Search

This research presents a study of the opportunities for lift-sharing between families with young children in the specific context of children's activities. The research identifies the key conditions that support lift-sharing in this context. Study 1 utilised travel pattern diaries and interviews on reasons for mode choice (with 13 families) to identify opportunities to decrease frequency of low occupancy car trips in families with young children. Frequency of short and/or low occupancy car trips was high. The potential for lift-sharing for children's activities was identified. There was a lack of existing research on factors influencing lift-sharing of this type; consequently Study 2 and Study 3 aimed to explore lift-sharing for children's activities. The online travel survey of 474 families in Study 2 assessed lift-share prevalence to establish socio-demographic and activity factors that influence lift-share prevalence, in general and for children's activities. These influential factors include: number of cars owned, number of seats in car, settlement type, income level, time of day and location of the activity, number of children attending, parking availability and number of close friends of the child attendee. Study 3 comprised interviews with 15 parents to gain a deeper understanding of how and why families participate in activity lift-sharing, and to understand how social exchange theory applies in this context. Key findings included time saving ability and social opportunity as salient motivating factors, trust being an initial imperative component of lift-share formation, variation in attitudes towards reciprocity, parent utilisation of synchronous and asynchronous communication channels being dependent on time gap to lift-share responsibility and recognition of the advantageous group communication affordances of instant group messaging applications. Five fundamental themes were identified across the results of the three studies. These included influential ‘household factors' and influential ‘activity factors,' and the concepts of and implications of ‘time,' ‘relationships,' and ‘technology and communications' that are fundamental to activity lift-share operation and formation. The knowledge gained on these topics has taken forward understanding of how lift-sharing between families with young children operates, and our understanding of the opportunities for children's activity lift-sharing.

Understanding travel mode choice behaviour is key to effective management of transport networks, many of which are under increasing strain from rising travel demand. Conventional approaches to simulating mode choice typically make use of behavioural models either derived from stated preference choice experiments or calibrated to observed average mode shares. Whilst these models have played and continue to play a key role in economic, social, and environmental assessments of transport investments, there is growing need to gain a deeper understanding of how people interact with transport services, through exploiting available but fragmented data on passenger movements and transport networks. This thesis contributes to this need through developing a novel approach for urban mode choice prediction and applying it to historical trip records in the Greater London area. The new approach consists of two parts: (i) a data generation framework which combines multiple data-sources to build trip datasets containing the likely mode-alternative options faced by a passenger at the time of travel, and (ii) a modelling framework which makes use of these datasets to fit, optimise, validate, and select mode choice classifiers. This approach is used to compare the relative predictive performance of a complete suite of Machine Learning (ML) classification algorithms, as well as traditional utility-based choice models. Furthermore, a new assisted specification approach, where a fitted ML classifier is used to inform the utility function structure in a utility-based choice model, is then explored. The results identify three key findings. Firstly, the Gradient Boosting Decision Trees (GBDT) model is the highest performing classifier for this task. Secondly, the relative differences in predictive performance between classifiers are far smaller than has been suggested by previous research. In particular, there is a much smaller performance gap identified between Random Utility Models (RUMs) and ML classifiers. Finally, the assisted specification approach is successful in using the structure of a fitted ML classifier to improve the performance of a RUM. The resulting model achieves significantly better performance than all but the GBDT ML classifier, whilst maintaining a robust, interpretable behavioural model.

Millions of children travel to and from school each day as part of their daily routine. A large percentage of children make this journey by car, and the numbers are steadily rising and this is leading to many environmental and health implications for children. The current economic climate has persuaded the British Government to look again at policies relating to all school travel funding to highlight areas where savings and cuts can be made. This is interesting because the home-to-school transport provision policy has been in place since the Education Act 1944 and this policy costs local authorities in England over £1 billion a year. Therefore, the focus of this thesis is threefold. Firstly, it seeks to determine the main issues within school travel and reports on the views of current professionals in the school travel industry. Structured in-depth interviews were carried out with 16 UK and US school travel experts. The questions focused on the current stakeholders of school travel, issues regarding school travel, bus use in school travel, and the challenges faced by transport planners to ensure school pupils have a safe and pleasant journey to school. Secondly this thesis quantifies the traffic and environmental impacts of the school choice policy in England. It achieves this by analysing School Census data from 2009 from the Department for Education. Multinomial logit modelling and mixed multinomial logit modelling are used to illustrate the current travel behaviour of English children in their journey to school and examine how there can be a significant reduction in vehicle miles travelled, CO2 emissions and fuel consumption if the school choice policy is removed. The results suggest that if all children attended their nearest school, this would result in reductions in their personal mobility, vehicle miles travelled and CO2 emissions. Finally, this thesis examines the policies relating to the funding criteria of home-to-school public school transport provision. Specifically, the paper employs a multilevel modelling technique to develop a series of relationships between bus usage by school and the level of spending by local education authorities on home-to-school bus travel provision while controlling for other factors such as school quality, land-use patterns and various proxies for household incomes. The results suggest that there is a significant effect of funding on the total school-level bus passenger mileage for primary (aged less than 11), secondary (aged 11 to 16) and Post 16 schools.

People’s perceptions, experiences, and personal attributes influence their daily travel and mode choices. Existing accessibility measures, an estimate of people’s potential ability to reach places, cannot capture these perceptual aspects of travel and its differences across individuals and groups, resulting in overestimation. Moreover, existing literature evaluates accessibility as either person-based (potential mobility of an individual) or place-based (potential mobility for a group residing over a large geographic space), with no specific guidelines for transitioning from one estimate to another. Although person-based estimates can effectively capture heterogeneity in travel perceptions, the findings cannot be converged for larger populations and applied in urban planning practices. On the other hand, place-based measures generalize the estimates for the overall population, assuming that the current travel environment provides equal access to all regardless of their socio-economic identities. Most importantly, both place-based and person-based measures are physicalist in nature and disregard the mobility barriers that travelers experience during their day-to-day travel as individuals or as part of their social cohorts.This dissertation develops a novel concept and method called inclusive accessibility that captures both physical and perceptual barriers of travel across individuals and socio-economic groups in evaluating their access to multimodal transportation systems. This study develops a systematic bottom-up approach that estimates inclusive accessibility for an individual first, aggregates the measure for socio-economic groups, and demonstrates the application of this concept in developing need-specific and equity-oriented infrastructure planning. We conceptualize inclusive accessibility as a subset of classic access considering hard constraints (physical space-time limitations of travel) and soft constraints (perceptions of safety comfort toward land use-transportation environment and mode-specific time preferences). Therefore, inclusive accessibility outlines the geographic space that is physically and perceptually accessible to individuals and groups. This dissertation presents a method for operationalizing the bottom-up inclusive access concept using a unique mobility survey dataset from Columbus, OH. The first and second studies estimate inclusive access for a person and mutually-exclusive social cohorts. In both studies, we apply a support vector regressor (SVR) and its ensemble on the survey dataset to predict walking perception score as soft constraints for individuals and groups across Columbus roads. We derive inclusive accessibility measures as STP constructs that consider space-time limitations from the transit and sidewalk network as hard constraints and minimize the individual or group-specific soft constraints to a certain threshold. Lastly, the third study demonstrates a multi-objective optimization method to design potential future infrastructure installation projects considering the soft constraint estimates of inclusive access and evaluate their equity impacts. The study reveals that the accessible geographic space for both individuals and groups shrinks substantially as we shift from classic to inclusive access, including travel perceptions and time preferences. The study also finds that perceptions of safety and comfort and its resultant inclusive access vary across individuals and socio-economic groups, highlighting income and racial disparities. The existing infrastructure and services favor high-income white populations, a representation of privileged populations over other communities. Meanwhile, low-income communities perceive higher mobility barriers while walking in Columbus due to their greater exposure to a poor-quality travel environment and risks. These findings emphasize that existing infrastructure and services yield unequal levels of access across communities, which incites the need for equity-oriented infrastructure planning. Building upon these insights, our example of designing sidewalk installation projects and its equity impacts demonstrates that even though installing new sidewalks improves perceptions of safety and comfort for all, the level of improvements varies across socio-economic communities depending on the locations of sidewalk installations.

Many large cities face significant challenges in providing safe, economical, equitable, reliable, and efficient mobility services. Moreover, mobility service options have been changing unprecedentedly due to the advent of emerging mobility services. Understanding the impacts of emerging mobility services (e.g., for-hire vehicles, bike sharing) on other travel modes (e.g., transit, non-motorized modes, personal car) and evolving travel behavior is critical to promote sustainable modes for urban residents and reduce auto-dependency. New York City (NYC) has been conducting annual mobility survey since 2017 to understand the emerging travel behavior and mode choice characteristics. This dataset presents a unique opportunity for conducting research on the effects of emerging mobility services and the mode choice behavior of urban residents. This dissertation investigates emerging travel behavior in NYC to promote sustainable mobility and reduce auto-dependency. Three specific research objectives are to- i) identify factors that influence the choice of sustainable and non-sustainable modes, ii) identify factors affecting mode shift behavior of city residents between transit and other modes, and iii) investigate the impacts of for-hire-vehicles (FHVs) and bike sharing services on transit ridership. Discrete choice modeling approach was adopted to identify factors influencing sustainable and non-sustainable mode choice. NYC residents with positive attitudes toward sustainable modes were more likely to use sustainable modes. Sustainable modes usage decreased with personal vehicle ownership. People with a higher level of education showed lower usage of non-sustainable modes. Increasing ADA compatibility of sustainable travel modes and associated transportation infrastructure could increase sustainable mode use among people with a disability. Residents of four boroughs in NYC (i.e., Brooklyn, Queens, Bronx, and Staten Island) were found to use non-sustainable modes significantly higher compared to Manhattan residents. Borough-specific targeted initiatives could increase sustainable mode attractiveness in all NYC boroughs. Negative attitudes towards transit services were significant factors in mode shift. Improving transit service reliability, enhancing comfort and safety, and increasing accessibility could improve users’ attitudes toward transit services. Accessibility of personal vehicles increased the mode shift from transit use. People with higher education were less likely to shift from transit. In addition, emerging modes significantly influenced the mode shift from transit. Negative binomial (NB) models with the Bayesian estimation technique were developed to study emerging mobility services' impacts on transit ridership. The number of FHV end trips within a subway station (quarter-mile distance) was found positively associated to higher transit ridership. FHV trips ending within subway station area act as first-mile connections and increased transit ridership. Marking designated areas for FHV trip pick-up and drop-off near all transit stations could improve the first and last-mile connection to transit and help increase the transit ridership. Citi Bike trips ending within a subway station service area did not significantly affect transit ridership. Bike sharing network should be well-connected to subway stations to encourage more people use the bike-sharing system to access the subway stations. By increasing the quantity of DOT bike racks near subway stations, NYC could develop a bike-supportive transportation system. This dissertation advanced the understanding of emerging travel behavior among residents in NYC and the impacts of emerging mobility services on transit ridership. Transportation agencies and policymakers can develop critical policies and investment decisions based on the identified significant factors to promote sustainable modes and reduce auto-dependency. Moreover, the research findings can be used to create a safe and sustainable mode-friendly environment and improve transit services that could increase positive attitudes/perceptions among city residents towards sustainable modes and increase transit ridership.

The commercial deployment of automated vehicles (AVs) is around the corner. With the development of automation technology, automobile and IT companies have started to test automated vehicles. Waymo, an automated driving technology development company, has recently opened the self-driving service to the public. The advancement in this emerging mobility option also drives transportation reasearchers and urban planners to conduct automated vehicle-related research, especially to gain insights on the impact of automated vehicles (AVs) in order to inform policymaking. However, the variation with urban form, the heterogeneity of mode choice, and the impacts at disaggregated levels lead to the dynamics of the impacts of AVs, which not comprehensively understood yet. Therefore, this dissertation extends existing knowledge base by understanding the dynamics of the impacts from three perspectives: (1) examining the role of urban form in the performance of SAV systems; (2) exploring the heterogeneity of AV mode choices across regions; and (3) investigating the distribution of energy consumption in the era of AVs. To examine the first aspect, Shared AV (SAV) systems are simulated for 286 cities and the simulation outcomes are regressed on urban form variables that measure density, diversity, and design. It is suggested that the compact development, a multi-core city pattern, high level of diversity, as well as more pedestrian-oriented networks can promote the performance of SAVs measured using service efficiency, trip pooling success rate, and extra VMT generation. The AV mode choice behaviors of private conventional vehicle (PCV) users in Seattle and Knasas City metropolitan areas are examined using an interpretable machine learning framework based on an AV mode choice survey. It is suggested that attitudes and trip and mode-specific attributes are the most predictive. Positive attitudes can promote the adoption of PAVs. Longer PAV in-vehicle time encourages the residents to keep the PCVs. Longer walking distance promotes the usage of SAVs. In addition, the effects of in-vehicle time and walking distance vary across the two examined regions due to distinct urban form, transportation infrustructure and cultural backgrounds. Kansas City residents can tolerate shorter walking distance before switching to SAV choices due to the car-oriented environment while Seattle residents are more sensitive to in-vehicle travel time because of the local congestion levels. The final part of the dissertation examines the demand for energy of AVs at disaggregated levels incorporating heterogeneity of AV mode choices. A three-step framework is employed including the prediction of mode choice, the determination of vehicle trajectories, and the estimation of the demand for energy. It is suggested that the AV scenario can generate -0.36% to 2.91% extra emissions and consume 2.9% more energy if gasoline is used. The revealed distribution of traffic volume suggests that the demand for charging is concentrated around the downtown areas and on highways if AVs consume electricity. In summary, the dissertation demonstrates that there is a dynamics with regard to the impacts and performance of AVs across regions due to various urban form, infrastructure and cultural environment, and the spatial heterogeneity within cities.

The dissertation's central question examined the impact of the built environment (BE) on people's trip-chain travel behavior. In contrast to trip-based research, which presumes an isolated, homogenous, single-leveled, and static relationship between BE and travel behavior, this collection of studies analyzes the interaction of a succession of decision-making processes with varying spatial limitations at different dimensions. This dissertation began by establishing a conceptual framework for evaluating the relationship between the built environment and trip-chain travel behavior. The research next evaluated the multidimensional BE impacts on the percentage of vehicle miles traveled (PVMT) and two critical trip-chain behaviors: trip-chain complexity and trip-chain mode choice. According to the tour complexity analysis, all non-auto passengers are opposed to increased tour complexity. Overall, only activity availability and density near the job and neighborhood housing type and density had a substantial effect on the BE factors. In comparison to tour complexity, the findings of the tour-level mode choice study indicate that BE characteristics are more influential in deciding mode choice. Among all BE characteristics, walkability (bike-ability) and street connection at the job and at home have the most impact. Additionally, the decision-making process for secondary mode selection is considered. The model findings suggest that the secondary mode choice is mostly determined by the primary mode choice, whereas BE characteristics close to sub-activity sites have a little effect. Only two walkability attributes have a significant influence on secondary mode choice when compared to all BE attributes close to sub-activity sites. Finally, the interdependence is evaluated by examining the relationship between tour complexity and tour-level mode selection. When the joint impacts are considered, the single-family ratio in the home zone has a substantial marginal influence on mode choice. The findings indicate that the single-family ratio in the home zone increases the probability of a drive-alone trip but decreases the probability of transit, bike, or walking tour.

Although cargo airships could potentially provide greater efficiency and flexibility in freight transportation and reduce demand on infrastructure, currently they remain a theoretical concept. This research aims to examine the factors determining the demand for their services from the perspective of transport logistics professionals. Specifically, the research examines the effects of freight type and freight modal attributes on freight mode choice decisions. As cargo airships is currently a hypothetical mode, an online freight mode choice experiment is conducted and the results analysed using a multinominal logit choice model. This analysis is explored further with the inclusion of qualitative data gathered through follow-up semi-structured interviews. The results of the online freight mode choice experiment established the potential demand for the cargo airship mode to be up to 27% of market share. The factor that provides the greatest variation in the choice model results is the different freight types, which include perishables, non perishables, oversize overmass (OSOM), and high value time sensitive (HVTS). Across these freight types the sensitivity of the modes (road, rail, sea and airship) to the variable attributes (price, time, reliability and frequency) varied substantially. Additionally, the willingness to pay (WTP) for the cargo airship mode for travel time saved is estimated to be $23.15 per tonne per hour. Simulations are further conducted from the choice model results to ascertain the potential market share of various different model cargo airships, including the exemplar model (ML868) by adjusting their speed (time) and cost (price) to reflect their predicted actual attribute levels. From these simulations it is established that there is potential demand in the Australian domestic freight market with the LMH-1 having the lowest market share of 7.56% and the ARH50 having the highest market share of 24.90%.

The field of transportation is on the cusp of major change. Innovations in how vehicles operate and are powered have the potential to elicit changes not seen since the introduction of the interstate highway system more than half a century ago. Predicting the impacts of new vehicle technologies has interested researchers and practitioners across disciplines and continents. This thesis makes a handful of such predictions. It is divided into three parts. In the first part, the results of two large-scale preference surveys and data from the U.S. Environmental Protection Agency’s (EPA) Smart Location Database (EPA, 2014) are used to estimate how land use characteristics impact Americans’ perceptions of, interest in, and willingness to pay for new vehicle technologies, while controlling for demographic attributes. The surveys were conducted by Quarles and Kockelman (2018) and Gurumurthy and Kockelman (2018) in 2017 and together represented over 4,000 U.S. households. Statistical models like the ordered probit and multinomial logit are used to estimate the impacts of demographics and land use characteristics on vehicle-related behavior. Various land use variables arise as significant depending on the question being asked of the respondents. For example, poor job accessibility via automobile is associated with higher levels of interest in automated vehicles (AVs), higher anticipated use of AV technology, a willingness-to-pay (WTP) for self-driving capability, and a greater reliance on AVs for some long-distance travel. No land use variable arises as significantly more predictive than the others at this national-level scale of analysis. The results emphasize the fact that land use policy must be considered at the local level, and that there is no “one size fits all” solution for managing future transportation behavior with land use action. The second part of this thesis evaluates the connection between land use and current travel behavior. Census tract-level measures of population and employment density (provided once again by the EPA’s Smart Location Database [EPA, 2014]) are evaluated across the nation to investigate the connection between the development conditions one experiences and his or her travel behavior. Travel data comes from the 2009 National Household Travel Survey (NHTS). The results highlight a stronger connection between population density and vehicle-miles traveled (VMT) and vehicle ownership, than with employment density. For both VMT and vehicle ownership, an improvement of only two to one can be expected by changing population density conditions in a census tract. In other words increasing population from the lowest density conditions to the highest results in a decline of VMT per capita per day from 20 miles to 10 miles. Similarly, vehicle ownership per capita generally ranges from 0.4 to 0.8. Notably, these improvements are not realized until the highest decile of population density (18 people per acre), thus indicating that simply building homes in rural or low-density suburban regions will likely have a negligible impact on transportation demand. Employment density was found to be less indicative of travel behavior. The third and final piece of the thesis predicts how an evolving light-duty vehicle (LDV) fleet will impact the amount of energy consumed by Americans and the emissions they create. Here, the results of a fleet evolution simulation, developed by Quarles et al. (2019), are used to project what a vehicle fleet with more electric (and fewer gasoline-powered) vehicles will mean for energy consumption and emissions on a per capita basis. Projections are based on historic fuel efficiency data and emission production rates from the Bureau of Transportation Statistics (BTS) and EPA (BTS, 2018b; BTS, 2018c; EPA, 2018a). Conclusions from these findings highlight the need for more efficient vehicles, better emissions control technologies on existing vehicle models and power plants, and a decreased reliance on highly-polluting energy sources for power generation. Policies aimed at achieving these objectives will help ensure that Americans’ future vehicle behavior and ownership will not create an undue burden on themselves or the environment in which they live. Although the analyses discussed in this thesis cover diverse topics such as human behavior, urban planning, and air quality, they establish the need for a proactive approach to cutting-edge vehicle technologies. If left to develop without any oversight or action, transportation network congestion will worsen, development will continue to sprawl, and the environment and public health will suffer. Policies aimed at limiting “empty” driving with AVs, increasing population density, and curbing vehicle and power plant emissions can help ensure the benefits of vehicle technology innovation are not realized at the expense of other considerations.

This thesis covers certain aspects of autonomous and shared autonomous vehicles (S/AVs), with a focus on dynamic ride-sharing (DRS). The first part investigates Americans’ preferences in adopting AVs. Rapid advances in technologies have accelerated the timeline for public use of fully-automated and communications-connected vehicles. Public opinion on self-driving vehicles or AVs is evolving rapidly, and many behavioral questions have not yet been addressed. This study emphasizes AV mode choices, including Americans’ willingness to pay (WTP) to ride with a stranger in a shared AV fleet vehicle on various trip types and the long-distance travel impacts of AVs. 2,588 complete responses to a stated-preference survey with 70 questions provide valuable insights on privacy concerns and crash ethics, safety and ride-sharing with strangers, long-distance travel and preferences for smarter vehicles and transport systems. While the starting sample data were relatively demographically unbiased, Texans were purposefully over-sampled, and all statistics adjusted/corrected (via sample weights) to match US demographics on gender, education, income, and age. Weighted results suggest that Americans are willing to pay, on average, $2073 to own AVs over conventional vehicles and an additional $1078 to maintain/include a manual driving option on such vehicles. Ride-sharing will be popular at 75¢ per mile, under most scenarios, and many Americans are willing to pay $1, on average, to anonymize their trip ends’ addresses. Most are also willing to let children 16 years of age and older have unsupervised access to AVs (both privately owned and shared). Nearly 50% of long-distance travel appears captured by AVs and SAVs in the future, rather than airlines, at least for one-way trip distances up to 500 miles. Two hurdle models (which allow for a high share of zero-value responses) were estimated: one to predict WTP to share a ride and another to determine WTP to anonymize location while using AVs. The first two-part model shows how travel time delays, person and household attributes, and land use densities can significantly affect Americans’ willingness to share rides. The second hurdle model suggests that traveler age, presence of children, household income, vehicle ownership and driver’s license status are major predictors of one’s WTP to obscure pick-up and drop-off locations. A binary logit was used to model current mode choice for long-distance (over 50 miles, one-way) travel (between one’s private car and an airplane), with household income as the leading predictor. On average, older Americans and/or those with children prefer such travel by car. Finally, a multinomial logit anticipated mode shifts when AVs and SAVs become available and affordable. Everything else constant, private cars remain preferred by older people, but SAVs may be used in the future for more business travel. In the second part of this thesis, a trip-matching framework is programmed to evaluate DRS opportunities for trips across Orlando, Florida. Transportation network companies (TNCs) are regularly demonstrating the economic and operational viability of DRS to any destination within a city, thanks to real-time information from smartphones. In the foreseeable future, fleets of SAVs may largely eliminate the need for human drivers, while lowering per-mile operating costs and increasing the convenience of travel. This may dramatically reduce private vehicle ownership and deliver extensive use of SAVs. Using AirSage’s cellphone-based trip tables across 1,267 zones over 30 consecutive days, this study anticipates DRS matches (by assigning independent travelers with overlapping routes in time and space to the same SAV) and simulates SAV travel across the Orlando network to determine optimal SAV fleet size. Those results suggest significant opportunities for DRS-enabled SAVs: Nearly 60% of the single-person trips can be shared with other persons traveling solo and with less than 5 minutes added travel time (to arrive at their destinations). This value climbs to 80% and 86% for 15 and 30 minutes of added wait or travel time, respectively. On the average travel day in Orlando, a fleet of just 30,000 SAVs can serve nearly 45% of those 3 million person-trips traveled solo. In other words, just 1 SAV per 100 person-trips is able to serve almost half of the region’s demand, helping reduce congestion while filling up passenger vehicle seats.

There are natural synergies between shared autonomous vehicle (AV) fleets and electric vehicle (EV) technology, since fleets of AVs resolve the practical limitations of today's non-autonomous EVs, including traveler range anxiety, access to charging infrastructure, and charging time management. Fleet-managed AVs relieve such concerns, managing range and charging activities based on real-time trip demand and established charging-station locations, as demonstrated in this paper. This work explores the management of a fleet of shared autonomous (battery-only) electric vehicles (SAEVs) in a regional (100-mile by 100-mile) discrete-time, agent-based model. The dissertation examines the operation of SAEVs under various vehicle range and charging infrastructure scenarios in a gridded city modeled roughly after the densities of Austin, Texas. Results indicate that fleet size is sensitive to battery recharge time and vehicle range, with each 80-mile range SAEV replacing 3.7 privately owned vehicles and each 200-mile range SAEV replacing 5.5 privately owned vehicles, under Level II (240-volt AC) charging. With Level III 480-volt DC fast-charging infrastructure in place, these ratios rise to 5.4 vehicles for the 80-mile range SAEV and 6.8 vehicles for the 200-mile range SAEV. However, due to the need to travel while "empty" for charging and passenger pickup, SAEV fleets are predicted to generate an additional 7.1 to 14.0% of travel miles. Financial analysis suggests that the combined cost of charging infrastructure, vehicle capital and maintenance, electricity, insurance, and registration for a fleet of SAEVs ranges from $0.42 to $0.49 per occupied mile traveled, which implies SAEV service can be offered at the equivalent per-mile cost of private vehicle ownership for low-mileage households, and thus be competitive with current manually-driven carsharing services and significantly less expensive than on-demand driver-operated transportation services. The mode share of SAEVs in the simulated mid-sized city is predicted to be between 14 and 39%, when competing against privately-owned, manually-driven vehicles and city bus service. This assumes SAEVs are priced between $0.75 and $1.00 per mile, which delivers significant net revenues to the fleet owner-operator, under all modeled scenarios, assuming 80-mile-range EVs and remote/cordless Level II charging infrastructure and $10,000-per-vehicle automation costs.

In the last several years, student mode choice has increasingly become an important area of study. Findings from these studies can be applied to regional travel demand modeling efforts, campus planning efforts, and sustainability initiatives, among others. This paper presents an analysis of student mode choice at the University of Texas at Austin, using statistical and geographic information systems analysis, based on the University of Texas Parking and Transportation Services mode choice survey administered during the spring 2014 semester. Results showed that within this sample, more students take alternative modes than drive alone, though the proportion of students driving alone to campus remains substantial. Among other conclusions, analysis also indicated clustering of respondent residential locations, and drive alone hotspots in several zip codes primarily in south/southeast Austin. These results point to a geographic area where it may be beneficial to concentrate resources aimed at inducing drivers to switch to an alternative mode of transportation, in order to support UT’s mobility and sustainability goals.

This dissertation examines the impact of parking pricing on travel demand and behavior, using the University of California (UC), Berkeley campus as a study site. Parking pricing is often implemented to recover costs or to serve as a source of revenue for cities or private parking operators. However, parking pricing can also be an effective transportation demand management tool. Parking price can be set at market rates or can be set to meet other objectives, such as reducing emissions or traffic. In either case, by increasing the direct cost of driving, parking pricing can lead travelers to shift to public transportation or non-motorized modes. Parking pricing can also help to reduce total distance traveled through cruising reduction, and through trip reduction or consolidation, and in so doing can decrease congestion, air pollution and other transportation externalities. Understanding the role of parking pricing in influencing travel demand and behavior is crucial for determining whether a flexible and variable pricing structure can be effective in managing parking demand and scarce land resources, yet at the same time, generating adequate economic revenue. The main objective of this dissertation is to analyze whether and to what extent changes in parking policies can alter transportation mode choice and parking preferences given different travel constraints, options and needs. Changes in parking policies examined in this dissertation not only include price, but also payment type (i.e. monthly, daily, or hourly), proximity of parking location to workplace and other incentives bundled together with specific parking options. Therefore, parking preference is defined as the pricing type and location of the chosen parking space. The types of parking pricing analyzed in this dissertation include paying by month, day, or hour, together with transit incentives bundled with different types of parking pricing options, while parking location is broadly divided into on-campus and off-campus parking. In order to better evaluate the impact of parking pricing and other transportation policies on travel behavior and demand, it is also necessary to understand how travel and parking behavior can be influenced by employment type and its respective flexibility of work schedule. In addition to accounting for the socioeconomic characteristics of the employees, this dissertation therefore investigates their job characteristics and the flexibility of their work schedule, both of which affect transportation mode choice and parking location because of their effects on time of travel, time, duration of stay at the workplace and frequency of commute trips. The UC Berkeley campus was selected as a study site to reevaluate current parking policies and to improve parking pricing to lower transportation demand and to reduce cruising for parking. The University is situated adjacent to the City of Berkeley's downtown, in the inner suburban ring of the San Francisco Bay Area. The campus is served directly by several AC Transit bus routes and a Bay Area Rapid Transit (BART) station. UC Berkeley is one of the largest employers and trip generators in the region, with more than 36,000 students, 1,377 faculty members, and more than 12,000 non-academic staff. As a result, it generates more than 50,000 trips per day, whereas there are only approximately 5,000 parking spaces available on campus. There is a clear constraint on parking availability and transportation demand management tools are vital in maintaining a relatively low driving mode share. Current parking policies are designed to cover current operating costs, but fall well short of replacement costs, with an annual budget of approximately $13 million, except for bond payments. Furthermore, there is a wide range of employment types, job levels, work schedules, residential locations, and socioeconomic characteristics at UC Berkeley, which reflect varying employee attitudes, commute and parking choices. Therefore, findings from this dissertation can be applied to other regions.UC Berkeley students are excluded in this study. Campus parking regulations restrict parking permits to students who live off campus at a distance of two miles or more, and only 26 percent of students meet this criterion. As a result, only eight parking lots or garages are available for student parking. The study focuses instead on faculty and staff transportation demand and parking behavior.A total of four different research methods were used to investigate attitudes and behavior, namely, open-ended interviews, focus groups, a transportation and parking survey, and discrete choice analysis. The combination of quantitative and qualitative methods provides complementary yet independent observations, as each method examines different facets of the research question. The survey was designed to examine current transportation demand and parking behavior, as well as potential changes in behavior under various parking pricing scenarios. Hence, it was used to collect both revealed preference (RP) and stated preference (SP) data.In-depth one-on-one interviews were conducted with a total of 86 UC Berkeley employees. The open-ended interviews were designed to understand the linkages amongst travel and parking behavior, work schedule and employment type. The purpose of interviewing is to understand the valuable lived experience and actions of a small sample of UC Berkeley employees, to gain a clearer perception of their current travel behavior, habits and preferences. An additional 10 focus groups with eight faculty members and 105 staff members were then conducted, prior to the final execution of the survey. Focus groups were used in this dissertation to allow a deeper understanding of the underlying reasons contributing to any potential changes in mode choice that cannot be captured by the interviews and survey alone. Findings from the interviews and focus groups were then incorporated into the final transportation and parking survey. The online version of the survey was mailed electronically to all campus faculty and staff members (approximately 12,000 employees) in December 2013, with a response rate of approximately 30 percent (n = 4,188). Data collected from the survey were used to develop multinomial logit (MNL) models for mode choice and parking choice. A RP-SP joint analysis was also conducted for the transportation mode choice model. Together, these research approaches illustrate current travel behavior and parking preferences. The also help determine the role of parking pricing in shifting transportation mode and parking location choices, show the differences in travel behavior and parking preferences according to University affiliation and provide insights into future UC Berkeley parking policies, as well as for other campus communities. Results from this study show that a considerable number of employees (23 percent) use a combination of various modes when commuting to campus, while others rarely switch to something other then their most preferred mode (77 percent). Most regular users of transit or non-motorized modes would drive occasionally too, with driving frequency ranging from once or twice a week to a few times a year. Employees who drive alone to campus are categorized by their frequency of car use in this dissertation, i.e. regular drivers, regular but flexible drivers and occasional drivers. Results show that regular drivers drive every day of the workweek mainly because of convenience, comfort, safety, low transit accessibility, and having dependents. Regular but flexible drivers live in residential locations without the availability of comparable transit services, or where biking or walking to campus is not a feasible option. However, they would use transit if services have improved or other transportation modes and not drive if they could. Hence, they are more flexible than regular drivers in terms of their driving frequencies and mode choice. They tend to not have any dependents and have arrival and departure times that are not affected by someone else's schedules. Lastly, occasional drivers are employees who have multiple transportation options and they could either be more cost sensitive or prefer to use transit or non-motorized transportation modes for other non-cost related reasons. In both cases, occasional drivers drive to campus under special circumstances, such as being late for work, bad weather, having to carry bulky and heavy belongs, or having to attend certain events after work. Results from the survey show that more employees in higher household income categories drive to campus than employees in lower income groups. Carpool and biking are two transportation mode choices that are not affected by income, as there are no substantial differences in the percentages of employees who carpool or bike across all income categories. On the other hand, the number of employees who use the bus is significantly higher for lower income groups than higher income groups and walking as a primary mode choice is most common for the lowest and highest income categories. Any changes in parking pricing on campus will tend to affect medium income groups the most. Work schedule and employment type have been found to affect parking location more than transportation mode choice. Driving alone is the most popular choice amongst all University affiliates and job categories as found in all three data sources, i.e. interviews, focus groups and survey. In general, almost half of the respondents drive alone to campus (49 percent), followed by transit (23 percent) and non-motorized transportation modes (16 percent). The remaining respondents carpool (seven percent), ride motorcycle (one percent), or use other forms of transportation, including being dropped off (four percent). However, not all University affiliates who drive have the same parking location preferences. Approximately 30 percent of employees who drive choose to park at of-campus parking locations. The flexibility of work schedule, which is directly related to employment type or University affiliation affects where an employee chooses to park more than transportation mode choice. Arrival and departure times, the number of hours spent on campus per day and the number of days on campus can all influence the parking location or parking type chosen. Lower income groups have been found to park less on campus than higher income groups. Changes in parking pricing are most likely to affect the parking location of employees with flexible work schedules more than their mode choice. Hence, employees who currently drive alone to campus will continue to do so, but may choose to park at a different location from where they currently park if parking prices are increased. Results from the mode choice models show that the respondents prefer alternatives with lower travel cost and time and that flexibility in work schedule may not decrease drive alone mode share. Staff members tend to prefer driving and parking on campus more compared to faculty members, while female respondents drive alone and park on campus more than male respondents do. Older employees with higher household income also prefer driving and parking on campus compared to younger employees and employees with lower household income. The average minimum annual salary at UC Berkeley is approximately $65,000 or $34 per hour. Comparing this value with the value of time estimates from the MNL models, it was found that the RP value of travel time of $16 per hour is 47 percent of the minimum average hourly income. The value of travel time is not surprisingly the lowest for low income respondents ($14 per hour) and the highest for high income respondents ($18 hour).The parking choice model shows that a daily parking payment is the most preferred parking option. The greater the parking fee refund and the availability of free transit passes, the greater the utilities of the monthly parking permits. When the transit pass also includes BART, it increases the utilities of the parking options. Survey respondents find parking options with shorter walking time more attractive and the longer an employee stays on campus, the greater the preference to drive and park on campus. Faculty members are less likely to choose a monthly parking permit than staff members. However, it is important to note that adjunct professors and other non-professor titled academic staff are included in the "faculty" category in the discrete choice analysis. The value of walking time from parking location to workplace is slightly lower at $15 per hour. The value of walking time shows that employees are willing to spend $0.25 to park a minute closer to their primary workplace on campus. Walking distance is thus a significant factor in influencing parking choices, in addition to parking costs and should be considered when pricing future parking facilities. The findings of this dissertation show that parking pricing plays an important role in regulating transportation demand by shifting mode choice, especially in a campus environment, where the majority of the employees who drive park on campus. However, it has a greater impact on influencing parking location, which can lead to changes in parking revenue for the University. Unlike existing studies, this dissertation examines the impact of parking pricing with the consideration of various payment type, parking location, transit incentives, flexibility of work schedule, income, and the willingness to walk from a parking space to the final destination. The use of qualitative research methods to complement results from a discrete choice analysis has also provided further insights to transportation mode choice and parking behavior.

This research develops an accessibility-based model of aggregate commute mode share, focusing on the share of transit relative to auto. It demonstrates the use of continuous accessibility -- calculated continuously in time, rather than at a single or a few departure times -- for the evaluation of transit systems. These accessibility calculations are accomplished using only publicly-available data sources. Multiple time thresholds for a cumulative opportunities measure of accessibility are evaluated for their usefulness in modeling transit mode share. A binomial logit model is estimated which predicts the likelihood that a commuter will choose transit rather than auto for a commute trip based on aggregate characteristics of the surrounding area. Variables in this model include demographic factors as well as detailed accessibility calculations for both transit and auto. The model achieves a pseudo-R-sqaure value of 0.597, and analysis of the results suggests that continuous accessibility of transit systems may be a valuable tool for use in modeling and forecasting. It may be possible to apply these techniques to existing models of transit ridership and mode share to improve their performance and cost-effectiveness.

In the two decades since the United States Congress passed the federal Intermodal Surface Transportation Efficiency Act, there has been a surge of interest in making urban transportation systems more sustainable. Many agencies, representing all levels of government, have searched for strategies to reduce private automobile use, including policies to shift local driving to pedestrian and bicycle modes. Progress has been made in a number of communities, but the automobile remains the dominant mode of transportation in all metropolitan regions.Sustainable transportation advocates are especially interested in routine travel, such as shopping and other errands, because it tends to be done frequently and for distances that could be covered realistically by walking or bicycling. According to the 2009 National Household Travel Survey, Americans made more trips for shopping than for any other purpose, including commuting to and from work. One-third of these shopping trips were shorter than two miles (3.2 km). However, 76% of these short shopping trips were made by automobile, while only 21% were made by walking and 1% by bicycling.In order to identify effective strategies to change travel behavior, practitioners need a greater understanding of why people choose certain modes for routine travel. Choosing to walk or bicycle rather than travel by automobile may help individuals get exercise, save money, interact with neighbors, and reduce tailpipe emissions. Yet, in some communities, non-motorized modes may also require more time and physical effort to run a series of errands, be less convenient for carrying packages and traveling in bad weather, and be perceived as having a higher risk of traffic crashes or street crime than driving.A mixed-methods approach was used to develop a more complete understanding of factors that are associated with walking or bicycling rather than driving for routine travel. An intercept survey was implemented to gather travel data from 1,003 customers at retail pharmacy stores in 20 San Francisco Bay Area neighborhoods in fall 2009. Follow-up interviews were conducted with 26 survey participants in spring and summer 2010 to gain a deeper understanding of factors that influenced their transportation decisions. The methodological approach makes several contributions to the body of research on sustainable transportation. For example, the study:--Explored multiple categories of factors that may be associated with walking and bicycling, including travel, socioeconomic, attitude, perception, and shopping district characteristics. Few studies of pedestrian or bicycle mode choices have included all of these categories of factors. Statistical models showed that variables in all categories had significant associations with mode choice.--Documented and analyzed short pedestrian movements, such as from a parking space to a store entrance or from a bus stop to home. These detailed data provided a greater understanding of pedestrian activity than traditional travel survey analyses. Walking was used as the primary mode for 65% of respondent trips between stops within shopping districts, and 52% of all respondents walked along a street or between stops at some time between leaving and returning home. Maps of respondent pedestrian path density revealed distinct pedestrian activity patterns in different types of shopping districts.--Used four different approaches to capture participant travel mode information. Respondents reported the primary mode of transportation they were using on the day of the survey, the mode they typically used, and all modes that they would consider using to travel to the survey store. They also mapped all stops on their tour and said what modes they used between each stop. These four approaches revealed nuanced travel habits and made it possible to correct inaccuracies in self-reported primary travel mode data.--Measured and tested fine-grained local environment variables in shopping districts rather than around respondents' homes. These variables characterized the shopping district area (e.g., sidewalks, bicycle facilities, metered parking, and tree canopy coverage), the main commercial roadway (e.g., posted speed limit, number of automobile lanes, and pedestrian crossing distance), and the survey store site (e.g., number of automobile and bicycle parking spaces and distance from the public sidewalk to the store entrance). This dissertation adds to the small number of studies that have explored how the characteristics of activity destinations are related to travel behavior.The study results contribute to the body of knowledge about factors that may encourage people to shift routine travel from automobile to pedestrian or bicycle modes. After controlling for travel factors such as time and cost, socioeconomic characteristics, and individual attitudes, mixed logit models showed that automobile use was negatively associated with higher employment density, smaller parking lots, and metered on-street parking in the shopping district. Walking was positively associated with higher population density, more street tree canopy coverage, lower speed limits, and fewer commercial driveway crossings. The exploratory analysis of a small number of bicycle tours found that bicycling was associated with more extensive bicycle facility networks and more bicycle parking. However, people were more likely to drive when they perceived a high risk of crime.Results also suggest the magnitude of mode shifts that could occur if short- and long-term land use and transportation system changes were made to each study shopping district. The mode choice model representing travel only to and from the study shopping districts (N = 388) was used to estimate respondent mode shares under the following three scenarios: 1) double population and employment densities in each study shopping district, 2) double street tree canopy coverage in each study shopping district, and 3) eliminate half of the automobile parking spaces at the survey store. Based on the model, the combination of these three changes could increase pedestrian mode share among the 388 sample respondents from 43% to 61% and decrease automobile mode share from 50% to 31%. This shift could eliminate 129 (13%) of the 983 respondent vehicle miles traveled (208 of the 1,580 respondent vehicle kilometers traveled), and 110 (36%) of the 308 times respondents parked their automobiles in the shopping district.The mode choice model of walking versus driving within survey shopping districts (N = 286) was used to test the combination of the following scenarios: 1) cluster separated stores around shared parking lots, 2) consolidate commercial driveways so that there are half as many driveway crossings along the main commercial roadway, 3) reduce all main commercial roadway speed limits to 25 miles per hour (40 kilometers per hour), and 4) install metered parking in all shopping districts. These changes could increase the percentage of the 286 sample respondents walking between shopping district activities from 32% to 54%. This shift could eliminate 29 (38%) of the 76 respondent vehicle miles traveled (47 of the 122 respondent vehicle kilometers traveled), and 105 (22%) of the 469 times respondents parked their automobiles in the shopping district. Note that these forecasted mode shifts are illustrative examples based on cross-sectional data and do not account for the process of modifying travel behavior habits.Qualitative interviews provided a foundation for a proposed Theory of Routine Mode Choice Decisions. This five-step theory also drew from survey results and other mode choice theories in the transportation and psychology fields. The first step, 1) awareness and availability, determines which modes are viewed as possible choices for routine travel. The next three steps, 2) basic safety and security, 3) convenience and cost, and 4) enjoyment, assess situational tradeoffs between modes in the choice set and are supported by many of the statistically-significant factors in the mode choice models. The final step, 5) habit, reinforces previous choices and closes the decision process loop. Socioeconomic characteristics explain differences in how individuals view each step in the process. Understanding each step in the mode choice decision process can help planners, designers, engineers, and other policy-makers implement a comprehensive set of strategies that may be able to shift routine automobile travel to pedestrian and bicycle modes.

Abstract: Trip chaining and mode choice are two important travel behavior decisions in activity-based travel demand modeling system. The hierarchy of these two decisions influences model’s predictive capability and policy sensitivity. This thesis is aimed at investigating the hierarchical relationship between these decisions and also the effects of socio-demographic characteristics on them. Structural Equation Modeling (SEM) technique is used for this investigation. A six week travel diary data collected in Thurgau, Switzerland in 2003 is used for model estimation. Model estimation results show that for work-tour, trip chain and mode choice decisions are simultaneous and it remains consistent across the six weeks. For weekday’s non-work tour, mode choice precedes trip chain whereas for weekend’s non-work tour trip chain precedes mode choice. The investigation of the effect of a number of socio-demographic characteristics on trip chaining and mode choice behaviors is also found useful for better understanding of these behaviors.

Abstract: Travel Demand Management (TDM) is now considered one of the most important aspects of transportation planning and operation. The prime objective of TDM is to develop a sustainable transportation system utilizing the existing infrastructure. It is now a well known fact that excessive use of single occupancy vehicle causes numerous problems like traffic congestion, environmental pollution etc. Thus, from TDM perspective, it is of great importance to analyze travel behaviour in order to influence people to reduce car use and choose more sustainable modes such as – carpool, public transit, park & ride, walk, bike etc. This study attempts an in-depth analysis of commuting mode choice behaviour using workplace commuter survey data from the City of Edmonton. Unlike traditional mode choice models, this study uses both instrumental and latent variables to better understand the choice process and analyzes their sensitivities with respect to TDM policies.

This study attempts to find a correlation between commuting modes in Washington DC and characteristics of the city and the people that they serve. It investigates why some census tracts have experienced increases in the commuting share of alternative transportation, such as public transit, walking, and bicycling, while others haven't. Findings demonstrate that demographic variables such as percent Hispanic and foreign born were the strongest predictors of change in commute mode share followed by distance to train station. Land use variables demonstrated weak correlations with variations in mode share due most likely to a lack of density gradient within the study area. The creation of variables to determine land use mix by census tract posed technical challenges as well. Recommendations include policy addressing rising demand for more diverse transportation systems be implemented and further research be conducted on creating more accurate land use variables to include in the model.

The aim of this study was to develop a framework to model travel choice behavior in order to estimate intercity travel demand at nation-level in the United States. Nested and mixed logit models were developed to study national-level intercity transportation in the United States. A separate General Aviation airport choice model to estimates General Aviation person-trips and number of aircraft operations though more than 3000 airports was also developed. The combination of the General Aviation model and the logit models gives the capability to estimate a full spectrum of intercity travel demand in the United States. The logit models were calibrated using a nationwide revealed preference survey (1995 American Travel Survey). Separate models were developed for business and non-business trip purposes. An airport choice model is integrated into the mode choice model to estimate both the market share between any origin-destination pair and other modes of transportation, and the market share split between airports associated with the origin-destination pairs. The explanatory variables used in the utility functions of the models are travel time, travel cost, and traveler's household income. The logit models are used to estimate the market share of automobile and commercial air transportation between 3091 counties and 443 commercial service airports in the United States. The model was also used to estimate market share for on-demand air taxi services. Given an input county-to-county trip demand table, the models were used to estimate county-to-county travel demand by automobile and commercial airline between all counties and commercial service airports in the United States. The model has been integrated into a computer software framework called the Transportation Systems Analysis Model (TSAM) that estimates nationwide intercity travel demand in the United States.

Artificial intelligence techniques have produced excellent results in many diverse fields of engineering. Techniques such as neural networks and fuzzy systems have found their way into transportation engineering. In recent years, neural networks are being used instead of regression techniques for travel demand forecasting purposes. The basic reason lies in the fact that neural networks are able to capture complex relationships and learn from examples and also able to adapt when new data become available. The primary goal of this thesis is to develop mode choice models using artificial neural networks and compare the results with traditional mode choice models like the multinomial logit model and linear regression method. The data used for this modeling is extracted from the American Travel Survey data. Data mining procedures like clustering are used to process the extracted data. The results of three models are compared based on residuals and error criteria. It is found that neural network approach produces the best results for the chosen set of explanatory variables. The possible reasons for such results are identified and explained to the extent possible. The three major objectives of this thesis are to: present an approach to handle the data from a survey database, address the mode choice problem using artificial neural networks, and compare the results of this approach with the results of traditional models vis-à-vis logit model and linear regression approach. The results of this research work should encourage more transportation researchers and professionals to consider artificial intelligence tools for solving transportation planning problems.

School-related traffic congestion is a large problem in North Carolina and throughout the nation. North Carolina Department of Transportation works with schools and municipalities to better design and retrofit campuses to accommodate the vehicular volumes that often queue onto adjacent streets and cause long delays. The large number of school trips made by automobile not only cause traffic problems, but safety problems as well, made evident by injury and fatality statistics. The goal of this research was to ultimately enhance student safety and reduce school-related traffic problems by gaining a better understanding of the household attributes and behaviors that influence school transportation mode choice in order to identify problems and prioritize solutions for school transportation, including school bus service improvements through automated vehicle location (AVL). The primary research objective was to calibrate a school transportation mode choice model for a selected North Carolina school district. Mode choice models were developed based on factors exhibiting statistical significance in estimating the choice of automobile or school bus for the morning and afternoon school trips of children in kindergarten through eighth grade in the Wake County Public School System. The variable expressing the convenience of the school bus service for a household based on parent work schedules, perceived problems, and AVL improvements was the most influential. Model transferability tests suggest that the models can be used statewide in schools or school districts where actual automobile usage ranges from about 30 to 55% of all school trips in the morning and 15 to 40% in the afternoon. The significant contribution of this research is the application of mode choice probability modeling to the grade school population. An assessment of the potential for AVL technologies to prompt a modal shift to the school bus was also included. Results suggest that 38% of students currently using the automobile for their morning school trips may shift to the school bus, indicating potential for a 16% overall modal shift to the school bus with the addition of AVL tracking and paging technologies.

TRANSIMS is a disaggregate, behavioral transportation planning package developed under US DOT's and EPA funding at the Los Alamos National Laboratory (LANL). It is an integrated system of travel forecasting models designed to give transportation planners accurate, complete information on traffic impacts, congestion, and pollution by simulating second-by-second movements of every person and every vehicle through the transportation network of a large metropolitan area. There is no built-in module for travellers' mode choices In TRANSIMS. The modes going with the shortest path are always taken. In Portland Study, a mode choice methodology implemented by a series of feedback processes is proposed. However, it uses aggregate, deterministic mode choice model. There is little solid theoretic ground for the format and coefficients of the generalized costs used in the calibration process. The accessibility to a mode, especially to Transit, was also not included in the model. In the thesis, a disaggregate and deterministic mode choice methodology in TRANSIMS is developed. The accessibility to each mode is analyzed and included in the model. The methodology is then implemented on the Blacksburg transportation planning study, namely Blacksburg_Lite. The analysis of the result is based on the indicator of mode choice, mode split between Transit and Auto. The indicator is close to that in survey data and converged fast. Therefore, this mode choice methodology could be used within TRANSIMS framework.