the jth day of week occurs during the mth month (either 4 of 5); the sum of the weights in the denominator is the number of calendar days in the month (i.e., 28, 29, 30, or 31)

d_m = the weighting for the number of days (i.e., 28, 29, 30, or 31) for the mth month in the particular year

While States and all in the transportation community can continue to use the traditional AASHTO method. FHWA is encouraging the adoption of the new formula which better reflects reality.

Volume Group Assignments

The State’s comprehensive traffic count program should be used to develop traffic volume group assignments for all roadway sections in a program that adequately monitors both high and low volume roads, including those off the State system. To facilitate this process, count station locations should be selected to represent expected AADT volume group breakpoints for the volume ranges of all required samples. This may require locating count stations at one per several miles in rural areas and more closely in urban areas. If there are homogeneous traffic sections as determined by prior counts or engineering judgment, more than one section may be represented by a single traffic count station as long as traffic does not vary more than 10%. Selection of count locations should be based on previous count experience on the section or adjacent sections, recent land use developments, and the existence of uncounted sections along the route.

Traffic Monitoring Program Elements

A detailed discussion of recommended procedures for developing, collecting, and processing travel monitoring data is contained in the Traffic Monitoring Guide (TMG). However, a general discussion of some elements of a typical traffic monitoring program and their applicability to the HPMS follows.

Count Cycle

A minimum of one-third of all NHS and Principal Arterial System (PAS) roadway sections (i.e., Interstates, Other Freeways and Expressways, and Other Principal Arterials) shall be counted each year; all other monitoring should be on a minimum six-year cycle. The roadway sections to be counted should be randomly selected from each sample stratum (volume group), with minor adjustments as necessary for strata with numbers of sections not divisible by three or having less than three samples. A single count may be used for several sections between adjacent interchanges on controlled access facilities.

Continuous Counts

Continuous Count Station (CCS) are used to provide 24/7 traffic count coverage for every day of the year at a limited number of locations using automated procedures. CCS data are also used to develop hourly, day-of-week, axle correction, monthly (seasonal), and annual adjustment factors which are then used to factor short term counts to an AADT. Analytical procedures to determine the appropriate level of effort and to develop the needed traffic estimates are described in the TMG.

Continuous count data are essential for converting short term counts to AADT. The State’s documentation of its continuous count program should identify the number of continuous counters on the rural and urban portions of the PAS/NHS system and the rest of the highway network. The process used to develop adjustment factors and their application should be thoroughly documented as well. Whenever possible, the State should have at least one continuous counter on each major PAS/NHS highway route. At a minimum, each continuous counter should have at least one full day of data for each day of the week for each month provided the State has an adequate automatic edit process based on the historic trend. If the new FHWA AADT method is utilized, at a minimum, each continuous counter should have at least one full time increment of data for each day of the week for each month of the year.

Short Term Counts

Short term counts cover lesser time periods than CCSs, 48-hour counts (two full 24-hour days) are required for all HPMS Full Extent and sample data including those off the State highway system except otherwise noted. Where axle correction factors are needed to adjust raw counts, they should be derived from facility-specific vehicle classification or weigh-in–motion (WIM) data obtained on the same route or on a similar route with similar traffic in the same area. Factors that purport to account for suspected machine error in high traffic volume situations shall not be applied to traffic counts used for HPMS purposes, including volume group assignment. In high volume situations and on controlled access facilities, it may be more appropriate to use continuous or short term ramp counts in conjunction with strategic mainline monitoring than to use short term counts on all mainline locations (see “ramp balancing” in the TMG for details).

Ramp Counts

Traffic counts are required on all Federal-aid highways including ramps associated with grade-separated interchanges. Ramp counts are important because many bottlenecks occur at major interchanges around the country and large amounts of Federal funds are expended to address these congestion issues. A minimum of one count every six years is required for ramps.

The same procedures used to develop AADTs on all HPMS roadway sections should be used to develop ramp AADTs. It is important that this volume data be an AADT for comparison to other AADTs and for reasonable trend analysis. States are encouraged to use adjustment factors developed based on either entrance or exit travel patterns or the functional class of the ramp, and to use this procedure consistently statewide. For example, the factors used for the mainline road with subordinate flow may be appropriate for use on the ramp. In other cases, the factors from intersecting roads connected to the ramp may be more appropriate for use. Good judgment and experience should be applied regarding factor use. As a minimum, 48-hour ramp counts should be adjusted with axle correction factors as needed.

Ramp counts should be available from freeway monitoring programs that continuously monitor travel on ramps and mainline facilities. Ramp balancing programs implemented by States on ramp locations and on high volume roadways could also be used to provide AADTs. In the case where no ramp counts are available, a State may use traffic matrix estimation. The State’s traffic modeling office may compute ramp traffic estimates as part of their modeling process.

Vehicle Classification

Data reported in the HPMS should reflect the use of continuous vehicle classification equipment to accurately report truck data, vehicle classification summaries, and develop monthly (seasonal), day of week and hour of day vehicle classification adjustment factors. Summary vehicle classification data reporting requirements are outlined in Chapter 3. Percent peak truck data (see Data Items 23 and 25), and truck AADT data (see Data Items 22 and 24) must be reported for each HPMS sample section as discussed in Chapter 4. Vehicle classification information must be reported in the summary travel data as discussed in Chapter 2.

The State’s vehicle classification program shall include:

• Data representative of all functional systems, both on and off the State system.
• Monitoring sessions for at least 48 hours to account for the changes in vehicle mix from day to day. Data for less than 24 continuous hours is not acceptable. At locations where vehicle class is not consistent throughout the week, such as on weekends, counts longer than 48 hours may be necessary to determine appropriate days to counts and to accurately represent average vehicle class data.
• Monitoring frequency shall be at a minimum, over a three-year cycle with one-third of the counts per year for all NHS and Principal Arterial System (PAS) roadway sections (i.e., Interstates, Other Freeways and Expressways, and Other Principal Arterials). All other locations should be monitored on no longer than a six-year cycle.
• Data reported in HPMS shall represent data for the reporting year. Prior year classification counts shall be adjusted with annual adjustment factors to represent current year data and to accurately develop percent trucks and truck travel trends.
• Monitoring activities should include all lanes in both directions.
• The Axle Class Algorithm used should be consistent among different equipment and vendors, and should be checked annually to ensure that it is working properly for all vehicle types.

Axle correction factors are to be developed based on data that represents all months (seasons) of the year. They should be applied to all counts that are based on axle sensors. The factor groups could be the same as for other adjustment factors or can be for each functional class and are to be updated each year based on that year’s vehicle classification data.

Vehicle classification programs shall be set up following the guidance in the TMG for monitoring homogenous roadway sections with one monitoring location. The limits of a homogenous traffic section for one vehicle class may differ from the limits of a different vehicle class.

Vehicle classification data used to report truck (both SU and CU) AADTs for HPMS shall be adjusted to represent average conditions for the entire year following the recommendations in the TMG. Adjustments to vehicle classification data should be based on factors developed using data from a permanent continuous vehicle classification (CVC) monitoring program established following the guidance in the TMG. As States fully develop and implement vehicle classification programs to provide sufficient and accurate data to develop adjustment factors, this should be an integral component of a comprehensive traffic monitoring program. States that do not have a complete program are still required to adjust raw count data using interim procedures they have developed. These interim procedures could focus on using data from traffic volume programs to develop adjustment factors if considered reasonable to represent truck travel patterns until more specific vehicle classification data becomes available.

The goal of developing a comprehensive vehicle classification program to provide truck AADTs based on truck characteristics is of utmost importance since various studies have concluded that truck travel oftentimes varies considerably from total traffic patterns and has different trends statewide and by functional class. All other vehicle types are also important and should receive the attention they deserve. Motorcycles in particular are a small percent of travel but have significant safety issues that require attention for estimating their travel exposure.

Vehicle Miles of Travel (VMT)

Estimates of Daily Vehicle Miles of Travel (DVMT) are developed by direct computation for all Federal-aid highways. Moreover, this information is generated via the HPMS software by taking the product of AADT and length (in miles) for all sections, and summing the section-specific results to the desired HPMS aggregation level (e.g., functional system, urban area, etc.). A comprehensive traffic monitoring program, good traffic volume procedures and practices, a well-distributed HPMS sample, and appropriate AADT estimation techniques will result in highly reliable DVMT estimates.

Specific HPMS requirements for reporting VMT can be found in Section 3.3 of this manual. Examples of good state practices for estimating VMT on non-Federal-aid highways are:

• Current traffic growth rate on collectors or higher systems;
• Limited sample of short term traffic counts;
• Combination of sample and estimated counts; and
• Area-wide average daily traffic based on documented methods.

One method which is not recommended is to use the residual of the statewide total VMT minus the Federal-aid highway system VMT because this obscures all other traffic data collected. Another discouraged method is the use of fixed percent of traffic growth (e.g. zero or one percent).

The monthly Traffic Volume Trends report is published by the FHWA based on a sample of traffic data from CCSs in the States. Annual VMT growth rates by functional system derived from these reports are used to validate HPMS traffic data. The goal is that all traffic information published by the FHWA and the States is valid and consistent.