2016 Transportation Research Board Annual Meeting

8:00 AM – 9:45 AM

Program:  Pedestrian and Bicycle University Education Joint Subcommittee of ANF20, ANF10

Daniel Hess, University at Buffalo, presiding

Lecture Program: Built for Success: What It Takes to Run a Modern Transportation Research Program

Evidence-Based Management of R&D Projects Intending Market Deployment

Joseph Lane, University at Buffalo

No abstract available

Lecture Program: Transportation Network Logistics

Implications of Cost Equity Consideration in Hazmat Network Design

Longsheng Sun, University at Buffalo

Mark Karwan, University at Buffalo

The hazmat network design problem (HNDP) aims to reduce the risk of transporting hazmat in the network by enforcing regulation policies. The goal of reducing risk can increase cost for different hazmat carriers. Since HNDP involves multiple parties, it is essential to take the cost increase of all carriers into consideration for the implementation of the regulation policy. While we can consider cost by placing upper bounds on the total increase, the actual cost increase for various OD pairs can differ, which results in unfairness among carriers. Thus we propose to consider cost equity issue as well in HNDP. Additionally, due to the existence of multiple solutions in current HNDP models and the possibility of unnecessarily closing road segments, we introduce a new objective considering the length of all the closed links. Our computational experience is based on a real network and we show results under different cost consideration cases.

10:15 AM – 12:00 PM

Lecture Program: FRP Composite Field Applications, Part 1: Existing Structures (Part 2, Session 840)

Advances in FRP Composites in Transportation Infrastructure

Jerome O'Connor, University at Buffalo

Wayne Frankhauser, Jr., Maine Department of Transportation

This document describes a domestic scan that was conducted in 2015 under NCHRP 20-68A U.S. Domestic Scan 13-03 “Advances in FRP Composites in Transportation Infrastructure”. An eight-person scan team visited states that have been proactive in the use of fiber-reinforced polymers (FRP) for bridges and other structures. Bridge engineers from these states shared their experiences, describing applications that worked well, but also the challenges of trying a technology without the benefit of national standards.

The team identified successful applications for bridges and other structures and collected project summaries, specifications, plans, etc. that can be shared with others. Some cost effective applications of these advanced materials were identified, such as for the rapid repair of concrete bridge girder that have been hit by an overhead truck or anytime a bridge needs immediate improvement and it is not feasible to close or load-restrict it while waiting for a long term solution. The service life of heritage structures can often be extended with FRP improvements and done so without altering its appurtenance significantly.

Despite numerous successful projects, the team identified several areas where improvement was needed. An inventory of FRP projects does not exist and a history of FRP work on particular bridges is not being satisfactorily documented and placed in bridge history files. In general, training has not been provided to bridge inspectors and maintenance staff who are responsible for taking care of bridges and sometimes they do not even know why a treatment has been applied. AASHTO guidelines for design, construction and maintenance have not been developed for most FRP applications.

Although there have been research and demonstration projects since the late 1980’s, a national initiative may be needed before the use of composite materials for retrofitting existing bridges and making new bridge components is commonplace. Currently, individual state efforts have resulted in redundant effort and in some cases repeated mistakes so a coordinated method of information sharing would be very beneficial to bridge owners. While the scan did not produce a strategic plan, the information collected will be helpful for leaders who are interested in producing a roadmap for further research and creation of standards for industry to follow.

This is a preliminary summary report, subject to further study and evaluation of the data collected.

Poster Program: Information Technology Applications for Travel Monitoring and Connected Vehicles (14)

Data Warehouse Development and Real-Time Incident Detection

Andrew Bartlett, University at Buffalo

Adel Sadek, University at Buffalo

In the traffic engineering field, study and analysis often requires the use of multiple datasets. The nature of these data often makes them difficult to work with, especially in conjunction with one another. The overall goal of this study was to not only design a solution to this problem for the Buffalo-Niagara Region of western New York, but to demonstrate its usefulness through a specific application. To achieve this, three objectives were designed: (1) outline the structure of a data warehouse for the Buffalo-Niagara region, (2) use the combined data in the prototype warehouse to examine its usefulness in the construction of a real-time incident detection system which not only detects incidents but also tries to predict incident characteristics, and (3) show the importance of the data warehouse by comparing the results of incident detection strategies which require different combinations of data. To meet these objectives a prototype data warehouse was first created, and then used in the creation and validation of three incident detection strategies: a speed threshold detection system, a binary probit model which uses only speed data, and a binary probit model which uses a combination of speed and volume data. The prototype data warehouse showed it was possible to construct a fully fleshed-out version for transportation data in the Buffalo-Niagara region with useful results. The speed threshold model which used a 10 minute speed drop of 10 mph to detect incidents had a 62.5% detection rate, as well as favorable false alarm and classification rates. The more complex binary outcome model which used only speed data detected incidents with a success rate of 70.4%, an improvement over the speed threshold model despite worse false alarm and classification rates. It was also able to predict incident type, number of blocked lanes, and incident severity with 75.9%, 70.4%, and 75.9% accuracy, respectively. The binary outcome model which used both speed and volume data had a more impressive detection rate of 75.5% with similar false alarm and classification rates and was slightly better at predicting incident type and severity (both with 77.6% accuracy) but slightly worse at predicting the number of blocked lanes (with 69.4% accuracy). Overall, the combined data model is the best strategy for both detecting incidents and predicting their characteristics, which emphasizes the importance of a transportation data warehouse.

Poster Program: Research Trends in Evacuation Transportation Modeling and Analysis

Statistical Analysis of Dynamics of Household Hurricane-Evacuation Decisions

Md Tawfiq Sarwar, University at Buffalo

Panagiotis Anastasopoulos, University at Buffalo

Satish Ukkusuri, Purdue University

Pamela Murray-Tuite, Virginia Polytechnic Institute and State University

Fred Mannering, Purdue University

With the increasing number of hurricanes in the last decade, efficient and timely evacuation remains a significant concern. Households’ decisions to evacuate/stay and selection of departure time are complex phenomena. This study identifies the different factors that influence the decision making process, and if a household decides to evacuate, what affects the timing of the execution of that decision. While developing a random parameters binary logit model of the evacuate/stay decision, several factors, such as, socio-economic characteristics, actions by authority, and geographic location, have been considered along with the dynamic nature of the hurricane itself. In addition, taking the landfall as a base, how the evacuation timing varies, considering both the time-of-day and hours before landfall, has been analyzed rigorously. Influential factors in the joint model include the relative time until the hurricane’s landfall, height of the coastal flooding, and approaching speed of the hurricane; household’s geographic location (state); having more than one child in the household, vehicle ownership, and level of education; and type of evacuation notice received (voluntary or mandatory). Two time intervals from 30 to 42 hours and 42 to 66 hours before landfall resulted in random parameters, reflecting mixed effects on the likelihood to evacuate/stay. Possible sources of the unobserved heterogeneity captured by the random parameters include the respondents’ risk perception or other unobserved physiological and psychological factors associated with how respondents comprehend a hurricane threat. Thus the model serves the purpose of estimating evacuation decision and timing simultaneously using the data of Hurricane Ivan.

Poster Program: Freeway Traffic Control (18)

Identifying On-Site Traffic Accidents using Both Traffic and Social Media Data

Zhenhua Zhang, University at Buffalo

Ming Ni, University at Buffalo

Qing He, University at Buffalo

Jing Gao; JIZHAN GOU; Xiaoling Li, Virginia Department of Transportation

Social media receives increasing attentions as crowdsourced information for traffic operations and management. One recent trending study is to use social media to detect on-site traffic accidents. However, it remains unknown how effective the social media based detection methods is as compared with traditional loop detector based method. In this paper, we first develop two prevailing accident detection models with traffic and tweet data, respectively. Then, we fuse such two models to jointly detect traffic accidents. The results show a promising improvement in both accuracy and precision. This study provides insights in utilizing social media data to assist accurate on-site traffic accident detection. 

2:00 PM – 3:45 PM

Poster Program: New Directions in Travel Surveys: Big Data, Smartphones, and Stated Preference (13)

Fast Online Travel Mode Identification using Smartphone Sensors

Authors: Xing Su, City University of New York (CUNY)

Hernan Caceres Venegas, University at Buffalo

Hanghang Tong, University at Buffalo

Qing He, University at Buffalo

Personal trips in modern urban society usually involve multiple travel modes. Recognizing a traveler's transportation mode is not only critical to applications like personal context-awareness, but also essential to urban traffic operations, transportation planning, and facility design. While most of recent practice often leverages infrastructure-based fixed sensors or GPS for traffic mode recognition, the emergence of smartphone provides an alternative promising way with its ever-growing computing, networking and sensing powers. 

In this paper, we propose a GPS and network-free method to detect a traveler’s travel mode using mobile phone sensors. Our application is based on the latest Android smartphones with multimodality sensors. By developing a hierarchical classification method with online learning model, we achieve almost 100% accuracy in the binary classification wheelers/non-wheelers travel mode, and an average of  97.1% accuracy with all of the six travel modes (buses, subways, cars, bicycling, walking, and jogging). Our system (a) performs significantly faster in computation and could adapt to each traveler's pattern by using the online learning model (b) works with a low sampling rate for sensing so that it saves the smartphone battery.

3:45 PM – 5:30 PM

Lecture Program: Innovations in Statistical Methods for Transportation Researchers and Practitioners

Panagiotis Anastasopoulos, University at Buffalo, presiding

Lecture Program: New Research on Alcohol and Driving

Drinking and Driving Behavior at Stop Signs

Jingyan Wan, University at Buffalo

Changxu Wu, University at Buffalo

YIQI ZHANG, University at Buffalo

Rebecca Houston, University at Buffalo

Chang Wen Chen, University at Buffalo

Panya Chanawangsa, University at Buffalo

Alcohol is one of the principal risk factors for motor vehicle crashes. One factor that contributes to vehicle crashes at intersections is noncompliance with stop signs. The present experiment investigated the effects of alcohol, drinking pattern and gender on driving behavior at stop signs. 30 subjects participated drinking and driving sessions during which they received a moderate dose of alcohol (0.08% BAC) or a placebo. A simulated driving tasks measured participants’ driving performance at stop signs in response to each dose. Results showed that alcohol led to impaired stop sign compliance, worse lateral and speed maintenance, and lower mean deceleration towards stop signs. More abrupt acceleration from stop signs was also observed among non-binge drinkers. Intoxicated female drivers, especially intoxicated female non-binge drinkers, exhibited longer duration of complete stop, greater stop accuracy, and more gradual and stable decelerating than other groups. In addition, under alcohol, male binge drinkers’ standard deviation of deceleration was lower compared with male non-binge drinkers. However, under placebo, binge drinkers’ complete stop duration was shorter than non-binge drinkers. Results indicated that alcohol increased willingness to take risks, impaired driving precision, and increased reaction time. It was also suggest that, intoxicated female drivers, especially intoxicated female non-binge drinkers, were more cautious than other groups, indicating that they were aware of their impairment and adjusted their behavior appropriately. Binge drinking was found increase males’ tolerance and, however, increased impulsivity of all drivers.

4:15 PM – 6:00 PM

Poster Program: Capacity and Level-of-Service Issues for Freeways and Other Facility Types (19)

An Exploratory Study on the Correlation between Twitter Concentration and Traffic Surge

Zhenhua Zhang, University at Buffalo

Ming Ni, University at Buffalo

Qing He, University at Buffalo

Jing Gao,  University at Buffalo

JIZHAN GOU; Xiaoling Li, Virginia Department of Transportation

Social media receives increasing attentions as a crowdsourced information source in traffic operations and management. The tweets, that are blogged and shared by the broad masses of people, may be associated with some major social activities. These tweets are called “Twitter concentrations” in this paper. The public activities behind Twitter concentrations potentially pose more pressure on traffic network and cause traffic surge within a specified time and location.  However, it still remains unknown how closely the Twitter concentration and traffic surge are correlated with each other. Our study fuses a set of tweets and traffic data collected during the whole year of 2014 in North Virginia Region, and mainly investigates the correlation between Twitter concentration and traffic surge in July. The results show the promise and effectiveness of our proposed methods and even provide insights in the causality of non-recurrent traffic surge.

Poster Program: Innovations in Rail Transit System Modeling, Optimization, and Simulation

Nonrecurrent Subway Passenger Flow Prediction from Social Media Under Event Occurrences

Ming Ni, University at Buffalo

Qing He, University at Buffalo

Jing Gao, University at Buffalo

Subway passenger flow prediction is strategically important in metro transit system management. The prediction under event occurrences turns into a very challenging task. In this paper, we adopt a new kind of data source -- social media to tackle this challenge. We develop a systematic approach to examine social media activities and identify event occurrences. Our results on real-world data demonstrate that there exists a strong positive correlation between passenger flow and the rates of social media posts. This finding motivates us to develop a novel approach for improved flow forecast. Specifically, we propose a parametric and convex optimization based approach, called Optimization and Prediction with hybrid Loss function (OPL), to fuse the linear regression and the results of seasonal autoregressive integrated moving average (SARIMA) model in the objective function jointly. The OPL hybrid model takes advantage of unique strengths of linear correlation in social media features and SARIMA model in time series prediction. Experiments on data related to a subway station show that OPL has the best forecasting performance compared with the state-of-the-art techniques. In addition, an ensemble model is developed to leverage the weighted results from OPL and support vector machine regression (SVR) together. As a result, the prediction accuracy and robustness further increases.

8:30 AM – 10:15 AM

Poster Program: Advances in Adaptive Signal Control and Optimization (19)

Qing He, University at Buffalo, presiding

10:15 AM – 12:00 PM

Lecture Program: Innovation in Passenger Rail Equipment and System Integration

Identification of Railcar Asymmetric Wheel Wear with Extreme Value Theory

Yu Cui, University at Buffalo

Qing He, University at Buffalo

Zhenhua Zhang, University at Buffalo

Zhiguo Li, IBM, Thomas J. Watson Research Center

Railcar asymmetric wheel wear results into severe wear on one wheel but little or no wear on the other wheel. The consequences of asymmetric wheel can be accelerated wear, mechanical failure and downtime and incurs high financial penalties. Therefore, identifying the asymmetric wheel wear is critical not only for cost effective maintenance, but also for safe operations.

Fortunately, increasing amount of various wayside detectors are instrumented along the railway which can monitor the health of railcar components and log large numbers of detailed information of railroad and. One can use this information to identify the asymmetric wheel wear as soon as possible. However, most elliptically contoured distributions are good at describing normal event but not good at dealing with the outliers which mainly locates in the tails of the distribution. Asymmetric wheel wear requires effective anomaly detection which mainly focused on the extreme values in the tail of a right-skewed distribution. In this paper, we employ the extreme value theory (EVT), which handles the unusually high or low data of distributions, to model the asymmetric wheel wear and derive an extreme value score to identify asymmetric wheel wear. Experiment results show that identification of asymmetric wheel wear can generate huge monetary benefit in terms of reducing yearly maintenance times of railcars.

2:00 PM – 3:45 PM

Poster Program: Urban Freight Innovations

Qian Wang, University at Buffalo, presiding

4:15 PM – 6:00 PM

Poster Program: Surface Transportation Weather and Its Effects on Traffic Networks

Measuring Freeway Route Travel Time Distributions Under Inclement Weather

Hernan Caceres Venegas, University at Buffalo

Ha Hwang, University at Buffalo

Qing He, University at Buffalo

This paper develops an efficient probabilistic model for estimating route travel time distributions, incorporating a variety of weather conditions with different potential traffic incident occurrence rates. Estimating the route travel time distribution from historical link travel time data is challenging due to the interactions among upstream and downstream links. Upon creating conditional probability function for each link travel time, we applied Monte Carlo Simulation to estimate the total travel time from origin to destination. A numerical example of three alternative routes in the City of Buffalo shows several implications. This study found that weather conditions, except for snow, incur minor impact on Off-Peak and Weekend travel time, whereas Peak travel time suffers great variations under different weather conditions. On top of that, inclement weather exacerbates route travel time reliability, even when mean travel time increases moderately. The computation time of the proposed model is linearly correlated to the number of links in a route. Therefore, this model can be used to obtain all the Origin-Destination travel time distributions in an urban region.

10:15 AM – 12:00 PM

Lecture Program: Pavement Performance Modeling and Measurement

Three-Stage Least-Squares Analysis of Post-Rehabilitation Pavement Performance

Md Tawfiq Sarwar, University at Buffalo

Panagiotis Anastasopoulos, University at Buffalo

Recent studies have provided improvements in the forecasting accuracy of pavement performance modeling, by statistically modeling pavement performance indicators as a system of seemingly unrelated regression equations (SURE). This approach accounts for cross-equation error correlation as a means to control for unobserved factors that lead pavements in poor condition to observe poor performance indicators. In the state of Indiana, the most common pavement performance indicators are the international roughness index (IRI), the rutting depth, and the pavement condition rating (PCR). Even though the first two can be accurately measured, the PCR is based on Engineers’ observations of the pavement surface. Therefore, it is possible that the PCR may be measured as a function of the observable IRI and rutting depth. This paper, explores this possibility by estimating a three-stage least squares (3SLS) model of IRI, rutting depth, and PCR, using data collected between 1999 and 2011 in Indiana. All three pavement performance indicators are found to be affected by traffic characteristics, prior pavement condition, treatment and surface characteristics, drainage performance, and weather conditions. In addition, the PCR is found to also be significantly affected by the IRI and rutting depth measurements. The results of the 3SLS and SURE models are counter imposed, with the 3SLS models providing significant improvements in the forecasting accuracy of the pavement performance. Pavement performance modeling with 3SLS has, therefore, the potential to help roadway Agencies cut costs through a more effective and efficient allocation of pavement management resources.

2:45 PM – 4:30 PM

Poster Program: Session: Transportation Network Modeling (60)

Implications of Cost Equity Consideration in Hazmat Network Design (16-4734)

Longsheng Sun, University at Buffalo

Mark Karwan, University at Buffalo

Changhyun Kwon, University of South Florida

No abstract available