Transit Travel Time Research Articles

Modeling transit travel times for predicting equity improvements

Modeling transit travel times for predicting equity improvements

Optimizing bikeshare service to connect affordable housing units with transit service

Affordable housing and transit accessibility have long been focal points for housing and urban development agencies at all levels, from national to regional. However, these elements are often considered and planned for independently, leading over time to an expanding spatial gap between affordable housing and transit services. This separation creates issues of mobility equity. Fortunately, bikeshare, an emerging mobility solution, provides the potential to bridge this spatial gap between affordable housing communities and transit services. To leverage this, we designed a bikeshare station (BSS) location optimization tool that aims to alleviate the impacts of this spatial divide through bikeshare services. Specifically, we developed a multimodal agent-based modeling (ABM) and simulation framework aimed at enhancing accessibility to a variety of destinations. To solve this optimization problem, we developed a genetic algorithm to determine the optimal locations for BSS. There are three main conclusions from this research. First, the strategic positioning of BSS will enhance accessibility for affordable housing community (AHC) residents by reducing average transit travel time and walk distance, or increasing the number of destinations accessible by transit. This underscores the advantages of combining bikeshare and transit systems to cater to the mobility requirements of AHC residents, especially for non-work-related trips in suburban areas. Second, the relationship between transit and bikeshare is twofold: complementation and substitution. In instances where a lengthy trip involves multiple transit transfers and significant waiting time, bikeshare could replace some transit trips. Lastly, but most importantly, this study advises a forward-looking approach for governments and operators to avoid the waste of public resources when planning bikeshare systems. The tool developed by this research can facilitate this forward-looking approach in practice.

Computer vision for transit travel time prediction: an end-to-end framework using roadside urban imagery

AbstractAccurate travel time estimation is paramount for providing transit users with reliable schedules and dependable real-time information. This work is the first to utilize roadside urban imagery to aid transit agencies and practitioners in improving travel time prediction. We propose and evaluate an end-to-end framework integrating traditional transit data sources with a roadside camera for automated image data acquisition, labeling, and model training to predict transit travel times across a segment of interest. First, we show how the General Transit Feed Specification real-time data can be utilized as an efficient activation mechanism for a roadside camera unit monitoring a segment of interest. Second, automated vehicle location data is utilized to generate ground truth labels for the acquired images based on the observed transit travel time percentiles across the camera-monitored segment during the time of image acquisition. Finally, the generated labeled image dataset is used to train and thoroughly evaluate a Vision Transformer (ViT) model to predict a discrete transit travel time range (band). The results of this exploratory study illustrate that the ViT model is able to learn image features and contents that best help it deduce the expected travel time range with an average validation accuracy ranging between 80 and 85%. We assess the interpretability of the ViT model’s predictions and showcase how this discrete travel time band prediction can subsequently improve continuous transit travel time estimation. The workflow and results presented in this study provide an end-to-end, scalable, automated, and highly efficient approach for integrating traditional transit data sources and roadside imagery to improve the estimation of transit travel duration. This work also demonstrates the added value of incorporating real-time information from computer-vision sources, which are becoming increasingly accessible and can have major implications for improving transit operations and passenger real-time information.

Accessibility of Opioid Treatment Programs Based on Conventional vs Perceived Travel Time Measures

Transportation barriers have long been associated with poorer health outcomes; this burden is especially acute for individuals with opioid use disorder (OUD), a chronic disease often associated with low socioeconomic status. Conventional travel time analyses may not fully account for experiential components of travel, thereby understating the true travel burden and overstating treatment accessibility to opioid treatment programs (OTPs). To develop a metric of feels-like accessibility for those using public transit to access OTPs that accounts for the realistic travel burden on individuals with OUD. This cross-sectional study integrated high-resolution transit schedules and operating hours of OTPs to measure feels-like accessibility. Feels-like accessibility considers the differential outcomes of out-of-vehicle travel components and more realistically reflects individuals' transportation burden than conventional accessibility measures. Gini indices and spatial regression models were used to investigate inequities in accessibility. Geocoded data for residential addresses of 1018 overdose fatalities in Connecticut in 2019 were used as a proxy for the treatment needs of individuals with OUD. Data were analyzed between May and August 2023. Conventional and feels-like accessibility scores. Fluctuations in public transit frequencies over the course of the day and the limited operating hours of the OTPs. Of the 1018 individuals in the study, the mean (SD) age at death was 43.7 (12.6) years, 784 individuals (77%) were men, 111 (11%) were African American, and 889 (87%) were White, with other racial and ethnic categories including 18 individuals (2%). A total of 264 individuals in the sample (26%) could not access an OTP within 180 minutes. For those who could access these facilities, the average 1-way travel time was 45.6 minutes, with individuals spending approximately 70% of their trip duration on out-of-vehicle travel components. The conventional accessibility metric underestimates individuals' travel burden to OTPs as well as the inequity in accessibility compared with the feels-like accessibility metric. For example, the median (range) conventional accessibility score, defined as the number of OTPs within 120 minutes of transit travel time, was 5.0 (0.0-17.0); the median (range) feels-like accessibility score, defined as the number of OTPs within 120 minutes of transit travel time weighted to account for in- and out-of-vehicle segments, was 1.0 (0.0-10.0). There is a considerable temporal variation in travel time and accessibility depending on the departure times. In this cross-sectional study of travel burdens, the calculated feels-like accessibility scores, which consider the differential outcomes of out-of-vehicle travel components (eg, walking and waiting), could better and more realistically reflect passengers' transportation burden. Policy recommendations derived from the conventional accessibility metric could be misleading, and decision-makers should use feels-like accessibility metrics that adequately capture individuals' travel burdens. In the context of access to OTPs, the findings from this study suggest that opening new OTP sites to address gaps in access due to distance to services or extending hours of operation at existing sites may ameliorate the travel burden for individuals.

Breaking down public transit travel time for more accurate transport equity policies: A trip component approach

In the Northern hemisphere, low-income transit users typically take shorter trips and use buses in a greater proportion, but there is limited knowledge on inequalities in the detailed characteristics of public transit trips.In this paper, we seek to understand and explain differences in segment components of transit trips across socio-economic groups. Trip components include access time, wait time, total travel time, overall trip speed and percent of trip time spent in access and wait (off-vehicle time).Using the Montreal metropolitan region travel survey (2013), we focus on the 46,186 transit trips made by 20,138 individuals for which household income was available. This data is augmented by time estimations of the aforementioned trip components, and combined with trips specific transit service measures calculated using General Transit Feed Specification (GTFS). We first assess the relationship between trip components and income using descriptive statistics. OLS regressions are then conducted to identify which individual and transport-based factors contribute to variations in trip component outcomes.Low-income transit users take more transit trips (especially bus and Metro) than their higher income counterparts and combine multiple transit modes through more frequent transfers. While average access and wait time are individually lower than those for trips of wealthier respondents, low-income total trip time is similar but trip speed is considerably lower. Lower-income populations also spend a greater share of their trips in access and wait time due to service and trip discrepancies including trip timing, proximity to metro stations and requirement for transfers. When controlling for other factors, the conditions under which trips by low-income individuals are taken mainly drive income differences.

State of the BART: Analyzing the Determinants of Bay Area Rapid Transit Use in the 2010s

Peaking on public transit—the concentration of ridership in peak times and directions into and out of central areas—has waxed in the U.S. over the past century, as public transit has lost more mode share at off-peak times, in off-peak directions, and among non-commute trips. A notable pre-pandemic manifestation of this chronic problem was on Bay Area Rapid Transit, the San Francisco Bay Area’s regional heavy rail system. While BART staved off an absolute ridership decline longer than most American transit operators in the mid- and late-2010s, it did so almost entirely due to peak gains in riders offsetting off-peak losses. As a result, the system experienced worsening passenger crowding at some times and places, expanding underutilization of capacity at many others, and the prospect of enormous expenditures to accommodate rising transbay passenger demand. To examine the factors driving transit use in the 2010s, we model peak and off-peak BART trips as a function of station area and system characteristics. We uniquely use origin–destination pairs as the unit of analysis in order to separately measure influences at both ends of the trip. We find that transfers and travel time most influence peak and off-peak BART ridership and that station-area employment and time-competitiveness with driving particularly influence peak patronage. Over time in our models, the associations between ridership and transit travel time weakened, while the associations between ridership and transfers, employment, and time-competitiveness with driving grew stronger. In sum, we find that the peaking problem plaguing public transit systems for decades worsened in the years leading up to the pandemic—on this one nationally significant U.S. transit system, at least—which poses potentially substantial financial challenges in the years ahead.

Longer trips to court cause evictions

Studying ∼200,000 evictions filed against ∼300,000 Philadelphians from 2005 to 2021, we focus on the role of transit to court in preventing tenants from asserting their rights. In this period, nearly 40% of tenants facing eviction were ordered to leave their residences because they did not show up to contest cases against them and received a default judgment. Controlling for a variety of potential confounds at the tenant and landlord level, we find that residents of private tenancies with longer transit travel time to the courthouse were more likely to default. A 1-h increase in estimated travel time increases the probability of default by between 3.8% and 8.6% points across different model specifications. The effect holds after adjusting for direct distance to the court, unobserved landlord characteristics, and even baseline weekend travel time. However, it is absent in public housing evictions, where timing rules are significantly laxer, and during the COVID-19 pandemic, when tenants had the opportunity to be present virtually. We estimate that had all tenants been equally able to get to the court in 10 min, there would have been 4,000 to 9,000 fewer default evictions over the sample period. We replicate this commuting effect in another dataset of over 800,000 evictions from Harris County, Texas. These results open up a new way to study the physical determinants of access to justice, illustrating that the location and accessibility of a courthouse can affect individual case outcomes. We suggest that increased use of video technology in court may reduce barriers to justice.

Estimating suburban preferences for integrated public transit and ridesourcing services through a revealed-preference/stated-preference survey

Low-density areas like suburbs can be challenging to serve with fixed-route transit. Demand responsive transit has been used to service these areas with mixed results, but suburban travellers may respond to ridesourcing integrated with transit more positively because of the improved booking and dispatching interface offered by ridesourcing. This paper explores how suburban residents perceive transit-integrated ridesourcing, both for cost and time attributes of the system and in comparison to other travel modes in the area (cycling, auto, taxi or Uber, and transit). A Hierarchical Bayes mixed logit model was estimated using the results of a revealed-preference/stated-preference survey, revealing that transit-integrated ridesourcing had similar preferences to fixed-route transit. Across time and cost variables, respondents were more sensitive to walk time, transfer time, the number of transfers, parking cost, and taxi or Uber fare, and were minimally impacted by changes in the IVTT deviation of their trip times. The change in utility for the length of transit or transit-integrated ridesourcing in-vehicle travel time versus auto in-vehicle travel time was found to have a strongly linear relationship, while other attributes exhibited non-linearity to different degrees, particularly walk time, transfer time, transit-integrated ridesourcing fare, and taxi or Uber fare. Respondents who were female, younger, taking home-based school trips, or had lower household incomes were more open to transit-integrated ridesourcing and transit than other demographic groups. The likelihood of taking each of the five travel modes was assessed after COVID-19, finding generally minimal to negative impact for transit-integrated ridesourcing. Insights into vehicle and transit pass ownership, existing travel alternatives, and a prior transit-integrated ridesourcing pilot in the area are presented.

A DSS Framework for Advanced Traffic Signal Control System Investment Planning

This article presents ongoing research on the initial development of a decision support system (DSS) framework for integrated emergency vehicle preemption and transit priority system investment planning. A conceptual intelligent DSS framework provides a holistic framework to perform analytical assessments of integrated emergency vehicle preemption and transit priority systems. Three analytical tools are presented for incorporation into future DSS design: the first addresses the potential impact of transit travel time reduction on transit operating costs; the second addresses the potential impact of reduced emergency vehicle crashes at signalized intersections on fire and rescue operating costs; and the third integrates fuzzy sets concepts and multiattribute decision-making methods to rank order transit signal priority strategy alternatives at the intersection level.

Transit Access to Subsidized Food Stores in the U.S. Midwest

In the United States, low-income individuals and some racial and ethnic minorities have higher rates of transit use and food insecurity. The Supplemental Nutrition Assistance Program (SNAP), which provides low-income households with money for groceries at authorized retailers, creates a unique food procurement environment for participants. In this article, transit networks’ effect on access to SNAP retailers is assessed by calculating transit travel times to all transitable SNAP stores of census block groups (CBGs) in thirty-three metropolitan areas in the U.S. Midwest. Low transit access CBGs are identified and compared to U.S. Department of Agriculture (USDA) low-income, low-access (LILA) census tracts. Store type transit access ratios are analyzed to further specify disproportionate access to healthy food retailers. Findings indicate low transit access is experienced in fringe-urban, majority white, car-dependent CBGs. Approximately 24,380 people receiving public assistance, 145,636 Black people, 135,731 non-Black Hispanic people, and 108,325 low-income people live within these CBGs, however. The USDA’s LILA census tracts are more expansive than this study’s low-income, low-transit-access (LILTA) tracts. Despite being classified as rural, LILTA-only tracts had sixty-eight households without a vehicle, on average. Future research and policy should engage populations in suburban and exurban regions to accommodate transportation needs of a nonnegligible portion of potentially food-insecure households.

What Do We Know When? Modeling Predictability of Transit Operations

Predictions of transit delays are crucial to passengers and operators. Passengers utilize the predictions to decide on departure time, route choice, and mode choice, whereas operators decide on schedules, timetables, rolling stock allocation, and control actions. We introduce the concept of predictability of transit travel times as the study of the reduction of the predicted variability with the temporal approaching of a predicted event. We evaluate predictability on a real-life test case in Zurich, Switzerland, spanning multiple transit lines over one year of operations. The concept is shown based on predictions obtained by a state-of-the-art Bayesian network approach, where we show how predictability (in general) can be modeled as an exponential decay phenomenon. The study of predictability of transit operations leads to additional insights for control actions and system analysis compared to other complementary concepts such as punctuality or regularity, for instance, concerning bunching, identification of bottlenecks, and passenger routing.

Development of Planning-Level Guidelines for Deploying Transit Signal Priority

Transit Signal Priority (TSP) has been proposed and implemented to reduce transit travel time, provide better schedule adherence, and improve transit efficiency. However, concerns have been raised about causing delays to cross street and left-turn movements. The cost-effective implementation of TSP requires methods to support the decisions associated with the implementation. The objective of this research was to examine the existing TSP guidance and propose enhanced planning-level guidelines for TSP deployment that can be applied to identify intersections where implementing TSP is beneficial. The proposed guidelines take consideration of bus delay, bus frequency, geometric and traffic feasibility, as well as the impacts on other movements. The effectiveness of the proposed guidelines was evaluated in comparison with existing guidelines using a microscopic simulation approach. The simulation results showed that the proposed TSP guidelines produce a better-balanced solution for minimizing both main street bus travel times and delays for other affected movements. The proposed guidelines also help to cut the TSP implementation costs by reducing the number of intersections identified for TSP implementation.

Effects of trip-level characteristics on autonomous vehicle ownership: A U.S. analysis

We examined how trip-level characteristics may influence autonomous vehicle ownership decisions. Using 639 stated preference survey responses in the US, we developed an integrated choice and latent variable model linking trip- and consumer-level characteristics to AV ownership choices. Results suggested parking cost, travel cost and travel time are important predictors for vehicle ownership at the trip-level; and monthly payment, car ownership importance and autonomous vehicle safety perceptions are significant predictors at the consumer-level. By examining elasticities of vehicle ownership market shares, we found while consumer-level characteristics have the most prominent effect on vehicle ownership decisions, trip-level characteristics have a small but measurable effect. Among trip-level characteristics, parking cost has the largest impact on vehicle ownership decisions, followed by travel cost of ridehailing (both conventional and autonomous), and transit travel time. Finally, we simulated future vehicle ownership decisions under three policy scenarios by leveraging trip- and consumer-level characteristics.

Bus Travel Time Prediction: A Comparative Study of Linear and Non-Linear Machine Learning Models

Congested roads are a global problem, and increased usage of private vehicles is one of the main reasons for congestion. Public transit modes of travel are a sustainable and eco-friendly alternative for private vehicle usage, but attracting commuters towards public transit mode is a mammoth task. Commuters expect the public transit service to be reliable, and to provide a reliable service it is necessary to fine-tune the transit operations and provide well-timed necessary information to commuters. In this context, the public transit travel time is predicted in Tumakuru, a tier-2 city of Karnataka, India. As this is one of the initial studies in the city, the performance comparison of eight Machines Learning models including four linear namely, Linear Regression, Ridge Regression, Least Absolute Shrinkage and Selection Operator Regression, and Support Vector Regression; and four non-linear models namely, k-Nearest Neighbors, Regression Trees, Random Forest Regression, and Gradient Boosting Regression Trees is conducted to identify a suitable model for travel time predictions. The data logs of one month (November 2020) of the Tumakuru city service, provided by Tumakuru Smart City Limited are used for the study. The time-of-the-day (trip start time), day-of-the-week, and direction of travel are used for the prediction. Travel time for both upstream and downstream are predicted, and the results are evaluated based on the performance metrics. The results suggest that the performance of non-linear models is superior to linear models for predicting travel times, and Random Forest Regression was found to be a better model as compared to other models.

Exploring the Utility of Gini Coefficients as a Measure of Temporal Variation in Public Transit Travel Time

Travel times by public transit vary dramatically throughout the day. However, these variations are not the same for all commutes. If one commute origin and destination are connected by high frequency transit all day, then that commute will have less variation in travel time throughout the day. This research illustrates how that variation can be measured with gini coefficients so that inequality in travel times can be better understood. Using Denver as a case study, gini coefficients are calculated with 120 travel times by public transit throughout the day for each of over 200,000 origin destination pairs. The utility of the gini coefficient as a predictor of the percent of the population that drives to work is then explored using a regression. The regression results indicate that the gini variable has utility in predicting the percent of the population that drives to work.

Safety Impacts of Transit Signal Priority Using a Full Bayesian Approach

Transit signal priority (TSP) is a strategy that prioritizes the movement of transit vehicles through a signalized intersection to provide better transit travel time reliability and minimize transit delay. Although TSP is primarily intended to improve the operational performance of transit vehicles, it may also have substantial safety benefits. This study explored the potential safety benefits of the TSP strategy deployed at various locations in Florida. An observational before–after full Bayes (FB) approach with a comparison group was adopted to estimate the crash modification factors (CMFs) for total crashes, rear-end crashes, sideswipe crashes, and angle crashes. The analysis was based on 12 corridors equipped with the TSP system and their corresponding 29 comparison corridors without the TSP system. The deployment of TSP was found to reduce total crashes by 7.2% (CMF = 0.928), rear-end crashes by 5.2% (CMF = 0.948), and angle crashes by 21.9% (CMF = 0.781), and these results are statistically significant at a 95% Bayesian credible interval (BCI) except for the rear-end crashes. On the other hand, sideswipe crashes increased by 6% (CMF = 1.060) although the increase was not significant at a 95% BCI. Overall, the results indicated that TSP improves safety. The findings of this study may present key considerations for transportation agencies and practitioners when planning future TSP deployments.

The Disparity in Transit Travel Time between Koreans and Japanese in 1930s Colonial Seoul

Transportation is a key element to understanding the socio-spatial structure of colonial cities and the lives of individuals living under colonial governance. This study investigates the disparity in transit-based travel time between colonial rulers (Japanese) and subjects (Koreans) in Colonial Seoul (Keijo) in 1936 using modern GIS and open-source transport analysis tools. Findings suggest a significant disparity in travel time to a major urban facility (i.e., City Hall) between the two population groups of the largest colonial city in the Korean peninsula.

Integrating transit data and travel time into food security analysis: A case study of San Diego, California

Research suggests an important role for planners and policymakers in enhancing neighborhood access to healthy food. Spatial analysis is often used to explore relationships between proximity to sources of food and health outcomes such as food security and risk for diet-related disease. Spatial models often account for proximity based on walking distance or automobile but advancements in geospatial technologies now allow for estimates of access by transit as well as travel time to understand neighborhood level food security. One approach for addressing this gap is to employ real-time travel estimates over transit and road networks to more precisely measure food access. In this paper, we compare the Google Distance Matrix API calculating both the distance and time between origins and destinations, with other spatial approaches to understanding grocery store and full-service supermarket access in this case study of San Diego, California. Results suggest that distances between residents and food outlets are constant across modes of transport but travel times are highly differentiated by mode. Trips to stores by transit are on average, three times longer than trips by car. This paper offers a new understanding of the potential for incongruence between spatial proximity and true accessibility in the built environment.

Identification of Affecting Factors on the Travel Time Reliability for Bus Transportation

Public bus transit travel time is affected by many factors, including traffic signals and traffic condition. Transit agencies have implemented transit signal priority (TSP) strategies to reduce transit travel time and improve service reliability. However, due to the lack of empirical data, these factors' collective impact and bus travel time strategies have not been studied at the stop-to-stop segment level. This research focuses on the factors affecting travel time reliability, emphasising the variability between operators and the policy implications of such differences. One-way analysis of variance (ANOVA) statistical methods have been used to assess the quality implications of public bus transportation time reliability. This research seeks to investigate the factors affecting the travel time (TT) reliability of bus transport. Studies were conducted along three bus routes serving different areas. Factors strongly related to TT reliability include route length, number of signalised intersections, day of the week, bus stops, departure delays, bus lane, passenger boarding and alighting, weather condition, and fare structure. Based on the proposed model factors affecting TT reliability, it was found that TT is strongly affected by the number of bus stoppings and also the length of the route. The reliability of all three routes during the weekday is low because of delays in departure. The number of signalised intersections along the route affects reliability. Meanwhile, more passengers boarding and paying cash increased the travel time reliability of buses.

Assessment of the socio-spatial effects of urban transport investment using Google Maps API

We analyze the spatially distributed impacts of transport investment in urban highways and public transport with a novel methodology based on the capabilities of online technology to replicate the (unobserved) condition without highways. This is based upon the intensive use of Google Maps API (GMA) to obtain travel times between each origin-destination pair at a highly detailed level to reveal the effects of new infrastructure on different zones and groups within a city. Santiago is used as a case study, as the city introduced 150 km of urban highways, a reorganization of surface transit, and new subway lines in a relatively short period. We show that the high-income segment of the population has been the most favored, simultaneously increasing the difference between transit and car travel times in those areas where car ownership is low, stimulating the acquisition of a car.