Phd Dissertation

Transportation network companies’ (tnc) impacts and potential on airport access

Publication Date

August 9, 2018

Author(s)

Karina Hermawan

Areas of Expertise

Public Transit, Shared Mobility, & Active Transportation

View Phd Dissertation

Abstract

When Transportation Network Company (TNC) services first emerged, there was extensive discussion in the popular press and among academics about the benefits that these “shared” services would bring. TNC as a form of ground transportation to and from the airport in contrast, is less often studied or permitted. TNC operations at airports are highly controversial. At Los Angeles International Airport for example, Uber and Lyft could not conduct pickups until about seven years after they were founded. Still, research on both airports and TNCs rarely intersect. This dissertation aims to fill the gap in the literature and address such questions as: which and how many airports have various types of TNC service (standard, pooled)? How do they impact other modes, vehicle-occupancy, congestion, and access at airports? Can their service be modified (i.e. through pricing or service improvement) to encourage higher uses of shared modes? Using Uber and Lyft websites, it documents all airports in the U.S. and internationally that permit TNC service and the type of services available. It analyzes airport passenger surveys to evaluate how much TNC replaces and complements transit and the net effects at several airports. Also using the passenger survey, Google Maps Directions API, and other sources, it estimates travel time and costs of the different modes to the airport, builds a discrete choice model of the access mode choices, and simulates various scenarios; some of the scenarios are a TNC price increase (to match the cost of taxis) or a price cut and travel time increase (to mimic Uber Pool and Lyft line which are carpool versions of TNCs). Finally, it assesses how a pooled TNC service to the airport would operate. We apply the pick-up and delivery problem to airport access requests (formed based on the airport passenger survey) and measure the number of private trips that would be eliminated when passengers are pooled. The motivation for understanding the consequences of making private TNCs more expensive, or pooled TNCs less expensive and more efficient (with shorter detours or travel time) is to identify effective tools to encourage modal shifts to vehicles with higher occupancy.

MS Thesis

Analysis of health impacts resulting from truck and rail emissions reductions attained through the san pedro bay ports clean air action plan programs

Publication Date

June 29, 2018

Author(s)

Tammie Kuo

Areas of Expertise

Freight, Logistics, & Supply Chain Safety, Public Health, & Mobility Justice

View MS Thesis

Abstract

Various policies have been implemented to deal with the air pollution generated by freight operations at the Ports of Los Angeles and Long Beach (also known as the San Pedro Bay Ports, or SPBP), including mandating cleaner vehicles and cleaner fuels, shifting container transport from trucks to trains for long distance travel, or shifting freight deliveries from peak to off-peak hours. The purpose of this thesis is to analyze the co-benefits of some of these policies on the reduction of regional pollutants, and on human health. Port-related emissions of nitrogen oxides (NOx) and particulate matter (PM2.5) are dispersed throughout the surrounding area using CALPUFF, and the resulting pollutant concentrations are compared between milestone years for the policies and a baseline year, 2005, using EPA’s BenMAP health analysis model. Results show that within the SPBP boundaries, the Clean Trucks program has cut heavy duty vehicle NOx emissions by 65-80% between 2005 and 2012, and PM2.5 levels have been reduced by over 95%. This has resulted in net annual savings related to cardiovascular and respiratory impacts of over $9 million. The Rail Line-Haul and Switcher Fleet Modernization program has achieved lower pollutant reductions, around 50% for NOx and 45% for PM2.5, but the broader range of this program’s impacts has resulted in even higher net savings of over $100 million between 2005 and 2012. These examples indicate that the Clean Air Action Plan has made a positive impact on quality of life for residents in the Los Angeles area.

Phd Dissertation

Impacts of Capacity Drop on Freeway Control

Publication Date

July 25, 2018

Author(s)

Felipe De Souza

Areas of Expertise

Infrastructure Delivery, Operations, & Resilience

View Phd Dissertation

Abstract

An unfortunate feature of freeway traffic flow at merge bottlenecks is the capacity drop (CD) phenomenon. It refers to a drop in the bottleneck outflow when a queue forms upstream to that bottleneck compared to the outflow observed before the formation of the queue. While its causes and exact mechanism are still open questions, this research concerns in the impacts of CD and how to mitigate them.

The distinct features of CD in a freeway corridor are assessed based on the behavior of equilibrium states in a model capable of replicating CD. The impacts are unveiled by comparing the system properties with and without the CD. The main finding is that the highest outflow occurs under uncongested equilibrium; however, it may not be reachable depending on the demands and initial conditions.

The local ramp metering control is investigated into more details. CD imposes a hysteresis on the system response with respect to the demand level. Also, we analyze the system in closed loop considering ALINEA, a well-known control algorithm. We establish the stability range with respect to parameters which is a necessary requirement for the controller to be effective. Further, we propose an extension of ALINEA to enlarge the stability range mitigating a performance loss that occurs when the on-ramp and the bottleneck are far apart.

Essential aspects of ramp metering are better captured with microscopic models; however, there were few evidences that such models can replicates CD. To that end, we propose a parameter calibration procedure that ensures the underlying model properly captures CD. The approach is tested with loop detector data from a merge bottleneck in which the CD is consistently observed.

All results with different approaches point to the direction that the existence of CD imposes additional challenges on the system control. Fortunately, in most cases the effects of CD can be mitigated with a properly designed control strategy, such as the ones tested and proposed in this research.

Phd Dissertation

Driver Response to Variable Message Signs in a 2D Multiplayer Real-time Driving Simulator

Publication Date

June 29, 2018

Author(s)

Si-Yuan Kong

Areas of Expertise

Infrastructure Delivery, Operations, & Resilience Travel Behavior, Land Use, & the Built Environment

View Phd Dissertation

Abstract

This research seeks to understand how information displayed by variable message signs (VMS) can affect driver route-choice and be better used for active traffic incident management. I study the effect of various VMS messaging strategies using a money incentivized behavioral experiment with a novel 2D real-time driving simulator that supports dozens of subjects driving on a shared virtual roadway where traffic incidents unpredictably occur. Drivers are shown a VMS display before choosing between two congestible routes. I conducted this experiment with students at the UCI Experimental Social Science Laboratory (ESSL) and with a more diverse sample of online subjects crowdsourced from the Amazon Mechanical Turk (MTurk) marketplace.

Chapter 1 will present the research motivation and methodology, the design and implementation of the experiment platform, and the results with student subjects. I find that subjects learned to efficiently operate the driving simulator, all tested VMS messaging strategies improved aggregate outcomes compared to the No VMS baseline, displaying messages didn’t cause highly volatile diversion rates, and subject gender exhibited consistent correlations with route choice.

Chapter 2 will discuss the reasons for replicating on MTurk, the methodological modifications necessary to conduct the experiment online, and how the MTurk results compare to the student results. I find that it’s viable but challenging to conduct real-time multiplayer experiments on MTurk, there are significant differences in individual characteristics between the MTurk and student subjects, and there are limited behavioral differences between the two groups.

Chapter 3 will introduce a framework using long short-term memory (LSTM) neural networks to predict driver route choice using real-time contextual data. I use varyingly limited vectors of data from my driving simulator experiments as the neural network’s input to predict driver route choice at the decision point between the two available routes. I find that the best performing model configuration can predict individual route choice with 74.0% average accuracy with in-sample cross validation and 72.2% average accuracy with out-of-sample validation.

Phd Dissertation

On the Complexity of Energy Consumption: Human Decision Making and Environmental Factors

Publication Date

June 29, 2018

Author(s)

Jaewoo Cho

Areas of Expertise

Travel Behavior, Land Use, & the Built Environment Zero-Emission Vehicles & Low-Carbon Fuels

View Phd Dissertation

Abstract

Given our rapidly changing society, the complexity of residential energy often hinders the efficacy of energy conservation policies designed to address our current social and environmental problems. Therefore, understanding this complexity appears to be essential to successfully building and efficiently implementing energy policies. The present dissertation attempts to advance our understanding of the dynamics and complexity of residential energy consumption by investigating various determinants and contextual factors through the three interrelated pieces of applied research. Using American Housing Survey (AHS) data, the first study investigates the dynamics of residential energy consumption at the micro level. It is found that the electricity consumption of households who have moved into new homes is generally lower than average, and their consumption is found to increase as the period of residence increases. The second study examines the relationship between the choice of energy-efficient systems and inter-agent dynamics. By employing a logistic regression model with two national datasets, the Residential Energy Consumption Survey (RECS) and the American Community Survey Public Use Microdata Sample (ACS PUMS), the empirical analysis reveals statistically significant differences in the installation of solar energy systems among households with different degrees of two major inter-agent issues—split incentives and split decision-making problems. The last study focuses on the complexity of residential energy consumption relevant to the surrounding environments, and it pays special attention to seasonality. Based on city-wide data from Chicago and using a special econometric model, the empirical analysis reveals the seasonal dynamics between urban forms and residential energy consumption. Through these three empirical studies, this dissertation explores the dynamics of residential energy consumption in various dimensions and reveals the complicated mechanisms that determine residents’ choices with respect to energy consumption. The evidence from this study is especially important because it reinforces the conclusion that there is no panacea when addressing energy issues. This study suggests that policy-makers and planners should instead thoroughly understand a wide range of contextual factors and their influences in order to develop more effective, context-specific energy policies that best fit each distinct geographical and socio-economic situation.

Phd Dissertation

Modeling Disruptions to Roadway Network Bridges, Restoration Workforce, and Vehicle-carried Information Flow for Infrastructure Management

Publication Date

June 29, 2018

Author(s)

Pierre Auza

Areas of Expertise

Infrastructure Delivery, Operations, & Resilience

View Phd Dissertation

Abstract

The ability to model the disruptions of adverse events on various systems, such as infrastructural and social, is an important tool to assessing these systems’ resilience. While previous research on system resilience concentrated on physical infrastructure such as transportation systems, two recent research topics include social resilience and dependencies across many infrastructure systems. For example, transportation is dependent on such systems as power, communications, and the workforces that are key to restoring these infrastructure systems. This dissertation contains three disruption modeling studies that have followed the evolution of resilience research over the past decade from physical systems to interrelated topics. The first study uses mesoscopic traffic simulation to evaluate seismic risk of potential travel time increases from earthquake damage to bridges in a roadway network. This analysis successfully obtained system risk curves of network-wide travel time increases. The second study shifts focus towards workforces that participate in restoring infrastructure systems. It identifies transportation and communications workers and calculates these workers’ exposure to the Peak Ground Accelerations (PGAs) of a 7.8 magnitude Southern California scenario earthquake. Indeed, for this scenario, transportation workers are exposed to statistically significant higher PGAs than non-transportation workers, and communication workers to significantly lower PGAs. The third study proposes a model for the travel time of information along communication-equipped vehicles physically traveling in a network. Vehicles are sampled as equipped vehicles, then their trajectories are analyzed to (1) estimate equipped vehicle link flow and turning movement counts and (2) estimate the frequency of equipped vehicles encountering each other on links and at nodes. This study compares two scenarios: the baseline scenario and a work zone scenario that corresponds to a bridge being damaged in the network. It is hypothesized that there would arise a difference in expected path travel times when (1) the representation of a specified subpath within the sample is increased and (2) when vehicles are routed along currently unused subpaths. This dissertation concludes with a discussion of the contributions of all three studies, as well as suggestions for future work.

Phd Dissertation

Paradigms of Identifying and Quantifying Uncertainty and Information in Constructing a Cognition-Modeling Framework of Human-Machine Transportation Systems

Publication Date

June 29, 2018

Author(s)

Jiangbo (Gabe) Yu

Areas of Expertise

Travel Behavior, Land Use, & the Built Environment

View Phd Dissertation

Abstract

This dissertation proposes a set of coherent cognition-based paradigms to allow greater sensitivity and adaptability to the emerging technologies and behavioral policies. These paradigms are derived from a cognition-based framework that explicates information source, medium, sensation, perception, and learning. The feasibility of the framework is demonstrated through an analytical example of multi-stakeholder decision processes and human-machine systems where the two types of entities can be incorporated into the same modeling scheme. Using the framework as guidance also reduces the challenges from information intractability and data redundancy of agent-based modeling practice.

The first paradigm follows the strict definition of information in Information Theory and models it as the change of uncertainty, which is applied in quantifying traveler information for the evaluation of dynamic message boards that present various contents at candidate locations in Los Angeles traffic networks.

The second paradigm is developed for a utility-based decision model under risk around the proposed concept, Elastic Surprise. This concept makes feasible the differentiation between probability misperception and perceived uncertainty. It is shown that conventional methods of decisions under risk such as Expected Utility Theory and Cumulative Prospect Theory are special cases. In addition, a specific form of Elastic Surprise under particular assumption on human’s cognition leads to Shannon’s information entropy and, hence, connects with the first Paradigm. The method is tested in conjunction with the Cumulative Prospect Theory on travel time equivalency under risk in a survey study. The results show improvement in data fitting and output interpretability.

Finally, guided by the framework, the paradigms are tested on a case study of multi-class multi-criteria dynamic traffic assignment where heterogeneous travelers’ risk preference on travel time is explicitly modeled. The algorithm approaches the user equilibrium through a stochastic quasi-gradient projection-based algorithm that shows the improvement in computational efficiency and cognitive implication of the agents’ decision rules. I also discuss the potential strategies and policies implication for system improvement.

research report

An Analysis of Travel Characteristics of Carless Households in California

Publication Date

May 1, 2018

Associated Project

An Analysis of Travel Characteristics of Carless Households in California

Author(s)

Jean-Daniel Saphores, Suman Mitra

Areas of Expertise

Travel Behavior, Land Use, & the Built Environment

View research report

Abstract

In spite of their substantial number in the U.S., the research team’s understanding of the travel behavior of households who do not own motor vehicles (labeled “carless” herein) is sketchy. The goal of this paper is to start filling this gap for California. We perform parametric and non-parametric tests to analyze trip data from the 2012 California Household Travel Survey (CHTS) after classifying carless households as voluntarily carless, involuntarily carless, or unclassifiable based on a California Household Travel Survey question that inquires why a carless household does not own any motor vehicle. We find substantial differences between the different categories of carless households. Compared to their voluntarily carless peers, involuntarily carless households travel less frequently, their trips are longer and they take more time, partly because their environment is not as well adapted to their needs. They also walk/bike less, depend more on transit, and when they travel by motor vehicle, occupancy is typically higher. Their median travel time is longer, but remarkably, it is similar for voluntarily carless and motorized households. Overall, involuntarily carless households are less mobile, which may contribute to a more isolated lifestyle with a lower degree of well-being. Compared to motorized households, carless households rely a lot less on motor vehicles and much more on transit, walking, and biking. They also take less than half as many trips and their median trip distance is less than half as short. This study is a first step toward better understanding the transportation patterns of carless households.

Related Publications

policy brief | May 1, 2020

Travel Varies Greatly Between Voluntary Versus Involuntary Carless Households in California

published journal article | March 19, 2019

How do they get by without cars? An analysis of travel characteristics of carless households in California

research report

Situational Awareness for Transportation Management: Automated Video Incident Detection and Other Machine Learning Technologies for the Traffic Management Center

Publication Date

May 29, 2018

Associated Project

Automated Video Incident Detection System

Author(s)

Craig Rindt

Areas of Expertise

Infrastructure Delivery, Operations, & Resilience Intelligent Transportation Systems, Emerging Technologies, & Big Data

View research report

Abstract

This report provides a synthesis of Automated Video Incident Detection (AVID) systems as well as a range of other technologies available for Automated Incident Detection (AID) and more general traffic system monitoring. In this synthesis, the authors consider the impacts of big data and machine learning techniques being introduced due to the accelerating pace of ubiquitous computing in general and Connected and Automated Vehicle (CAV) development in particular. They begin with a general background on the history of traffic management. This is followed by a more detailed review of the incident management process to introduce the importance of incident detection and general situational awareness in the Traffic Management Center (TMC). The authors then turn their attention to AID in general and AVID in particular before discussing the implications of more recent data sources for AID that have seen limited deployment in production systems but offer significant potential. Finally, they consider the changing role of the TMC and how new data can be integrated into traffic management processes most effectively.

Phd Dissertation

Essays in the Economics of Transportation and the Environment

Publication Date

April 14, 2018

Author(s)

Amine Mahmassani

Areas of Expertise

Transportation Economics, Funding, & Finance

View Phd Dissertation

Abstract

This thesis uses applied econometrics and traffic experiments to identify environmental and behavioral factors that contribute to externalities in traffic networks, as well as evaluate mechanisms designed to address them.

The first chapter examines whether exposure to ambient fine particulate matter (PM 2.5) increases the likelihood of getting into a vehicle collision. PM 2.5 has been shown to affect alertness and cognition, which may in turn impair driving ability. Variation in daily AQI level from PM 2.5 was exploited to identify a possible causal effect on daily car accident rates in nearby cities. This approach yielded no evidence of a causal effect on vehicle accidents, perhaps due to endogeneity of PM 2.5 with other factors correlated with accident frequency. An alternative instrumental variables approach exploited exogenous shifts in wind direction relative to nearby coal power plants – a significant point source of PM 2.5. This specification found that a one-standard deviation in PM 2.5 AQI increases the car accident rate by 13.2 percent.

The second chapter investigates if the presence of multiple states in traffic networks adversely impacts the speed at which users learn route-choice equilibria. To answer this question, data were generated from several sessions of a repeated binary route-choice experiment with human subjects. Exogenous random state changes were introduced as discrete, varied reductions in roadway capacity. The sessions were comprised of either a “simple” network treatment with only two states, or a “complex” network treatment with five states. Reinforcement learning models estimated from the experimental data show that learning was significantly impaired in the complex five-state treatment but not the simple two-state treatment. Simulations based on the learning behavior estimated from each treatment showed that the impaired learning from the five-state treatment resulted in disproportionately slower (and sometimes non-existent) equilibrium convergence compared to learning with two-states.

This third chapter demonstrates the workability of a truth-telling mechanism for efficiently allocating freeway capacity. I conduct a traffic experiment using an interactive multi-user driving simulator in which I allocate human subject drivers to freeway lanes using an optimal tolling scheme where users reveal their valuation of the road through a Vickrey-Clarke-Groves mechanism. I find that the mechanism generally elicits truthful values of time from subjects. However, there are also significant and persistent deviations from truth-telling caused largely by difficulty in understanding the complexity of the mechanism as well as stochasticity in travel time outcomes. Nevertheless, I show that the mechanism dominates alternatives under a plausible set of assumptions.