Sarah Hernandez

Wavelet-k Nearest Neighbor Vehicle Classification Approach with Inductive Loop Signatures

In this study, a new vehicle classification algorithm was developed with inductive loop signature technology. There were two steps to the proposed algorithm. The first step was to use the Haar wavelet to transform and reconstruct inductive vehicle signatures, and the second step was to group vehicles into FHWA vehicle types through the use of the k nearest neighbor (KNN) approach with a Euclidean distance classifier. To determine the proper proportion of the wavelet to apply for reconstruction and feature extraction, transformed signatures were examined with percentages of large components of their corresponding wavelets. To implement the KNN approach, a library of vehicle signature templates for each FHWA vehicle class was composed. The proposed vehicle classification algorithm demonstrated promising classification results, with a 92.4% overall accuracy. The algorithm can be applied to the real world without the concerns about recalibration and transferability that arise with the use of signature data from single loops. Two additional vehicle classification schemes were applied for performance evaluation. For the inductive signature performance evaluation classification scheme, which aimed to facilitate emission analysis and easy interpretation, the overall accuracy was 94.1%. For the axle-based vehicle classification scheme proposed in this project, which aimed to group vehicles by use and the number of axles, the overall accuracy was 93.8%. Future research will focus on refinement of the signature template library for each FHWA vehicle type to further improve the performance of the proposed vehicle classification algorithm. The selection of the value of k for the KNN approach will be investigated also.

Multi-Criteria Sustainability Assessment in Transport Planning for Recreational Travel

ABSTRACT A transport planning framework is considered that incorporates a multi-criteria, composite sustainability index (CSI) with elastic decision-maker preferences, and applied to a case study of an outdoor recreational destination. A stated preference survey is conducted on transit alternatives to access the United States. Mojave National Preserve from Barstow, California, located 160 kilometers away. A binary logit model is developed to relate policy variables to sustainability dimensions. A revised CSI is applied to evaluate eight alternatives under three decision-making schemes. Findings suggest that a zero-emissions train service with two round trips per day is preferred over the other alternatives under all three schemes.

Truck Body Configuration Volume and Weight Distribution: Estimation by Using Weigh-in-Motion Data

Weigh-in-motion (WIM) systems measure truck volumes, assist in pavement design and management, and enforce truck size and weight regulations. Although WIM systems provide truck classification based on the FHWA axle configuration classification scheme, more specific vehicle characteristics such as body configuration are necessary for freight planning and pollution monitoring. A modified decision tree model was developed to estimate truck volumes and gross vehicle weight (GVW) distributions by body configuration for five-axle semi-tractor trailers (3S2) with the use of existing WIM system measurements such as axle spacing and vehicle length. This method allows more information to be extracted from axle-based measurement data to leverage the significant investments in existing WIM systems better. Data for model development were collected at three WIM sites spanning rural and urban locations in California and described more than 7,500 3S2 trucks stratified into five trailer body categories: vans, tanks, platforms, 40-ft intermodal containers, and other. Model estimates of trailer body configuration volumes differ by only 8% from actual volumes when averaged across all body configurations on an independent test data set. A normalization procedure was designed to improve the models robustness against systematic and random calibration inaccuracies at WIM sites. An algorithm based on Gaussian mixture models was developed to estimate GVW distribution by body configuration. Results show that estimated GVW distributions statistically capture the actual GVW distribution of each body configuration and are temporally and spatially transferable.

Using Data from a State Travel Demand Model to Develop a Multi-Criteria Framework for Transload Facility Location Planning

The majority of freight is transported within the U.S. by road. However, the use of alternative modes, such as rail and barge, is associated with lower transportation and infrastructure maintenance costs, release of highway capacity, increased safety, and lower emissions. Thus, there is a latent opportunity for shippers and consumers to benefit from modal shift. In this context, strategically located freight-transfer facilities to improve rail and barge access is key. Moreover, for states with lower commodity tonnages and access to short-line rail and navigable waterways, transload facilities have significant potential to shift freight to underutilized modes. This paper develops a multi-criteria assessment framework to identify strategic locations for transload facilities at the state level. Using a statewide travel demand model (STDM) as the main data source, this framework provides a sketch-planning tool to support decision-making for state Departments of Transportation and economic development agencies. The multi-criteria quantify four measures of facility potential: (a) interaction with the transportation network, (b) amount of freight transported between major freight routes, (c) spatial aggregation, and (d) directionality aggregation. Each criterion is estimated and combined at the county level to produce a multi-criteria score, which defines a county’s potential to support transload movements. Using this score, counties are ranked, and facilities prioritized. The framework is applied to Arkansas and validated using the STDM for base (2010) and forecast (2040) years.

Development of a Cost Estimation Framework for Potential Transload Facilities

As the cost of transportation continues to rise and there is a growing push for a more environmentally friendly transportation network, optimizing mode distributions becomes an attractive solution. One way to optimize mode distributions is through strategically located freight transfer facilities, such as transload facilities. While there are many benefits to this type of facility, such as the emissions savings or pavement damage reductions garnered by shifting commodity tonnage onto alternative modes of transport, it is also essential that transportation planners understand the costs associated with building transload facilities. Unfortunately, literature does not provide an adequately disaggregated and scalable cost estimation approach that could be applied to various configurations and types of transload facilities. In this research, a cost estimation framework was created to determine the basic cost of transload facilities by type using unit costs from a construction cost database, equipment costs from local dealers, the projected commodity tonnage, design recommendations from literature, and survey responses from local facilities. A case study based upon proposed facilities in Arkansas was completed to illustrate the effectiveness of this methodology. While there is currently no construction design for these facilities, this framework yielded costs consistent with those expected. A key finding was that storage costs could account for up to 81% of a transload facility’s costs. Overall, this cost framework is believed to balance general scalability with accuracy well to provide reasonable cost estimations for constructing new or expanded facilities.

Cellular Models: HD Patient-Derived Pluripotent Stem Cells

Huntingtons disease (HD) is an autosomal dominant neurodegenerative disorder caused by expanded polyglutamine (polyQ)-encoding repeats in the Huntingtin (HTT) gene. Traditionally, HD cellular models consisted of either patient cells not affected by disease or rodent neurons expressing expanded polyQ repeats in HTT. As these models can be limited in their disease manifestation or proper genetic context, respectively, human HD pluripotent stem cells (PSCs) are currently under investigation as a way to model disease in patient-derived neurons and other neural cell types. This chapter reviews embryonic stem cell (ESC) and induced pluripotent stem cell (iPSC) models of disease, including published differentiation paradigms for neurons and their associated phenotypes, as well as current challenges to the field such as validation of the PSCs and PSC-derived cells. Highlighted are potential future technical advances to HD PSC modeling, including transdifferentiation, complex in vitro multiorgan/system reconstruction, and personalized medicine. Using a human HD patient model of the central nervous system, hopefully one day researchers can tease out the consequences of mutant HTT (mHTT) expression on specific cell types within the brain in order to identify and test novel therapies for disease.

Truck Activity Monitoring System for Freight Transportation Analysis

Understanding truck activity is an essential component of strategic freight planning and programming. However, recent studies have revealed a significant void in the availability of detailed truck activity data. Although some existing detectors are capable of providing truck counts by axle configuration, higher-resolution data that indicate truck body configuration, industry served, and commodity carried cannot be obtained from existing sensors. This paper presents the newly developed Truck Activity Monitoring System, which leverages existing in-pavement traffic sensors to provide truck activity data in California. Existing inductive loop detector sites were updated with inductive signature technology and advanced truck classification models were implemented to provide detailed truck count data with more than 40 truck body configurations. The system has been deployed to more than 90 detector locations in California to provide coverage at state borders, regional cordons, and significant metropolitan truck corridors. An interactive geographic information system website provides users with advanced visual analytics and access to archived data across all deployed locations. The case studies presented in this paper demonstrate the potential of the data obtained from this system in analyzing and understanding current and historical industry-specific truck activity.

Estimation of Average Payloads from Weigh-in-Motion Data

Average payloads define the ton-to-truck conversion factors that are critical inputs to commodity-based freight forecasting models. However, average payloads are derived primarily from outdated, unrepresentative truck surveys. With increasing technological and methodological means of concurrently measuring truck configurations, commodity types, and weights, there are now viable alternatives to truck surveys. In this paper, a method to derive average payloads by truck body type and using weight data from weigh-in-motion (WIM) sensors is presented. Average payloads by truck body type are found by subtracting an estimated average empty weight from an estimated average loaded weight. Empty and loaded weights are derived from a Gaussian mixture model fit to a gross vehicle weight distribution. An analysis of truck body type distributions, loaded weights, empty weights, and resulting payloads of five-axle tractor trailer (FHWA Class 9 or 3-S2) trucks is presented to compare national and state-level Vehicle Inventory and Use Survey (VIUS) data and the WIM-based approach. Results show statistically significant differences between the three data sets in each of the comparison categories. A challenge in this analysis is the definition of a correct set of payloads because the WIM and survey data are subject to their own inherent misrepresentations. WIM data, however, provide a continuous source of measured weight data that overcome the drawback of using out-of-date surveys. Overall, average payloads from measured weights are lower than those for the national or California VIUS estimates.