2021

Existing technologies like smartphones and Dedicated Short Range Communication (DSRC) have significant untapped potential to enhance the experience of non-motorized users as they navigate the transportation system—and not just their interactions with vehicles. This research project will explore the near-term opportunity to interface DSRC-enabled infrastructure with a smartphone-based platform to improve the safety and user experience of pedestrians and bicyclists via a novel connected countdown timer-based display. For the bicyclist, this display would indicate the approximate remaining green time of a downstream traffic signal and enable the bicyclist to request extra time to make it through a changing signal. For the pedestrian, this display would non-numerically depict the amount of time remaining before the walk signal is displayed. The primary research question for both displays is: Does the extra information provided by the display improve bicyclist and/or pedestrian safety and compliance with the traffic signal? The research will be conducted in a three-year phased study: a widescale survey (phase one) will identify display concepts to be tested in simulator environments for safety and compliance (phase two), then scaled to field testing (phase three).

Many public agencies and advocacy organizations have worked to decrease the reliance of Americans on the private automobile. While Americans routinely use and own automobiles at high rates, researchers have noted negative consequences, including air pollution, congestion, and deaths and injuries from collisions. In this context, a better understanding of cohort influences on travel behavior can help predict future trends in auto ownership and use.

In this paper, I examine the influence of lifecycle, cohort, and period effects on vehicle ownership in the U.S. over the last two decades. The analysis relies on data from the 2000 Census and 2005, 2010, and 2015 one-year American Community Survey Public Use Microdata Samples (PUMS). I model the likelihood of living in a zero-vehicle household by generation and year, and then use median polish analyses to test the presence of cohort effects.

I find that that the Silent Generation (born prior to 1946) are less likely to live in households with cars than more recent cohorts. However, Millennials differ from previous generations, as they too are less likely to live in households with automobiles. Although additional data are needed to draw a definitive conclusion, the findings suggest that Baby Boomers may represent a generation of “peak” automobile users. 

The engineering properties of asphalt mixtures change with time. Shortly after placement, asphalt concrete (AC) layers are more susceptible to rutting. As the pavement ages, the AC layer becomes stiffer, more brittle, and thus more susceptible to cracking. Current protocols provide guidelines for the selection of materials, the determination of the material proportions (e.g. aggregates and binder content), and the evaluation of the engineering properties (e.g. cracking and rutting potentials) of any given AC mix. However, these protocols do not provide any check on the impact of aging on the mixture. This project seeks to investigate and propose protocols to simulate realistically the short- and long-term aging states required to evaluate the asphalt mixture resistance to rutting and cracking, respectively. Such conditioning protocols must be valid for all types of AC mixes including, but not limited to mixtures that incorporate reclaimed asphalt pavement (RAP), reclaimed asphalt shingles (RAS), recycling agents, and other additives. By investigating existing and novel laboratory methods, this study seeks to propose protocols that simulate the two critical aging states needed to design properly an AC mixture, as well as provide a deep understanding of how curing affects the physical and engineering performance of a mixture. The study leverages existing research studies and available performance data along with a systematic and innovative test matrix to optimize the curing conditions.

COVID-19 has totally disrupted the daily life of almost all Americans. In addition to the significant loss of life, it has significantly impacted many economic sectors. Given that transportation is central to the economy, the impact of COVID-19 on mobility patterns has been unlike anything we have experienced before and is yet to be fully understood. As the virus continues to spread around our communities, some of its impacts could become permanent. As such, it is important to understand which personal mobility impacts due to the pandemic are temporary and which are long lasting/permanent.

This study aims at understanding the heterogeneity of personal mobility patterns throughout the COVID-19 pandemic. More specifically, the study aims at 1) identifying any spatiotemporal variations in personal mobility trends throughout the pandemic and examine whether any changes to regional personal mobility trends are related to regional spikes in COVID-19 spread, 2) exploring the dynamics of daily profiles of different trip types across different socio-demographic groups, and 3) revealing any changes in personal mobility throughout the pandemic between groups with different pre-COVID19 personal mobility behavior. This study achieves these goals by using point of interest (POI) data from approximately 60,000 U.S. panelists. We supplement the POI dataset with a survey of ~1000 panelists to measure the economic well-being, mental health, personality, employment status, and social distancing behaviors throughout the pandemic. These additional dimensions will help further explain any determinants of change in personal mobility during the pandemic.

Transportation planners tend to characterize walkability based on destination accessibility and pedestrian infrastructures such as sidewalks and crosswalks. Urban designers, meanwhile, emphasize how streetscape characteristics such as spatial enclosure and active facades are important for creating inviting pedestrian environments. The degree to which streetscapes influence walkers’ preferences in real-world settings, however, remains unclear. It is difficult to represent streetscapes in stated-preference surveys in ways that are sufficiently realistic to produce naturalistic responses. At the same time, it is difficult to disentangle streetscape effects on naturalistic revealed preferences, such as route choices, amongst confounding factors such as directness.

This study addressed these challenges by conducting a stated preference experiment in a field setting, offering subjects choices between real-world stimuli. Confounding factors were accounted for through semi-structured interviews conducted alongside the preference experiment. During the experiment, subjects compared opposing block faces with contrasting streetscapes, choosing which side of the street they would prefer to walk along and describing which factors contributed to their choice. They also ranked block faces along the entire route against one another. This produced a multidimensional dataset describing preferences in both quantitative and qualitative terms.

Counter to expectations, subjects preferred streetscapes fronted by open spaces with trees over those with enclosing buildings. However, subjects did prefer more active facades and particularly disliked long stretches of blank wall. This suggests that transportation planners aiming to improve walkability may find it useful to look beyond the roadway and advocate for design codes that encourage green features and active building facades.

Well-designed transportation demand management (TDM) strategies are needed to fully leverage the potential of pooled on-demand mobility to lessen congestion, energy use, and greenhouse gas (GHG) emissions by reducing private-vehicle ownership and enabling higher vehicle occupancy rates. While organized ridesharing has historically been focused on commuters, with large employers playing a central role in incentivization and facilitation, app-based ridesharing services and on-demand pooling provided by transportation network companies (TNCs) and microtransit offer the opportunity to drastically expand the pooling market, particularly in concert with effective incentive mechanisms. Key findings will be presented from a stated preference survey distributed in four California metropolitan regions from August to December 2018: Los Angeles, Sacramento, San Diego, and the San Francisco Bay Area. Important socio-demographic, TNC trip purpose, and general mobility profiles of TNC users will be examined. In addition, results from a discrete choice analysis of the choice to use ride-alone TNCs, door-to-door shared rides, or indirect shared ride services will be discussed, which reveal significant factors in pooling demand sensitivity including variation across trip contexts, metropolitan regions, socio-demographics, travel behavior, and attitudes and perceptions toward sharing. The talk will conclude with a presentation of preliminary results from a study of congestion pricing optimization using the Berkeley Integrated System for TRansportation Optimization (BISTRO), in which the heterogeneity of TNC mode choices are accounted for by implementing this mode choice model into an agent-based simulation and activity-based travel model of the City of San Francisco.

The rise of e-commerce and availability of fast and/or free delivery engendered a new wave of online platforms for ordering food and household goods. Such platforms are evolving the modes and mechanisms through which grocery shopping occurs given many platforms offer same-day delivery or curbside pickup options. Such services may help cities, households, and businesses remain resilient through the COVID-19 pandemic, which has drastically impacted travel for in-person shopping, global supply chains, and food business operations. To investigate the role online platforms may play in shifting travel behavior for household provisioning during and beyond the pandemic, a longitudinal survey was being administered across five U.S. states—AZ, FL, MI, WA, and OR. The first wave captured changes in household provisioning before and after the state-of-emergency declaration for COVID-19, including the use of online platforms for pickup or delivery. The second wave, fielded in December 2020, focused on adoption of, attitudes toward, and barriers to using online platforms for household food shopping in response to COVID-19. Initial results from the first two waves shed light on the variance in effects of online platforms across demographics and built environments. Further, the complementary versus substitutional nature of online-ordering and in-person shopping is explored. The culmination of these survey data pose implications for policy related to urban freight, public health in access to food, and continued response to the COVID-19 pandemic.

Pedestrian infrastructure plays a critical role in city life, and many of its flaws have been exposed during the coronavirus pandemic. As city residents move outdoors to dine, recreate, and social distance the cracked, inaccessible walkways, or even complete lack of sidewalks has become much more apparent. There is an apparent need in many American cities for improving sidewalk quality and quantity to facilitate safer, healthier, and more accessible communities. Some cities struggle to fund, maintain, and build sidewalks because of fractured policies. Many leave sidewalk repairs and maintenance in landowners’ or developers’ hands, which can often raise issues of equity and right-of-way. My research will answer how cities can most efficiently and equitably overhaul or upgrade their current sidewalk funding and management systems. I plan on using spatial analyses and asset management principles to quantitatively analyze and compare what cities across the United States and world are doing to determine the most efficient and equitable way to fund and prioritize sidewalks. With this research, I plan to recommend or rank funding mechanisms and prioritization hierarchies or tools for city sidewalks. In addition to preparing a one-page summary and infographic of data to make available to city departments of transportation and public works, I will complete and submit a full research to the Transportation Research Board in the summer of
2021.

Transit goals have typically focused on commuter trips but facilitating urban last-mile freight logistics is a potential strategy to mitigate the demands of parcel distribution on the transportation network. Presently, most parcel lockers operate out of private businesses, but consumer surveys have found that transit users may be interested in locker facilities at transit connections. The implementation of an unmanned, secure, common carrier parcel locker system could have benefits for non-transit users as well. Consolidation of deliveries would benefit courier companies by allowing operations at increasingly competitive rates; retailers and consumers benefit from low shipping rates. This evaluation includes a case study of the light rail stations, transit centers, and transit malls in the greater Portland, OR metro. Their potential of hosting transit sites is reviewed based on ridership (the number of ons/offs at transit facilities), selecting a balance of central city and suburban locations, the size of populations in influence areas (whether transit users of not), and a framework for prioritizing locations based on best-practice equity metrics. The suitability of locating lockers at park-and-ride facilities, transit centers, transit malls, and high volume bus/light rail stops is presented, with contextual considerations pertinent to facility type. Particular attention is given to the park-and-ride facilities outside the urban core for their suitability in reaching populations in lower-density developments, as well as the potential of consolidated distribution points in city resiliency plans.

Mass evacuations, particularly those at a statewide level, represent the largest single-event traffic movements that exist. These complex events can last several days, cover thousands of miles of roadway, and include hundreds of thousands of people and vehicles. Often, they are marked by enormous delays and heavy congestion and are nearly always criticized for their inefficiency and lack of management. However, there are no standardized methods by which to systematically quantify traffic characteristics at the proper scale. Several recent evacuations have occurred in the United States. Wildfire evacuations have been ordered in the state of California while Hurricanes have led to evacuations in the state of Florida. It has generally been accepted that the evacuation from a regional wildfire is fundamentally different than the evacuation from a hurricane. Hurricane evacuations generally encompass larger areas when compared to wildfire evacuations and provide several days of advanced warning. Whereas, wildfires impact smaller areas with significantly shorter warning time. On the other hand, at the broadest level, evacuees and their vehicles move in both time and space. This research seeks to develop a better understanding of the travel flow principles that govern the evacuation process and its impact on the mobility of a community, for different hazard types. The goal of this research is to build upon the prior knowledge and expand the scientific understanding of the evacuation process by systematically analyzing evacuations from hurricane and wildfire events.

Negotiating horizontal curves is a high-risk tactical driving maneuver. Drivers must simultaneously and adeptly control their steering adjustment, speed, and lane positioning, as well as accurately perceive the curvature of the road segment and adjust to proprioceptive cues. To investigate driving curve negotiation ability among at-risk populations, researchers often use driving simulators, which allow for precise collection of numerous outcome variables in a safe, controlled environment. Published works often include subsets of variables that capture categories of driving behavior, such as dynamic movement (e.g., longitudinal velocity, longitudinal acceleration) or steering control (e.g., variability in lane positioning, number of lane departures, steering velocity, number of steering reversals). However, to date, these driving behavior categories have been theoretically but not quantitatively associated. The proposed study will conduct Principal Components Analyses (PCA) using a dataset including adolescent and young adult drivers with autism spectrum disorder (ASD), attention-deficit/hyperactivity disorder (ADHD), and neurotypical development (TD) to establish how extracted components of simulated driving metrics (i.e., quantitative “categories” of data) relate to typically utilized qualitative variable categories used in driving simulator research (e.g., dynamic movement, steering control). PCA will be conducted using both the entire dataset for a global examination of horizontal curve negotiation metrics as well as diagnostic status subsets to examine if relationships between driving variables produce differential patterns among different types of drivers. Improved quantification of horizontal curve negotiation can inform future training and education protocols to ensure safe driving among drivers, including those with diagnoses that may put them at higher risk.

We propose a novel method for verifying safety properties of neural networks. A motivating example is given for verifying safety of neural networks used to advise aircraft of imminent mid-air collisions. Neural networks are useful for this purpose because they provide a collision-avoidance policy that has a small data footprint which is necessary for small unmanned aircraft. Such collision-avoidance networks should obey certain input-output safety specifications if an aircraft is to seek certification from the Federal Aviation Administration (FAA). One such property is that the network should advise an evasive maneuver if an intruder aircraft is directly ahead. However, certifying safety of neural networks has only recently been possible. The proposed method builds on the current literature to provide marked advantages when identifying the existence of unsafe advisories which the neural network may output. In addition, the proposed method can be used for safety analysis of a closed-loop dynamical system by computing forward and backward reachable sets in time. To accomplish this, we note that neural networks with the ReLU activation are piecewise-affine (PWA) functions. Forward and backward reachability problems can be solved exactly for such functions at relatively little computational cost. Our approach uses linear programming to incrementally define the explicit PWA representation associated with a neural network, allowing for the incremental evaluation of safety properties.

Autonomous vehicles (AV) differ significantly from traditional passenger vehicles in both their behavior and physical characteristics. As such, the validity of the guidance provided in the Manual for Assessing Safety Hardware, Second Edition (MASH 2016) is questionable in AV applications. Impact angles, speeds, and vehicle weights specified in MASH 2016 are inextricably linked to the traditional vehicles underlying the estimates. For AV applications, these parameters must be estimated from the ground-up, stepping outside the guidance of MASH 2016. In this paper, a conservative method for evaluating existing infrastructure to support AV traffic is proposed. The method integrates traditional structural analyses with unconventional methods of estimating impact conditions. This methodology was developed for the Jacksonville Transportation Authority, who, when faced with unique challenges in maintaining and expanding their Automated Skyway Express, opted to convert the system from monorail to AV traffic. Leading AV developers were surveyed to develop a portfolio of potential candidates for the conversion. Then, estimated impact conditions were compared against the capacity of the system’s existing concrete parapets. Ultimately, safe operating speeds for each AV candidate were recommended on the bases of structural capacity and vehicle stability. All but one AV candidate were deemed capable of safely operating at the desired speed of 25 mph without any modifications to the barrier. Although the methodology was developed for a particular case, it is applicable to future implementations of AVs on existing infrastructure, provided the roadway is confined similarly to the Skyway deck.

Over the past two decades, superloads, or vehicles with gross vehicle weight over 150,000 pounds, have been increasingly utilized to transport heavy loads, such as prestressed concrete girders, automotive presses, transformers, wind turbine components, and other substantial loads. Since these superload have a significant effect on the infrastructure system in comparison to regularly permitted vehicles, they should be subject to special consideration in the permitting and operation process. Despite the great research effort that has been made to improve the superload permitting process, few studies have been performed on the characterization and prediction of superload. Superload has its own distinct characteristics that differ from other vehicle loads. Thus, there is a need to better understand the characteristics of superload and to develop a procedure to predict vital superload attributes for enhanced accuracy in the permitting process. In this research, the major focus was to develop an analytical procedure for the characterization and prediction of superload using advanced gradient boosting machine (GBM) learning algorithms. A similar study will be performed on overweight vehicles, which are vehicles
between 80,000 pounds and 150,000 pounds. Additionally, artificial intelligence and big data will be utilized in order to derive the optimal management and maintenance plan for bridges based on the findings of all permit vehicles. The study will be using a probability-based algorithm that combines data from multiple bridges, pavement data, traffic data, environmental data, and human factor data to provide an objective maintenance plan that ensures the integrity of Florida’s infrastructure.

High-capacity on-demand mobility has the potential to overhaul public transportation by closing the gap in the ridership versus coverage dilemma. However, as evidenced in a recent report by the Eno Center for Transportation, most microtransit pilot programs in the US have experienced difficulties with maintaining public reception and quality of service. A key technical challenge for the successful deployment of a microtransit service is the optimization of system-level operations. In this research, we bridge theory and practice by taking a principled approach to tackle the optimization problems arising in on-demand microtransit (e.g., vehicle routing, passenger matching). Given the link between system-level performance and quality of service, our research may enable a better assessment of the potential of this mode of transportation.

Many cities and regions across the country are considering reducing or eliminating single-family zoning as  a way to increase housing affordability. An important question brought forth by these policy changes is how  much effect they will have on the built environment. While large swaths of metropolitan areas may be rezoned  to allow more dense housing, construction of new housing in many of these areas may not be economically  feasible. In this research, I evaluate where development of new multifamily dwellings would be economically  feasible in Southern California, if single-family zoning were relaxed. I do this by estimating the construction  cost for a number of building plans, ranging from duplexes to 6-unit developments, that could fit on existing  single-family lots. For each residential parcel with a potential zoning change, I evaluate which of these building   plans could be constructed on the lot. I compare the construction costs of all  geometrically feasible buildings to their rental values in that location. I estimate that 10% of parcels would be  redeveloped under a zoning change, producing 580,000 new units. If zoning changes were instead confined  only to areas within 1/2 mile of transit, approximately 40% of this redevelopment would occur.  The results are highly sensitive to model assumptions, however, but all results show a spatial pattern with development concentrated in higher income areas, lessening gentrification concerns. Ultimately the outputs of this model will be fed into a travel demand model to understand the transport impacts of this policy change.

The advancement of in-vehicle sensors provides abundant datasets to estimate parameters of car-following models that describe driver behaviors. The question of parameter identifiability of such models (i.e., whether it is possible to infer its unknown parameters from the experimental data) is central to the well-posedness of the inverse problem, and yet still remains open. This article presents both structural and practical parameter identifiability analysis on four common car-following models: (i) the constant-time headway relative-velocity (CTH-RV) model, (ii) the optimal velocity model (OV), (iii) the follow-the-leader model (FTL) and (iv) the intelligent driver model (IDM). The structural identifiability analysis is carried out using a differential geometry approach, which confirms that all of the tested car-following systems are structurally locally identifiable, i.e., the parameters can be uniquely inferred under almost all initial condition and all inputs by observing the space gap alone. However, the analysis fails to detect specific initial conditions and parameter sets that cause non-identifiability. We propose a optimization-based numerical direct test to determine if the model is identifiable given a specific experimental setup (initial conditions and input are known). Direct test conclusively finds distinct parameters under which the CTH-RV and FTL are not identifiable under the given initial condition and input trajectory. When small deviations in the output are permitted (e.g., due to measurement error), the numerical test finds that none of the models are practically identifiable.

My research compares the decision-making processes within highway agencies regarding roadside vegetation management policy. I am conducting interviews with landscape architects at state highway agencies and followed a snowball method to continue interviewing personnel in the maintenance and safety departments, geometric design departments, and others depending on results that emerged.
Roadside vegetation policy runs the gamut among states, from managing native plantings and pollinator habitat to states that routinely clear roadsides of both woody and herbaceous growth. As storm resiliency in the form of tree removal is incorporated into more state practices, as tree mortality from pests and drought increases, and as environmental regulations are loosened, the trend among states may bend toward clear-cutting.

Highway agencies remove trees from roadsides for several reasons, most frequently with the goal of eliminating run-off-the-road crash fatalities. While a roadside clear of trees is maintained as an absolute rule of road design, and has been since the 1960s, its relationship to road safety is still disputed. Further, if treated as a maintenance project, highway agencies are not required to analyze the ecological effects of tree removal on habitat reduction, erosion, carbon sequestration, or any other environmental metrics, even if the removal constitutes acres of land. Further, agencies generally have no public process to evaluate tree removal decisions, though some voluntarily engage the community prior to taking action.

This research offers insights into best practices for collaborative decision-making among staff at highway agencies, often siloed into engineering, environmental, and maintenance units.

This poster investigates the ways that transit users learn about, make decisions around, and ultimately change the ways they commute via public transit.  The setting for this research is the newly instituted Southeastern Pennsylvania Transportation Authority (SEPTA) bus Route 49 in Philadelphia – the first new bus route added to SEPTA’s network in a decade. Using a mixed methods research design that incorporates quantitative survey data collected onboard, as well as qualitative data from semi-structured interviews with select users, this poster addresses the following research questions: 1) How do users learn about the ways they travel to work via public transit and thereby gain “commute knowledge”? 2) What role do communications from formal versus informal actors play in the “commute knowledge” acquisition process? 3) How did users commute before the new bus? 4) How did users ultimately decide to change their commute to a new route? Based on initial findings, users appear to gain substantial levels of “commute knowledge” from informal mechanisms, such as conversations with coworkers, or merely from seeing the bus go by. Respondents tend to discuss the new route with their peers, signaling robust communication networks when it comes to making transit choices. Overwhelming numbers of Route 49 users switched to this route from another public transit mode. Users often make decisions around whether or not to use the Route 49 day-of or even during the commute itself. Together, these findings describe a universe of social and informal communications and sometimes spontaneous actions that shape commute knowledge.

Research on health and transport has increased significantly in the past 20 years, both across health and transportation fields. Researchers and practitioners have called for the use of health indicators in transportation, which come amidst the growing emphasis on the use of indicators for transportation plans and projects in general. The underlying hope is that new procedural arrangements, such as measuring and tracking indicators, can turn policy goals into practice. However, it is unclear if these indicators, if used, have any influence on transportation decisions. Much of the research on indicators is focused on their development and use rather than their influence.
 
This research proposal explores the use and influence of health indicators in transportation using a mixed methods approach. First, an inventory of municipal health indicators will be undertaken, addressing a gap in knowledge regarding the extent and use of health indicators. Then, a survey to municipal actors in a transportation decision-making process will explore which factors (indicator, user, organizational, and political) best explain use and influence. Finally, a common case will be selected from among the municipalities to further explore the leading theories suggested by the survey results and uncover the mechanisms by which use and influence occur.

Drivers with autism spectrum disorder (ASD) and attention-deficit/hyperactivity disorder (ADHD) may be at risk for compromised driving safety due to unsafe driving behaviors. This study examined perceived driving styles, how a driver actively chooses to drive, among drivers with ASD and ADHD relative to typically developing (TD) drivers and how perceived driving styles may subsequently affect driving behavior. Based on behavioral manifestations of each disorder, it was hypothesized that compared to TD drivers, drivers with ASD would report higher levels of anxious and patient/careful driving while drivers with ADHD would report higher levels of reckless/careless and anger/hostile driving. 48 participants (nADHD=16, nASD=16, nTD=16) matched on age, gender, and driving experience completed the Multidimensional Driving Style Inventory assessing 4 driving styles (Reckless/Careless, Anxious, Angry/Hostile, Patient/Careful). Participants drove in a driving simulator along a two-lane bidirectional road in a daytime suburban setting. Drivers with ADHD reported greater Reckless/Careless driving style compared to drivers with ASD, t(45) = 3.59, p = .001, but not the TD drivers. Drivers with ASD drove at higher speeds than drivers with ASD, t(45) = 3.01, p = .004, and TD drivers, t(45) = 2.17, p = .035. A hierarchical regression indicated a significant interaction such that drivers with ASD reporting greater reckless/careless driving style, β = -0.56, p = .048, and greater anxious driving style, β = -0.56, p = .020, drove at slower speeds compared to TD drivers. Drivers with ASD who perceive themselves as having maladaptive driving styles may drive slower to selfcalibrate their driving.

Motivated by concepts employed in travel habits and mobility biographies research, this project explores the relationship between key life events and shifts in routine transportation behavior. It uses data scraped from Reddit, a popular social media and content-sharing website, to observe real-world conversations about personal travel histories. Specifically, it focuses on understanding the relationship between changes in travel patterns and key events that correspond with these changes. Directed content analysis techniques are used to analyze 437 comments from three US cities: Atlanta, Boston, and Washington, DC.

The results suggest two distinct pathways through which routine travel behaviors change: 1) through influence of a key event, and 2) through reevaluation of available options. In the first pathway, while key events hasten change by making people consider their options, they play a causal role by either altering the transportation choices available to a person or altering their transportation needs. In the second, people reconsider their current patterns and opt to make a change, but these changes appear to happen unprompted by a key event.

These results inform a goal of sustainable transport policy in two ways. First, the best way to leverage key events into less car-dependent behavior is to ensure that viable alternatives are present when these events happen. Second, people are capable of noticing and reacting to incremental changes in the quality of available options and will likely respond to them.

This work begins to address the problems faced by free-floating scooter systems that must effectively utilize and maintain the continued participation of crowdsourced contractors, while also provide an environmentally friendly, car-free alternative to its users. We show the value of using simulation-based heuristics to place vehicles into a high-quality configuration in an attempt to improve fleet utilization. Distribution planning methods that seek to minimize the total vehicle miles traveled for the set of independent contractors are then used to achieve the found configuration. We also introduce the problem of assigning repositioning tasks in such a way that encourages positive placement decisions by the contractors.

Approach guardrail transitions (AGTs) are commonly used to safeguard rigid hazards, including bridge railings and concrete parapets. AGT installations provide a safe transition in lateral stiffness between semi-rigid guardrail and rigid bridge rail installations. For many state DOTs, intersecting roadways or other roadside obstacles create space constraints that limit the ability to effectively install an AGT within the desired area. Thus, a need exists to minimize the length of AGTs tangent to the roadway.

My research project will entail a full evaluation of previous AGT crash test data and flared guardrail studies in addition to simulation studies and full-scale crash testing, which will be used to identify and evaluate critical flare rates for AGT installations. The research project is comprised of two phases. During the first phase, an extensive review of AGT crash test reports and flared guardrail studies was performed to evaluate AGT performance and identify a critical AGT design for further evaluation. Computer simulation studies were then conducted to determine maximum flare rates for the identified critical AGT design. The second phase of the research effort will use full-scale crash testing to evaluate the critical flare rate identified by the Phase 1 study. A minimum of three full-scale crash tests will be conducted under the Manual for Assessing Safety Hardware (MASH) test level three criteria to evaluate the flared AGT installation. Finally, researchers will develop recommendations and guidance for flared thrie-beam transitions. If applicable, the system will be submitted to FHWA for an eligibility letter and Hardware Guide drawings will be prepared for the system.

Previous research suggests that practice in diverse environments during the pre-licensure phase is related to better driving performance in novice teen drivers. However, little research has examined the relationship between pre-licensure practice (PLP) and driver self-regulation (DSR) of specific driving environments. Objective: to examine the influence of PLP diversity on DSR over the first six months of licensure.

The sample included fifty-six licensed teen drivers (Mage = 16.17). Baseline data were collected within two weeks of licensure and at follow-ups 3- and 6-months post-licensure. At baseline, participants reported frequency of PLP in various environments with a parent; at each timepoint, participants reported driving exposure and DSR in specific environments. A Principal Component Analysis extracted factors from the PLP items. Two factors emerged, Complex Practice and Simple Practice. Longitudinal multi-level models examined impact of PLP diversity on DSR over time.

DSR decreased over time (p < .001). More Complex Practice = less total DSR at baseline (p < .001). More bad weather practice = less baseline regulation of bad weather driving (p < .001). More nighttime practice = less baseline regulation of driving at night (p < .001) and steeper decrease in night DSR longitudinally (p < .01).

Some types of practice may better prepare teens for real-world driving. Practice in complex environments was associated with less DSR at baseline. Direct relationships between practice in night and bad weather conditions and post-licensure DSR in those environments was found. Findings may inform teen driver interventions targeting PLP on DSR post-licensure.

The rapid growth of ridesourcing services in North America over the past ten years has introduced a new area of research. Since this area of research is constantly changing, the objective of this paper is to provide a comprehensive literature review of the latest research and summarize findings relating to ridesourcing users’ traveler behavior. In total, 44 studies were reviewed, and six main traveler-focused categories were identified: demographics; frequency and time of use; trip purpose; reason for using ridesourcing services; relationship between ridesourcing and other modes; and transportation system impacts. The results pertaining to demographics revealed that ridesourcing users are likely younger with higher incomes and education levels, are full-time students or employed, and live in urban areas. Most ridesourcing trips occur on weekends and at night, with the most common trip purpose being for social events. Common reasons for using ridesourcing were to avoid driving under the influence, parking difficulties, and faster travel and wait times. Ridesourcing was found to substitute for taxis and personal vehicles; however, the results were mixed for public transit. Some studies suggest that ridesourcing can increase both vehicle miles travelled and the number of vehicles on the road; however more research is needed in this area to have conclusive findings. As both the use of ridesourcing services and research involving ridesourcing continue to grow, it is important to understand the trends of who is using these services and how travel behavior might be changing during this period of expansion.

People with low income generally spend a much larger proportion of their income on transportation than those who generate higher incomes, and some may even find the cost of owning a car to be infeasible. People without access to a personal vehicle are often reliant on pay-per-use modes, such as public transit, to get to their place of employment. Creating transit routes that connect residents of low-income neighborhoods to jobs that pay higher incomes is something that transit agencies can do to help elevate the socioeconomic statuses of people from these neighborhoods. Indianapolis, Indiana is an example of a transit-equipped, metropolitan city with pockets of poverty that have prevailed for years. In 2017, one quarter of Indianapolis’ residents lived in either low-income neighborhoods or concentrated poverty. This study seeks to assess the extent to which Indianapolis’ transit system is advantageous for providing low-income people access to better employment opportunities. This is done by utilizing ArcGIS tools to examine the proximity of financially sustainable jobs to bus stops along routes which circulate near low-income and impoverished Indianapolis neighborhoods. Financially sustainable jobs are classified as those which pay their employees' wages, which meet the Indiana Institute of Working Families’ self-sufficiency standards. Findings from this study can provide transit agencies with additional equity-related criteria to meet when creating and modifying transit routes. Long-term effects of implementing these criteria include mitigating and decentralizing poverty, especially in areas that have been historically plagued by poverty.

Through an internship with the City of Oakland’s Planning and Building Department, I am conducting preliminary exploratory research on zoning protocols to reduce proximity and exposure to air pollution from transportation. This research must balance the need to support businesses to maintain a strong industrial economic base with the need to protect the health of nearby residential communities from the impacts of the emissions from said businesses. The air pollution sources under City jurisdiction are primarily mobile sources from permitted businesses that generate transportation activities within its industrially zoned land. These types of businesses are also called magnet sources. We hope to create a protocol the City can use to (1) assess and estimate the impact of a business's air pollution contribution from their magnet source and (2) inform and improve City policy on its conditional use permitting process and existing non-conforming magnet sources. Our protocol will be informed by inventories of traffic volumes from local streets and freeways, surveys of truck trip volumes in a sample case study neighborhood, and truck trip estimates for industrial businesses based on academic models. This effort will include collaboration with the Bay Area Air Quality Management District’s air modeling team to set pollution thresholds and understand modeled air pollution estimates per business and land-use type.

In April, 2020, one month into COVID-19 lockdown, the San Francisco Municipal Transportation Agency (SFMTA) announced its new Slow Streets program. This emergency response closed select city streets to thru traffic, providing more space to physically distance for those who want to travel by foot, bike, wheelchair and other modes. Since then, the city has added even more streets for a total of 24 implemented corridors in the program.

Moving forward, the city now has to decide how to build a Slow Streets program that will be sustainable in the long term. To help with this plan, SFMTA issued a public questionnaire that asks where respondents live, what their opinion is of Slow Streets and if they recommend certain corridors for future Slow Streets. This capstone project will set out to analyze citizen responses to interpret the project’s effectiveness in communicating to San Francisco residents. The research will primarily focus on two research questions: What are San Franciscans’ perceptions of Slow Streets and do they understand how the program works? How are citizen responses different based on socio-demographic data in neighborhoods? Based on this analysis, the capstone will recommend future steps SFMTA can take in communicating about Slow Streets to San Franciscans.

The curb is a monopoly good; network infrastructure owned entirely by cities. It differs from most  network infrastructure, however, in not being managed by prices. Cities allocate curb space with a confusing patchwork of metering, time limiting, permitting, and—perhaps most notably—fines. Urban parking stands out among network infrastructure for its low quality, low payment, and the large share of its revenue derived from punishment. I develop a theory of why this is so, which emphasizes the curb’s public ownership and low production costs. I then use both qualitative and quantitative methods to examine some of the theory’s implications. I draw on descriptive data from a variety of US cities, but focus primarily on the city of Los Angeles.  I show that in a politicized parking environment, city officials are more likely to exempt affluent neighborhoods from payment, and more likely to levy fines on poorer neighborhoods.

The Every Student Succeeds Act (2015) gave foster youth additional legal protections in school, including the right to transportation and the right to remain at their school despite any moves, similar to protections already in place for students experiencing homelessness and students with disabilities.  California’s compliance with this mandate was relatively more difficult than other states’, as only eight percent of students in California travel by school bus, compared with 35 percent nationally.  Thus, California schools could not simply tap into these existing services to provide transportation for foster youth.

Ridehailing offers a solution to this gap.  HopSkipDrive, a ridehailing company designed to transport children, engages in contracts with school districts and counties to provide school transportation for these vulnerable student populations.  In 2018–2019, HopSkipDrive provided 32,796 trips to school in Los Angeles County, with massive time savings over the logical alterative: transit.  HopSkipDrive offers time savings of nearly 70 percent compared with the same trips simulated on transit.  HopSkipDrive’s trips average 28 minutes in duration, yet on transit only 30 percent would have taken less than 45 minutes.  This is despite 90 percent of all origins and destinations being located within a half-mile of a transit stop.  This service has important social equity implications beyond just time savings offered to vulnerable student populations, as HopSkipDrive contract trips tend to originate in neighborhoods with high percentages of low-income households and people of color.

We study the linked tactical design problems of fleet sizing and partitioning a service region into vehicle routing zones for same-day delivery (SDD) systems. Existing SDD studies generally focus on operational problems, while limited work on SDD tactical design does not consider a priori customer partitioning or clustering, which have been shown to improve the efficiency of vehicle routes in other logistics contexts. Using continuous approximations to capture average-case system behavior, we consider first the problem of independently maximizing the area of a single-vehicle zone to minimize fleet size. We characterize area-maximizing dispatching policies and leverage these results to develop a procedure for calculating optimal areas as a function of a zone's distance from the depot. We demonstrate how to derive fleet sizes from optimal area functions and propose an associated Voronoi partitioning approach. We test our approach in a computational study that examines on two different service regions, demonstrating efficacy in terms of prediction accuracy and robustness.

In recent decades, studies and interventions in transportation equity have attempted to measure and improve the extent to which the U.S. transportation system serves and includes diverse groups, such as women, the elderly, and ethnic and racial minorities. However, rather less study and intervention has focused on the diversity of the transportation system itself, including: diversity of modes; diversity of trip purposes; diversity of ways of experiencing and enacting transportation; and diversity of ways of valuing, understanding and relating to spatial mobility. In this presentation, I contrast the relative homogeneity of transportation in the U.S. today with the relative diversities of transportation in other times and places, while also highlighting minor mobility subcultures of the contemporary U.S. I critically analyze barriers to transportation diversity and discuss a politics of possibility for helping diverse mobility cultures to flourish.

Access to opportunities measures used in planning practice and research assume that pedestrians can complete public transit trips and traverse the first and last mile without difficulty. But physical accessibility depends on characteristics of the built environment and requires a full, unbroken path between origin and destination. Due to limitations associated with data availability and scale, existing access to opportunities measures generally do not include characteristics like curb cuts and obstructions even though these features directly affect physical accessibility. For people with disabilities, these characteristics are extremely important because they rely on public transit for their everyday mobility and participation in society.
 
My research uses mixed methods to understand how first and last mile challenges prevent people with disabilities from effectively reaching destinations using public transit. I employ a synthetic population along with detailed sidewalk data to understand how access to opportunities varies for people with disabilities at a fine scale in Seattle, Washington. This work demonstrates where sidewalk investments would have the greatest impact and recommends mitigating policies and strategies that transportation agencies can implement to enhance conditions for people with disabilities.

In the United States, transit is largely a lifeline social service for low-income people (Taylor & Morris, 2015). This function has eroded, however, to the extent that low-income people have been displaced from transit-rich areas (Bardaka, Delgado, & Florax, 2018). We test this idea using data from Southern California. We do so in two parts. First, we match changes in Census tract-level rents and housing burdens to changes in  transit boardings, and show that across two time periods—2008-2012 and 2013-2017—as housing prices rise boardings fall.  This finding is robust to a broad array of controls. It only suggests, however, that neighborhood change is associated with falling neighborhood ridership. It remains possible that lower-income people no longer living in high-transit neighborhoods ride just as much once they move elsewhere. To address this possibility, in the  second part of our analysis we will use migration data to follow people who left Census tracts with large ridership losses disproportionately located in lower-density census tracts with fewer transit stations and lower transit accessibility. If our hypothesis holds, these people relocated to tracts with comparatively poor transit infrastructure that discouraged transit use.

Bus ridership has decreased in major cities across the United States: it is currently at its lowest level in thirty years. One of the few exceptions is Seattle, where bus ridership has seen consistent increases the past few years. The goal of this research is to assess what factors have led to Seattle’s success with buses, and through that case to provide insight into how cities can prioritize transit in a way that can attract new riders, change perceptions of the bus, and compete with other modes. The findings will inform debate on the sources of transit ridership growth and decline, and lessons for embattled transit systems. This has significant implications for sustainability, congestion, and the financial health of transit systems and their ability to provide service for populations depending on them.

Habit is a key determinant of travel behavior, where people who have regularly chosen a mode in the past are more likely to use it in the future (Schneider 2013, Verplanken et al. 2014, et cetera). But mode choice habits can be interrupted when people experience significant life changes (Bamberg et al. 2003, Janke & Handy 2019). The abrupt and long-term workplace closures affected by the COVID-19 pandemic have potential to cause precisely the type of significant life change that may interrupt routine mode choices. The extent that pandemic-era workplace closures cause long-term travel behavior changes will be the subject the study for the next decade and beyond, but we can already see the resultant short- and medium-term changes in mode choice. We use in this study both a panel and repeat cross-sectional survey to model changes in mode share among a university population (undergraduates, graduate students, faculty, and staff) before and during the 2020 campus closure due to COVID-19. This university – University of California, Davis – is a particularly bicycle-friendly campus for the United States; it had a high pre-pandemic bicycle and transit mode shares and a low vehicle mode share relative to state- and nationwide patterns (34% bicycled to campus, 20% rode transit, and 31% drove alone in the fall of 2019). In the second wave of our panel survey, administered in spring 2020, the share of people commuting by single-occupancy vehicles had nearly doubled from the fall (57% drove alone) and in the third wave, administered in fall 2020, 41% of respondents drove alone. Many of the “new” drivers were people who had regularly bicycled or taken transit before the pandemic, and we see commensurate drops in bicycle and transit mode shares. Drawing from travel behavior theory, we explore and estimate the influence of factors such as habit, attitudes, perceived safety, and enjoyment on the event of switching routine modes. If these mode switches to single-occupancy vehicles are sustained and widespread, such trends will contravene local, state, and national policy goals to reduce driving in the interests of improving air quality, mitigating climate change, and easing parking demand.

Introduction: The onset of the COVID-19 pandemic brought sudden and drastic changes to daily life. Several trends converged that influenced changes in travel behavior including loss of income and benefits from widespread layoffs or reductions in work hours, and reduced transit capacity due to service cuts or boarding restrictions. Healthcare access was also impacted. Staff at overwhelmed hospitals were faced with diverting patients or delaying transfers. Even healthcare facilities with minimal COVID-19 patient loads postponed elective care. Overall, hospitals and other facilities tightened restrictions on patients’ visitor/guest policies. This qualitative study explores how the COVID pandemic impacted transportation access to healthcare and the innovative solutions used to address those barriers—from the perspective of care coordinators (social workers, nurses and other professionals in the medical field who support patients in utilizing medical care, including addressing transportation issues).

Methods: Fifteen in-depth interviews were conducted in October and November of 2020 with care coordinators employed at hospitals in two regions in North Carolina that serve patients in their respective regions and around the state, as well as treatment centers and social service agencies.

Results/Conclusion (TBD): This project will identify transportation barriers to accessing healthcare experienced during the COVID-19 pandemic, from a cross-disciplinary perspective. Moreover, the study will examine the role of emerging virtual services (e.g., telehealth, etc.) during the pandemic to address reduced access to healthcare. 

The COVID-19 pandemic disrupted typical travel patterns across the United States. Governmental responses to the pandemic have varied from state to state, ranging from stay-at-home orders to non-essential business closures to, in some cases, nearly no requirements. In addition, authorities dispersed public health messaging to help people make educated choices about their behaviors and travel. The “flatten the curve” campaign, for example, was a wide-spread health education effort; the message encouraged people to take immediate action to quell the spread of the disease so that the number of cases did not exceed local hospitals’ capacities to care for COVID-19 patients. Both emergency governmental action and public health education campaigns have changed trip making behaviors.

Many solutions have been presented to assist connected and automated vehicles (CAVs) interpret the geometries of roadways through optical means, such as cameras or lidar point cloud maps, or external augmented global navigation satellite system (GNSS) positioning systems. A new method of representing road data based on constitutive segment relationships and GNSS-based anchor points is proposed, called the Midwest Virtual Road Corridor (MVRC). Critical points of roadways, such as the left-side edge boundary per travel direction, are extracted along the entire length of the roadway. Consecutive points of the road are used to develop orthogonal curvature vectors, road tangent vectors, and continuity relationships based on the Midwest Discrete Curvature (MDC) method. In this study, road data was smoothed, and characterized to identify unique road curvature profiles as either curved or straight from any general curve geospatial dataset. This technique was evaluated by discretizing three testing road sections from Google Earth: A general highway curve consisting of multiple spiral transitions, straight, and constant radius curves. A constant radius curve and a highway interchange. Lastly, a matrix is constructed by combining the road segment anchor points, road tangent vectors, curvature vectors, and supplementary road information that can provide an ultra-compact, high-fidelity, and precise virtual road corridor for CAVs. The MVRC data matrix can be implemented to assist many of the existing and in-development CAV guidance solutions.

Determining the location and severity of sinkhole hazards through geospatial methods is fundamental to predict the risk of sinkhole occurrence by developing a susceptibility model that is objective and repeatable. Sinkhole formations are associated with both the geomorphology and changes in the hydrological conditions. Hence, this research investigates the use of satellite-based temperature data, precipitation data, volumetric moisture content data, and the national soil survey data for reliable sinkhole nowcasting. First, historical time series data of conditional factors contributing to the triggering of sinkholes was assembled from high-quality satellite-based precipitation data such as the National Aeronautics and Space Administration (NASA) Earth and Precipitation Satellite Missions; meteorological conditions extracted from the Moderate Resolution Imaging Spectroradiometer (MODIS), Daily Surface Weather and Climatological Summaries (Daymet V3),  and the Global Precipitation Measurement (GPM) mission with Integrated Multi-satellitE Retrievals (IMERG) for GPM.  This data was then combined with the spatial and temporal hydrological changes in the ratio of water to soil volume (volumetric water content) from surface water to groundwater recharge (soil moisture data) that was acquired from the NASA Soil Moisture Active Passive (SMAP) Earth satellite mission. Lastly, geological data were collected from the soil survey performed by the USDA National Resource Conservation Service (NRCS) as collapse tends to be in areas with underlain carbonate (limestone and dolomite) and evaporate (salt, gypsum, and anhydrite) rocks.  The study also examined the aggregation of satellite data to downscale resolution errors. 

Emergent mobility technologies, such as UAM, must be understood in the context of urban transportation networks in order to better inform policy and infrastructure decisions for cities. When evaluating the impact of new modes and evaluating future infrastructure provision, cities typically assess the intent, need, possibilities, judgment, tradeoffs, and investment costs, and simulations can often inform this analysis. However, there has yet to be a comprehensive simulator of the short- and long-term effects of UAM on urban environments. This metropolitan-scale transportation and land-use simulation will leverage powerful parallel computing architectures, advanced optimization architectures for network design, and extensive multinomial logit modeling for both activity generation and long-term decision-making processes.

This project develops an integrated simulation pipeline for air and ground transportation, to enable door-to-door simulations of multi-modal (ground and air) transportation. This integrated simulation capability will be used to support an optimization-based approach for multi-modal transportation, in order to enable optimized design and placement of a network of vertiports relative to a number of measures, including cost, capacity, and travel time. These outputs are then incorporated into long-term land use models to understand the evolution of cities over time.

Near-completed work includes the development and coupling of both the ground and air microsimulators, with aerial expertise provided by NASA. In addition, advanced optimization and clustering techniques, travel time analyses, and vertiport capacity design have also been conducted. Ongoing work includes calculating mode choice of UAM, and incorporating the aforementioned simulator inputs and outputs with long-term land use simulation. 

Community colleges (CCs) continue to serve as an important ladder of opportunity for residents in the United States. These opportunities include, but are not limited to, gaining job training skills, acquiring educational credits in advance of transferring to a four-year university, and earning an equivalent of a high school diploma. Travel to CCs is primarily served by private automobiles as most CCs do not provide on-campus housing for students. This could serve as a barrier to students initially enrolling in CCs, establishing successive enrollment at CCs, and achieving academic success at CCs. This research poster provides a breakdown of the literature related to transport to and from CCs, existing literature gaps, and a research plan for investigating the relationship between CC transport access and egress and academic achievement at CCs. Additionally, the variation in this relationship across the significantly heterogeneous student populations often found at CCs will be explored.