Traffic Monitoring Guide Research Articles

Challenges in Monitoring Regional Trail Traffic

This study reports traffic monitoring results at 30 locations on a 972-mi shared-use trail network across the east-central United States. We illustrate challenges in adapting the principles in the Federal Highway Administration’s Traffic Monitoring Guide to a regional trail network. We make four contributions: 1) we use factor analysis and k-means clustering to implement a stratified random process for selecting monitoring sites; 2) we illustrate quality assurance procedures and the challenges of obtaining valid results from a multi-state monitoring system; 3) we describe variation in trail traffic volumes across five land use classes in response to daily weather and seasons; and 4) we report two performance measures for the network: annual average daily trail traffic and trail miles traveled. The Rails to Trails Conservancy deployed passive infrared traffic monitors in 2015 through 2017. Site-specific regression models were used to impute missing daily traffic volumes. The effects of weather were consistent across land use classes but the effects of temporal variables, including weekend and season of year, varied. A plan for short-duration monitoring is presented. Results confirm the FHWA monitoring principles and the difficulties of implementing them regionally.

Up on The 606

The 606 is the world’s first multiuse elevated trail, extending for 2.7 mi (4.35 km) through diverse neighborhoods whose per capita of open space is one of the lowest in Chicago. The trail connects six ground-level parks and is managed for recreation, but it also serves as a cross-town transportation connector and was funded partially with transportation dollars. Managers sought information about trail use to maintain a safe and harmonious experience for users, to plan operations and maintenance, and to document the benefits of trail development. The use of The 606 was examined during the first 6 months of 2016, and on the basis of those results, its use for the entire year was projected. Automated traffic monitoring with active infrared counters followed procedures in the FHWA Traffic Monitoring Guide. Screenline calibration tests revealed relatively high rates of occlusion owing to user type and traffic volume, yielding an adjustment factor of 1.239. Most users were pedestrians, but proportions varied by day of the week and time of day. Average daily traffic volumes between January 1 and June 30 at counters near the east and west ends of the trail were 3,500 and 3,000, respectively, with peak daily traffic exceeding 10,000. A regression model using weekdays and weekends, location on the trail, and temperature variables explained 80% of the daily use variation. Model extrapolation with historical weather averages estimated annual traffic volumes at 1.46 million and 1.3 million for the two sites, and a combined total annual miles traveled of 3.7 million (5.95 million km). Management implications and future research directions are highlighted.

Monitoring and Modeling of Urban Trail Traffic: Validation of Direct Demand Models in Minneapolis, Minnesota, and Columbus, Ohio

Transportation planners and engineers need estimates of nonmotorized traffic volumes and analytical tools to plan and manage infrastructure for bicycling and walking. Direct demand models are useful, comparatively simple tools for the estimation of volumes from nonmotorized traffic counts and do not require detailed information from travel behavior inventories. However, few demand models for bicycling and walking have been validated. This paper extends the practice of nonmotorized traffic monitoring and modeling in three ways. First, procedures recommended in the FHWA Traffic Monitoring Guide are followed to present estimates of annual average daily traffic (AADT) for each segment of the urban trail networks in two major U.S. cities: Minneapolis, Minnesota, and Columbus, Ohio. Second, independent variables constructed from nationally available data sets and the local characteristics of each trail system are used to estimate and validate direct demand models for AADT. Third, to assess the potential for the general application of the models, the results of cross-city validations are presented. Our results confirm that FHWA monitoring procedures can be used to characterize the variation in traffic flows on urban trail networks. Direct demand models for each city have reasonably good fits, but the predicted traffic volumes for more than one-third of the segments exceed the actual volumes by more than 60%. Cross validation results indicate that these models cannot yet be applied as predictive tools across cities. More experimentation is needed to assess the feasibility of developing generalized direct demand models for trails and other nonmotorized transportation networks.

Determining the optimal number of seasonal adjustment factor groupings when estimating annual average daily traffic and investigating their characteristics

Although cluster analysis is recommended by the US Traffic Monitoring Guide (TMG) to supplement the development of seasonal adjustment factor groupings (SAFGs), the relationships among SAFGs' characteristics remain undiscovered, while the determination of the optimal number of clusters is an ambiguous task exposed to great subjectivity. Statistical indicators provide a mathematical solution by removing engineering judgment without taking into consideration any guidelines or other criteria, necessary for transportation planners to generate ‘practical and sensible’ groupings. The method examined in this study aims to overcome the above weaknesses incorporating into the methodology a series of statistics, recommendations, and previous research findings. The investigation of the relationships among (1) the within-group variation, (2) the total number of sites, (3) the minimum number of stations within a cluster, (4) the optimal number of clusters, and (5) the geographical size of the groups constitutes the main objectives of this research. According to the results, the cluster variability declines as the available number of stations increases. When the minimum number of stations within a cluster increases, the weighted coefficient of variation inflates as well, with the rate of increase depending on sample size. The average number of automatic traffic recorders per cluster is analogous to the sample size, while the optimal number of clusters varies conversely with the minimum number of stations within a cluster. The application developed for the conduct of the analysis minimizes the computational time needed, while it can be easily implemented by engineers to automate the process recommended by the TMG, enhancing the current state of practice.

Application of Seasonal Adjustment Factors to Subsequent Year Data

Seasonal adjustment factors are used by state highway agencies to convert short-term traffic counts into estimates of annual average daily traffic (AADT). These factors are typically calculated from continuous traffic counter data. The protocol for the process is described in FHWA's Traffic Monitoring Guide. This protocol specifies that temporary counts collected during a year should be adjusted by seasonal factors computed from continuous data collected during the same calendar year. The authors had suggested that, if the prior year seasonal factors could be used, AADT estimates would be available soon after temporary counts were collected. This study investigated whether the seasonal factors from sequential years could be applied without loss of statistical accuracy. No significant loss of accuracy was found in estimating AADT when prior-year seasonal adjustment factors were applied to current-year short-term counts of traffic volume. Likewise, no significant loss of accuracy was found when vehicle miles of travel were computed by using prior-year seasonal adjustment factors.

WIM data quality and its influence on predicted pavement performance

In this study, quality of weigh-in-motion (WIM) traffic data is checked based on a set of algorithms implemented in a program developed in visual basic application (VBA). As a first step, several algorithms are written to check whether the data are within the acceptable range prescribed by the traffic monitoring guide (TMG). Algorithms are also written to checks whether the number of axles is consistent with the vehicle class and the spacing of axles. Next, the VBA program determines the frequencies of vehicle class distribution, directional distribution, lane distribution, monthly distribution, hourly distribution, gross vehicle weight distribution and steering axle weight distribution. To test the quality of data, gross weight and steering axle weight distributions are tested by a set of rules recommended by TMG. It is evident from this study that bending plates provide high quality WIM data compared to piezoelectric WIM systems. It is possible that piezoelectric sensors are affected by temperature and surface condition. To this end, the effect of WIM data quality is evaluated by determining the influence of axle load spectra on the pavement performance predicted by the mechanistic empirical pavement design guide (MEPDG). Using the data collected by three bending plate WIM sites, a subroutine is written to produce a positive and negative calibration bias in the WIM data. Next, single, tandem, tridem and quad axle load spectra are developed for 0, 10, 20 and 30% positive and negative weight measurement bias by using an algorithm based on the axle spacing. From the results, it is observed that the influence of weight measurement bias on predicted longitudinal and alligator cracking is critical and can lead to overestimation or underestimation of the pavement thickness. The effect of weight data bias on predicted rutting is moderate. However, weight measurement error does not affect predicted transverse cracking and international roughness index (IRI).

Validation of TMG Traffic Data Check Algorithms

Several state DOTs utilize the statistics presented in the Federal Highway Administration (FHWA)’s version of the traffic monitoring guide (TMG) to conduct weigh-in-motion (WIM) data quality checks for DARWin-ME traffic load spectra data preparation. Because the data check statistics in the guide were originally developed in the early 1990s, there is a need to reexamine its continuing use in up-to-date truck traffic streams. This paper aims to validate the TMG data check algorithms using WIM data from the long-term pavement performance (LTPP) special pavement study (SPS) traffic pooled-fund study. Cluster analysis is performed to identify similar traffic patterns, and the statistical summary for each data check criterion is obtained. Subsequently, the statistics based on the pooled fund data are compared to those recommended in the TMG. The results of the paper conclude that the data check criteria in the TMG guide are still valid for most of the sites; however, users should be alert that the guide also acknowledges WIM stations with exceptional patterns. This study presents a framework to check the suitableness of using the guide for such cases.

Sensitivity of NCHRP 1-37A Pavement Design to Traffic Input

This study deals with the sensitivity of the NCHRP 1-37A Pavement Design Guide predictions to traffic data input. A number of traffic data collection scenarios are simulated with the use of extended coverage (more than 299 days per year) weigh-in-motion (WIM) data from the Long-Term Pavement Performance (LTPP) database. These scenarios consist of combinations of site-specific, regional, and national data, including total truck counts, truck counts by class, and axle-load distributions by axle type. For each simulated scenario, traffic input to the NCHRP 1-37 Pavement Design Guide is estimated with the methodologies prescribed in the Traffic Monitoring Guide. For discontinuous time coverage scenarios, all possible time coverage combinations are considered, and the range in the estimated traffic input parameters is computed. Pavement life predictions are obtained under mean traffic input for all traffic data collection scenarios and under low percentile input for the discontinuous traffic data collection sc...

Sensitivity of NCHRP 1-37A Pavement Design to Traffic Input

This study deals with the sensitivity of the NCHRP 1-37A Pavement Design Guide predictions to traffic data input. A number of traffic data collection scenarios are simulated with the use of extended coverage (more than 299 days per year) weigh-in-motion (WIM) data from the Long-Term Pavement Performance (LTPP) database. These scenarios consist of combinations of site-specific, regional, and national data, including total truck counts, truck counts by class, and axle-load distributions by axle type. For each simulated scenario, traffic input to the NCHRP 1-37 Pavement Design Guide is estimated with the methodologies prescribed in the Traffic Monitoring Guide. For discontinuous time coverage scenarios, all possible time coverage combinations are considered, and the range in the estimated traffic input parameters is computed. Pavement life predictions are obtained under mean traffic input for all traffic data collection scenarios and under low percentile input for the discontinuous traffic data collection scenarios. For each scenario and confidence level, pavement life errors are predicted for the life estimated by the complete extended coverage WIM data set. Two sources of error are identified: one from specifying the mean value for all traffic input and one from specifying the lowest percentile for all traffic input simultaneously (the latter applies to discontinuous coverage scenarios only). Overall error is computed by the addition of the range in the mean of the second error component to the range of the first error component. The results are in the form of plots of pavement life prediction error, versus confidence level, versus traffic data collection scenario.