MIT Transit Lab

This paper proposes a general Network Performance Model (NPM) for urban rail systems performance monitoring using smart card data. NPM is a schedule-based network loading model with strict capacity constraints and boarding priorities. It distributes passengers over the network given origin-destination (OD) demand, operations, route choice, and effective train capacity. A Bayesian simulation-based optimization method for calibrating the effective train capacity is introduced, which explicitly recognizes that capacity may be different at different stations depending on congestion levels. Case studies with data from the Mass Transit Railway (MTR) network in Hong Kong are used to validate the model and illustrate its applicability. NPM is validated using left behind survey data and exit passenger flow extracted from smart card data. The use of NPM for performance monitoring is demonstrated by analyzing the spatial-temporal crowding patterns in the system and evaluating dispatching strategies.

The explosive growth of ridesourcing services has stimulated a debate on whether they represent a net substitute for or a complement to public transit. Among the empirical evidence that supports discussion of the net effect at the city level, analysis at the disaggregated level from a geospatial perspective is lacking. It remains unexplored the spatiotemporal pattern of ridesourcing's effect on public transit, and the factors that impact the effect. Using DiDi Chuxing data in Chengdu, China, this paper develops a three-level structure to recognize the potential substitution or complementary effects of ridesourcing on public transit. Furthermore, this paper investigates the effects through exploratory spatiotemporal data analysis and examines the factors influencing the degree of substitution via linear, spatial autoregressive, and zero-inflated beta regression models. The results show that 33.1% of DiDi trips have the potential to substitute for public transit. The substitution rate is higher during the day (8:00-18:00), and the trend follows changes in public transit coverage. The substitution effect is more exhibited in the city center and the areas covered by the subway, while the complementary effect is more exhibited in suburban areas as public transit has poor coverage. Further examination of the factors impacting the relationship indicates that housing price is positively associated with the substitution rate, and distance to the nearest subway station has a negative association with it, while the effects of most built environment factors become insignificant in zero-inflated beta regression. Based on these findings, policy implications are drawn regarding the partnership between transit agencies and ridesourcing companies, the spatially differentiated policies in the central and suburban areas, and the potential problems in providing ridesourcing service to the economically disadvantaged population.

Many subway systems operating near capacity face challenges in meeting reliability and level of service targets. This paper examines the effectiveness of various strategies to relieve congestion and increase capacity using a microscopic, agent-based, urban heavy rail simulation model. The Massachusetts Bay Transportation Authority’s (MBTA) Red Line serves as the testbed for the analysis. The Red Line operates very close to its capacity. Bottlenecks on the Red Line and possible strategies to mitigate them are discussed, including skip-stop, station consolidation, and dwell time control. The results show that, compared with the no strategy case, skip-stop and consolidation are effective in reducing runtimes and passenger journey times, increasing train throughput, and maintaining headway regularity during peak periods. Performance under these two strategies is also robust to dispatching irregularity and increases in passenger demand. The dwell time control strategy mitigates congestion and disturbances in operations to some extent, but is less effective and robust.

Although automatically collected human travel records can accurately capture the time and location of humanmovements, they do not directly explain the hidden semantic structures behind the data, e.g., activity types. This work proposes a probabilistic topic model, adapted from Latent Dirichlet Allocation (LDA), to discover representative and interpretable activity categorization from individual-level spatiotemporal data in an unsupervised manner. Specifically, the activity-travel episodes of an individual user are treated as words in a document, and each topic is a distribution
over space and time that corresponds to certain type of activity. The model accounts for a mixture of discrete and continuous attributes---the location, start time of day, start day of week, and duration of each activity episode. The proposed methodology is demonstrated using pseudonymized transit smart card data from London, U.K. The results show that the model can successfully distinguish the three most basic types of activities---home, work, and other. As the specified number of activity categories increases, more specific subpatterns for home and work emerge, and both the goodness of fit and predictive performance for travel behavior improve. This work makes it possible to enrich human mobility data with representative and interpretable activity patterns without relying on predefined activity categories or heuristic rules.

Although transit buses have a relatively small impact on greenhouse gas emissions, they have a larger impact on urban air quality, have commercially available electric models, and have historically commercialized clean technologies that enabled deployment in other heavy-duty vehicles. This paper investigates what factors affect transit agencies’ decisions to go beyond electric bus pilots to larger scale deployments, with the goal of identifying strategies to enable an accelerated transition to an electrified fleet. This mixed methods analysis utilized quantitative total cost of ownership analysis and qualitative interviews to study the barriers and drivers of electric bus investment for transit fleets in three case study states: California, Kentucky, and Massachusetts. A total cost of ownership analysis estimated electric buses are already more cost-effective than diesel buses in many agency contexts, but are sensitive to key parameters such as annual mileage, fossil fuel costs, and electricity tariffs and supporting policies that vary widely. Though multiple agencies in California reported planning to fully electrify their fleets, outside California where less supportive policies exist, fewer agencies reported planning to procure additional electric buses, primarily owing to high first cost and undesirable tradeoffs with maintaining transit service levels. Interview respondents also reported other substantial barriers such as oversubscribed grant programs, charging infrastructure costs, electricity costs, and additional operational complexity, suggesting a need for multiple complementary policies to overcome these barriers and ensure agencies can transition to a new technology without affecting transit service.

Monitoring rail transit system performance is important for effective operations planning. The number of times passengers are denied boarding is becoming a key measure of the impact of near-capacity operations on customers and is fundamental for calculating other performance metrics, such as expected waiting time for service. This paper reviews existing methods and proposes a denied boarding probability distribution inference method for closed Automated Fare Collection (AFC) urban rail systems. Using AFC (tap-in and tap-out) and Automated Vehicle Location (AVL) data, the method relaxes some of the limitations of existing approaches. The problem is modeled using a mixture distribution framework that incorporates a priori structural information. It is data-driven and requires neither observations of denied boardings, nor assumptions about access/egress time distributions. Also, for comparison purposes, the paper presents an event-based deterministic transit assignment model with explicit capacity constraints. While the network assignment works at the network level and requires train capacity, the mixture model works at the station level, requires no external parameters, and can be easily applied to any station and for any time period. A case study illustrates the application of the proposed methods using actual data and compares the results against existing methods, and also survey data. The results demonstrate the mixture model’s robustness and applicability for monitoring denied boarding.

Transportation Demand Management (TDM), long used to reduce car traffic, is receiving attention among public transport operators as a means to reduce congestion in crowded public transportation systems. Though far less studied, a more structured approach to Public Transport Demand Management (PTDM) can help agencies make informed decisions on the combination of PTDM and infrastructure investments that best manage crowding. Automated fare collection (AFC) data, readily available in many public transport agencies, provide a unique platform to advance systematic approaches for the design and evaluation of PTDM strategies. The paper discusses the main steps for developing PTDM programs: a) problem identification and formulation of program goals; b) program design; c) evaluation; and d) monitoring. The problem identification phase examines bottlenecks in the system based on a spatiotemporal passenger flow analysis. The design phase identifies the main design parameters based on a categorization of potential interventions along spatial, temporal, modal, and targeted user group parameters. Evaluation takes place at the system, group, and individual levels, taking advantage of the detailed information obtained from smart card transaction data. The monitoring phase addresses the longterm sustainability of the intervention and informs potential changes to improve its effectiveness. A case study of a pre-peak fare discount policy in Hong Kong’s MTR network is used to illustrate the application of the various steps with focus on evaluation and analysis of the impacts from a behavioral point of view. Smart card data from before and after the implementation of the scheme from a panel of users was used to study policy-induced behavior shifts. A cluster analysis inferred customer groups relevant to the analysis based on their usage patterns. Users who shifted their behavior were identified based on a change point analysis and a logit model was estimated to identify the main factors that contribute to this change: the amount of time a user needed to shift his/her departure time, departure time variability, fare savings, and price sensitivity. User heterogeneity suggests that future incentives may be improved if they target specific groups.

Travel demand management (TDM) is used for managing congestion in urban areas. While TDM is well studied for car traffic, its application in transit is still emerging. Well-structured transit TDM approaches can help agencies better manage the available system capacity when the opportunity and investment to expand are limited. However, transit systems are complex and the design of a TDM scheme, deciding when, where, and how much discount or surcharge is implemented, is not trivial. The paper proposes a general framework for the optimal design of promotion based TDM strategies in urban rail systems. The framework consists of two major components: network performance and optimization. The network performance model updates the origin-destination (OD) demand based on the response to the promotion strategy, assigns it to the network, and estimates performance metrics. The optimization model allocates resources to maximize promotion performance in a cost effective way by better targeting users whose behavioral response to the promotion improves system performance. The optimal design of promotion strategies is facilitated by the availability of smart card (automated fare collection, AFC) data. The proposed approach is demonstrated with data from a busy urban rail system. The results illustrate the value of the method, compare the effectiveness of different strategies, and highlight the limits of the effectiveness of such strategies.

Customer fare product choices can affect both ridership and revenue, so they are strategically important for transit agencies. Nearly all major agencies offer choices between pay-per-use and pass products, and with each potential fare change, agencies face decisions about whether to modify pass “multiples”—the number of rides needed to “break even” on a pass purchase. However, the simple elasticity spreadsheet models often used to analyze the potential ridership and revenue impacts of fare changes make little or no adjustment for shifts in fare product choices. This paper reviews different options for incorporating product choice into fare policy scenario models, and it presents a ridership and revenue prediction procedure that combines a multinomial logit fare product choice model with the logic of an elasticity spreadsheet model. This combination facilitates evaluation of complex fare changes that are likely to alter fare product market shares while maintaining much of the flexibility and simplicity of a traditional spreadsheet model. Additionally, the proposed model uses only preexisting, revealed-preference automated fare collection data rather than requiring customer surveys. The proposed model is demonstrated using examples at the Chicago Transit Authority (CTA). The CTA experienced a large shift from passes to pay-per-use following a fare change in 2013, illustrating the potential value of accounting for fare product choices in fare scenario evaluation.

The paper studies the impacts of built environment (BE) on the first- and last-mile travel modal choice. We select Singapore as a case study. The data used for this work is extracted from the first- and last-mile trips to mass rapid transit (MRT) stations in the Household Interview Travel Survey of Singapore in 2012 with nearly 24,000 samples. The BE indicators are quantified based on four “D” variables: Density, Diversity, Design, and Distance to transit. We also take into account sociodemographic and trip-specific variables. Mixed logit (ML) modeling frameworks are adopted to estimate the impact of BE and the heterogeneity of taste across the sample. Based on the availability of light rail transit (LRT) in different areas, two modeling structures are implemented with binary ML models for non-LRT areas where “walk” and “bus” are the available travel modes, and multinomial ML models for areas where LRT is an additional alternative. The modeling results shed light on the following findings: BE—especially distance to MRT station, transportation infrastructures, land-use mix, and socioeconomic activities—significantly influences the first- and last-mile travel behaviors. Those who live or work close to MRT stations and in an area with high socioeconomic activities and land-use mix may have stronger preferences to walk for the first- and last-mile trips. The impact of physical BE (i.e., distance, infrastructures) is relatively homogeneous among the sample, while the impact of socioeconomic BE factors (i.e., floor space density, entropy) tends to vary across the sample.

With subway systems around the world experiencing increasing demand, measures such as passengers left behind are becoming increasingly important. This paper proposes a methodology for inferring the probability distribution of the number of times a passenger is left behind at stations in congested metro systems using automated data. Maximum likelihood estimation (MLE) and Bayesian inference methods are used to estimate the left behind probability mass function (LBPMF) for a given station and time period. The model is applied using actual and synthetic data. The results show that the model is able to estimate the probability of being left behind fairly accurately.

Although stable in the short term, individual travel patterns are subject to changes in the long term. The ability to detect such changes is critical for developing behavior models that are adaptive over time. We define travel pattern change as "abrupt, substantial, and persistent changes in the underlying pattern of travel behavior" and develop a methodology to detect such changes in individual travel patterns. We specify one distribution for each of the three dimensions of travel behavior (the frequency of travel, time of travel, and origins/destinations), and interpret the change of the parameters of the distributions as indicating the occurrence of the pattern change. A Bayesian method is developed to estimate the probability that a pattern change occurs at any given time for each behavior dimension. The proposed methodology is tested using pseudonymized smart card records of 3,210 users from London, U.K. over two years. The results show that the method can successfully identify significant changepoints in travel patterns. Compared to the traditional generalized likelihood ratio (GLR) approach, the Bayesian method requires less predefined parameters and is more robust. The methodology presented in this paper is generalizable and can be applied to detect changes in other aspects of travel behavior and human behavior in general.

This paper proposes and simulates an integrated autonomous vehicle (AV) and public transportation (PT) system. After discussing the attributes of and the interaction among the prospective stakeholders in the system, we identify opportunities for synergy between AVs and the PT system based on Singapore’s organizational structure and demand characteristics. Envisioning an integrated system in the context of the first-mile problem during morning peak hours, we propose to preserve high demand bus routes while repurposing low-demand bus routes and using shared AVs as an alternative. An agent-based supply-side simulation is built to assess the performance of the proposed service in fifty-two scenarios with different fleet sizes and ridesharing preferences. Under a set of assumptions on AV operation costs and dispatching algorithms, the results show that the integrated system has the potential of enhancing service quality, occupying fewer road resources, being financially sustainable, and utilizing bus services more efficiently.

Service reliability is a major concern for public transportation agencies. Transit services experience natural variability in operations performance, due to factors such as congestion, changes in demand, and operator behavior. This variability leads to irregular headways, resulting in longer passenger waits and decreased effective capacity as gaps in service form. Real-time control strategies allow controllers to regulate service and improve performance. This research tested the effectiveness of a headway-based dispatching strategy at a terminal on the Massachusetts Bay Transportation Authority (MBTA) Green Line in Boston, a complex, four-branch light rail line. Terminal personnel were provided with tablet computers showing departure times optimized by an even-headway policy. When optimized departure times were adhered to, peak period headway variability was reduced by 40%. The average wait was shortened by 15% (30 sec), and the 90th percentile wait was shortened by 21% (90 sec). The results show that adopting headway-based dispatching at terminals of high-frequency lines promises significant benefits to service and passengers if operational changes are accompanied by improved supervision.

Demand for public transportation is highly affected by passengers’ experience and the level of service provided. Thus, it is vital for transit agencies to deploy adaptive strategies to respond to changes in demand or supply in a timely manner, and prevent unwanted deterioration in service quality. In this paper, a real time prediction methodology, based on univariate and multivariate state-space models, is developed to predict the short-term passenger arrivals at transit stations. A univariate state-space model is developed at the station level. Through a hierarchical clustering algorithm with correlation distance, stations with similar demand patterns are identified. A dynamic factor model is proposed for each cluster, capturing station interdependencies through a set of common factors. Both approaches can model the effect of exogenous events (such as football games). Ensemble predictions are then obtained by combining the outputs from the two models, based on their respective accuracy. We evaluate these models using data from the 32 stations on the Central line of the London Underground (LU), operated by Transport for London (TfL). The results indicate that the proposed methodology performs well in predicting short-term station arrivals for the set of test days. For most stations, ensemble prediction has the lowest mean error, as well as the smallest range of error, and exhibits more robust performance across the test days.

For intelligent urban transportation systems, the ability to predict individual mobility is crucial for personalized traveler information, targeted demand management, and dynamic system operations. Whereas existing methods focus on predicting the next location of users, little is known regarding the prediction of the next trip. The paper develops a methodology for predicting daily individual mobility represented as a chain of trips (including the null set, no travel), each defined as a combination of the trip start time t, origin o, and destination d. To predict individual mobility, we first predict whether the user will travel (trip making prediction), and then, if so, predict the attributes of the next trip (t, o, d) (trip attribute prediction). Each of the two problems can be further decomposed into two subproblems based on the triggering event. For trip attribute prediction, we propose a new model, based on the Bayesian n-gram model used in language modeling, to estimate the probability distribution of the next trip conditional on the previous one. The proposed methodology is tested using the pseudonymized transit smart card records from more than 10,000 users in London, U.K. over two years. Based on regularized logistic regression, our trip making prediction models achieve median accuracy levels of over 80%. The prediction accuracy for trip attributes varies by the attribute considered---around 40% for t, 70-80% for o and 60-70% for d. Relatively, the first trip of the day is more difficult to predict. Significant variations are found across individuals in terms of the model performance, implying diverse travel behavior patterns.

Previous research has combined automated fare-collection (AFC) and automated vehicle-location (AVL) data to infer the times and locations of passenger origins, interchanges (transfers), and destinations on multimodal transit networks. The resultant origin–interchange–destination flows (and the origin–destination (OD) matrices that comprise those flows), however, represent only a sample of total ridership, as they contain only those journeys made using the AFC payment method that have been successfully recorded or inferred. This paper presents a method for scaling passenger-journey flows (i.e., linked-trip flows) using additional information from passenger counts at each station gate and bus farebox, thereby estimating the flows of non-AFC passengers and of AFC passengers whose journeys were not successfully inferred.

The proposed method is applied to a hypothetical test network and to AFC and AVL data from London’s multimodal public transit network. Because London requires AFC transactions upon both entry and exit for rail trips, a rail-only OD matrix is extracted from the estimated multimodal linked-trip flows, and is compared to a rail OD matrix generated using the iterative proportional fitting method.

Innovative analyses of origin–destination (OD) data derived from automatic fare collection and automatic vehicle location systems in public transport networks enable planners to gain new insights into how passengers travel in the network and the quality of service provided, and can even inform decisions about network improvements. Particularly in large, complex networks, systematic, data-driven approaches to network evaluation and planning are essential. New methodologies are needed to transform OD data into informative metrics and planning recommendations. This paper proposes a framework for this process and applies it to London’s public transport network. Though there are many ways to improve public transport networks, this paper focuses on the addition of new bus routes to reduce circuity. The proposed framework includes three steps that combine OD-level analysis with spatial aggregation methodologies for the identification of corridors for new bus services. First, bus stops and rail stations were clustered into geographic zones. Second, a subset of zonal OD pairs with circuitous service were identified as candidates for improvement through new bus routes, based on performance standards established with user-defined parameters. Third, an algorithm that clusters OD pairs into corridors was applied to identify promising corridors for new bus services. This paper discusses corridors identified for new services in the London case study.

Autonomous vehicles (AVs) represent potentially disruptive and innovative changes to public transportation (PT) systems. However, the exact interplay between AV and PT is understudied in existing research. This paper proposes a systematic approach to the design, simulation, and evaluation of integrated autonomous vehicle and public transportation (AV+PT) systems. Two features distinguish this research from the state of the art in the literature: the first is the transit-oriented AV operation with the purpose of supporting existing PT modes; the second is the explicit modeling of the interaction between demand and supply. We highlight the transit-orientation by identifying the synergistic opportunities between AV and PT, which makes AVs more acceptable to all the stakeholders and respects the social-purpose considerations such as maintaining service availability and ensuring equity. Specifically, AV is designed to serve first-mile connections to rail stations and provide efficient shared mobility in low-density suburban areas. The interaction between demand and supply is modeled using a set of system dynamics equations and solved as a fixed-point problem through an iterative simulation procedure. We develop an agent-based simulation platform of service and a discrete choice model of demand as two subproblems. Using a feedback loop between supply and demand, we capture the interaction between the decisions of the service operator and those of the travelers and model the choices of both parties. Considering uncertainties in demand prediction and stochasticity in simulation, we also evaluate the robustness of our fixed-point solution and demonstrate the convergence of the proposed method empirically. We test our approach in a major European city, simulating scenarios with various fleet sizes, vehicle capacities, fare schemes, and hailing strategies such as in-advance requests. Scenarios are evaluated from the perspectives of passengers, AV operators, PT operators, and urban mobility system. Results show the trade off between the level of service and the operational cost, providing insight for fleet sizing to reach the optimal balance. Our simulated experiments show that encouraging ride-sharing, allowing in-advance requests, and combining fare with transit help enable service integration and encourage sustainable travel. Both the transit-oriented AV operation and the demand-supply interaction are essential components for defining and assessing the roles of the AV technology in our future transportation systems, especially those with ample and robust transit networks.

Planners must understand how public transportation systems are used in order to make strategic decisions. Smart card transaction data provides vast, detailed records of network usage. Combined with other automatically collected data sources, established inference methodologies can convert smart card transactions into complete linked journeys made by individuals within the public transport network. However, for large, multi-modal public transport networks it can be challenging to summarize the journey records meaningfully. This paper develops a method for categorizing origin–destination (OD) pairs by public transport mode or combination of used modes. By aggregating across OD pairs, this categorization scheme summarizes the multi-modal aspects of public transport network usage. The methodology can also be applied to subsets of data filtered by time of day or geography. The categorization results can inform performance analysis of OD pairs, allowing planners to make comparisons between pairs served by different combinations of modes. London Oyster card data is analyzed to illustrate how the OD pair categorization can characterize a network, allowing planners to quickly assess the roles of different modes, and perform OD pair analysis in a multi-modal public transport network.

Transit agencies increasingly deploy planning strategies to improve service reliability and real-time operational control to mitigate the effects of travel time variability. The design of such strategies can benefit from a better understanding of the underlying causes of travel time variability. Despite a significant body of research on the topic, findings remain influenced by the approach used to analyze the data. Most studies use linear regression to characterize the relationship between travel time reliability and covariates in the context of central tendency. However, in many planning applications, the actual distribution of travel time and how it is affected by various factors is of interest, not just the condition mean. This paper describes a quantile regression approach to analyzing the impacts of the underlying determinants on the distribution of travel times rather than its central tendency, using supply and demand data from automatic vehicle location and farecard systems collected in Brisbane, Australia. Case studies revealed that the quantile regression model provides more indicative information than does the conditional mean regression method. Moreover, most of the coefficients estimated from quantile regression are significantly different from the conditional mean–based regression model in terms of coefficient values, signs, and significance levels. The findings provide information related to the impacts of planning, operational, and environmental factors on speed and its variability. On the basis of this information, transit designers and planners can design targeted strategies to improve travel time reliability effectively and efficiently.

Regularity is an important property of individual travel behavior, and the ability to measure it enables advances in behavior modeling, mobility prediction, and customer analytics. In this paper, we propose a methodology to measure travel behavior regularity based on the order in which trips or activities are organized. We represent individuals’ travel over multiple days as sequences of “travel events”—discrete and repeatable behavior units explicitly defined based on the research question and the available data. We then present a metric of regularity based on entropy rate, which is sensitive to both the frequency of travel events and the order in which they occur. The methodology is demonstrated using a large sample of transit smart card transaction records from London, UK. The entropy rate is estimated with a procedure based on the Burrows-Wheeler transform. The results confirm that the order of travel events is an essential component of regularity in travel behavior. They also demonstrate that the proposed measure of regularity captures both conventional patterns and atypical routine patterns that are regular but not matched to the 9-to-5 working day or working week. Unlike existing measures of regularity, our approach is agnostic to calendar definitions and makes no assumptions regarding periodicity of travel behavior. The proposed methodology is flexible and can be adapted to study other aspects of individual mobility using different data sources.

The value of accessibility concepts is well-established in transportation literature, but so is the low adoption of accessibility-based instruments by practitioners. Based on the premise that leveraging accessibility concepts to address public involvement challenges could promote their adoption in planning practice, this research investigates mechanisms to promote social learning among participants in public workshops. Potential mechanisms of learning include specific tool-based interactions and how such interactions reinforce structures of learning such as alignment and imagination. This paper details iterative testing of these mechanisms with a tool called CoAXs (short for Collaborative ACCESSibility-based stakeholder engagement system), through focus groups and exploratory workshops. A mixed-methods analysis of the workshops supports the expectation that alignment and imagination correlate positively with social learning, as measured by reported learning and dialog quality. Specific interactions with the accessibility-based features of CoAXs in turn correlate positively with alignment and imagination, at individual and group levels of analysis. These findings, while not robustly generalizable, suggest that effective targeted stakeholder engagement for public transport can be structured around collaborative accessibility mapping. Adoption for broader public participation requires further development, especially the incorporation of actual day-to-day experiences such as unreliable operations.

Unit costs measured as bus operating costs per vehicle mile have increased considerably above the inflation rate in recent decades in most transit agencies in the United States. This paper examines the impact of (lack of) productivity growth, union bargaining power, and contracting out on cost escalation. We draw from a 17-year (1997–2014) and a 415-bus transit agency panel with 5780 observations by type of operation (directly operated by the agency or contracted out). We have three main findings: first, the unit cost increase in the transit sector is far worse than what economic theory predicts for industries with low productivity growth. Second, contracting out tends to reduce unit costs, and the results suggest that the costs savings from private operations can be only partly explained by lower wages in the private sector. Interestingly, we find that the cost savings from contracting out are larger when the transit agency also directly operates part of the total transit service. However, while overall unit costs are lower in contracted services, cost growth in large private bus operators is no different than cost growth in large public transit operators. Third, unique transit labor laws that lead to union bargaining power are a likely driver of the unit cost growth above inflation. Overall, these factors reflect inherent characteristics of the bus transit sector, such as the nature of low productivity growth and union legislative power related to the need for public subsidy. They drive increases in both transit fares and public subsidy at rates higher than inflation, and play an important role in the deterioration of transit agencies’ financial sustainability.

This paper explores the possibility of using subway maps as a planning tool to influence passenger route choice to mitigate congestion. Specifically, it tests whether extending the appearance of an overcrowded subway line on the Washington DC subway map would encourage passengers to use other underutilized lines. The experiment was conducted through the Mechanical Turk, a crowdsourcing platform, with 3056 participants, producing 21,240 route choice decisions on the official and six alternative maps. Results show that redesigned maps significantly affect participants’ route choices. Depending on the specific design, a 20% length increase of the overcrowded line could move 1.9–5.7 percentage points of ridership to an alternative, underutilized line. The change could remove up to 10 passengers per car during the highest peak, reducing the number of highly congested half-hour periods (max load = 100–120 passengers per car) on the overcrowded line from 4 to 1, and the number of crush periods (max load > 120 passengers per car) from 3 to 1. This is done at minimal or zero cost. The paper calls for more attention from transit agencies to the planning potential of transit maps.

For public transportation agencies, the ability to provide personalized and dynamic passenger information is crucial for improving the efficiency of demand management and enhancing customer experience. This requires understanding and especially predicting individual travel behavior in the public transportation system, which is challenging because of the heterogeneity among passengers and the variability of their behaviors. This paper presents, to the best of the authors' knowledge, the first attempt to predict individual spatiotemporal behavior of public transportation passengers using smartcard data. In this study, each trip is coded as a combination of trip start time, an entry station and an exit station. A passenger's daily mobility is represented as a chain of travel decisions. The authors propose a new modeling framework, inspired by Bayesian n-gram models used in natural language processing, to estimate the probability distribution of the next decision in the sequence. Empirical analysis using Oyster card data from London shows promising results. It is found that the exact time of travel is most challenging to predict, but the difference between the predicted time and the true value is usually small. Model performance varies greatly across individuals for the prediction of entry and particularly exit stations. Overall, the proposed model shows significant improvement over the regular n-gram models, or Markov chain-based models in general. The improvement is even larger for weekend trips when travel behavior is flexible, irregular, and considerably less predictable.

Urban public transport systems must periodically close certain links for maintenance, which can have significant effects on the service provided to passengers. In practice, the effects of closures are mitigated by replacing the closed links with a simple shuttle service. However, alternative shuttle services could reduce inconvenience at a lower operating cost. This paper proposes a model to select shuttle lines and frequencies under budget constraints. We propose a new formulation that allows a minimal frequency restriction on any line that is operated and minimizes passenger inconvenience cost, which includes transfers and frequency-dependent waiting time costs. This model is applied to a shuttle design problem based on a real-world case study of the Massachusetts Bay Transportation Authority network of Boston, Massachusetts. The results show that additional shuttle routes can reduce passenger delay compared to the standard industry practice, while also distributing delay more equally over passengers, at the same operating budget. The results are robust under different assumptions about passenger route choice behavior. Computational experiments show that the proposed formulation, coupled with a preprocessing step, can be solved faster than prior formulations.

The paper presents a methodology for assigning passengers to individual trains using: (i) fare transaction records from Automatic Fare Collection (AFC) systems and (ii) Automatic Vehicle Location (AVL) data from train tracking systems. The proposed Passenger-to-Train Assignment Model (PTAM) is probabilistic and links each fare transaction to a set of feasible train itineraries. The method estimates the probability of the passenger boarding each feasible train, and the probability distribution of the number of trains a passenger is unable to board due to capacity constraints. The access/egress time distributions are important inputs to the model. The paper also suggests a maximum likelihood approach to estimate these distributions from AFC and AVL data. The methodology is applied in a case study with data from a major, congested, subway system during peak hours. Based on actual AFC and train tracking data, synthetic data was generated to validate the model. The results, both in terms of the trains passengers are assigned to and train loads, are similar to the "true" observations from the synthetic data. The probability of a passenger being left behind (due to capacity constraints) in the actual system is also estimated by time of day and compared with survey data collected by the agency at the same station. The left behind probabilities can be accurately estimated from the assignment results. Furthermore, it is shown that the PTAM output can also be used to estimate crowding metrics at transfer stations.

The near ubiquity of smartphones has the potential to transform how researchers, companies, and public transit agencies understand travel behavior. This research analyzes how an emerging class of automatically-collected data based on smartphone GPS and sensor information - referred to here as mobile activity-tracking data - can be used in a transit agency to better understand travel behavior. Through a collaboration with Transport for London, multiple weeks of mobile activity-tracking data of London residents was collected between 2015 and 2016 using an application called Moves. Using this case study, this paper discusses the benefits of this new data and how it compares with other data at TfL and elsewhere and examines the process of collecting the data. Using the resulting data, this paper then compares the resulting trip records from the mobile activity tracking data with those from the automatic fare card data collected during the same period and same individuals. By comparing mobile activity tracking with an established, well-researched data source like AFC, the authors observe that while the trip match rate between the two data sources is high (68%) but not perfect. Next, the paper proposes a probabilistic framework to identify between motorized trip modes using mobile activity tracking data and and the public transit network. Specifically, the model uses both spatial characteristics, such as distance to public transit network, and trip characteristics such as speed in order to identify the trip mode as bus, rail, subway, or motorized non-public transit. Using logistic regression, classification tree, and random forest, this model achieves an accuracy of 90%, 91%, and 92% respectively.

High-frequency transit systems are essential for the socioeconomic and environmental well-being of large and dense cities. The planning and control of their operations are important determinants of service quality. Although headway and optimization-based control strategies generally outperform schedule-adherence strategies, high-frequency operations are mostly planned with schedules, in part because operators must observe resource constraints (neglected by most control strategies) while planning and delivering service. This research develops a schedule-free paradigm for high-frequency transit operations, in which trip sequences and departure times are optimized in real-time, employing stop-skipping strategies and utilizing real-time information to maximize service quality while satisfying operator resource constraints. Following a discussion of possible methodological approaches, a simple methodology is applied to operate a simulated transit service without schedules. Results demonstrate the feasibility of the new paradigm.

Origin-destination matrices provide vital information for service planning, operations planning, and performance measurement of public transportation systems. In recent years, methodological advances have been made in the estimation of origin-destination matrices from disaggregate fare transaction and vehicle location data. Unlike manual origin-destination surveys, these methods provide nearly complete spatial and temporal coverage at minimal marginal cost. Early models inferred destinations on the basis of the proximity of possible destinations to the next origin and disregarded the effect of waiting time, in-vehicle time, and the number of transfers on path choice. The research reported here formulated a dynamic programming model that inferred destinations of public transportation trips on the basis of a generalized disutility minimization objective. The model inferred paths and transfers on multileg journeys and worked on systems that served a mix of gated stations and ungated stops. The model is being used to infer destinations of public transportation trips in Boston, Massachusetts, and is producing better results than could be obtained with earlier models.

The growing ubiquity of smartphones offers public transit agencies an opportunity to transform ways to measure, monitor, and manage service performance. The potential of a new tool is demonstrated for engaging customers in measuring satisfaction and co-monitoring [Editor’s note: This is the authors’ word, meaning “agencies using public feedback to supplement official monitoring and regulation.”] bus service quality. The pilot project adapted a smartphone-based travel survey system, Future Mobility Sensing, to collect real-time customer feedback and objective operational measurements on specific bus trips. The system used a combination of GPS, Wi-Fi, Bluetooth, and accelerometer data to track transit trips while soliciting users’ feedback on trip experience. Though not necessarily intended to replace traditional monitoring channels and processes, these data can complement official performance monitoring through a more real-time, customer-centric perspective. The pilot project operated publicly for 3 months on the Silver Line bus rapid transit in Boston, Massachusetts. Seventy-six participants completed the entrance survey; half of them actively participated and completed more than 500 questionnaires while on board either at the end of a trip, at the end of a day, or both. Participation was biased toward frequent Silver Line users, the majority of whom were white and of higher income. Indicative models of user-reported satisfaction reveal some interesting relationships, but the models can be improved by fusing the app-collected data with actual performance characteristics. Broader and more sustained user engagement remains a critical future challenge.

Unit costs measured as operating costs per vehicle mile in the public transit sector have increased significantly above the inflation rate in recent decades in the United States, regardless of mode and location. This paper examines the impact of (lack of) productivity growth, union bargaining power, and contracting out on cost escalation. The authors draw from a 17-year (1997-2014) and 438-agency panel of 8,276 observations by mode (bus vs. rail) and type of operations (directly operated by the agency vs. contracted out). They have three main findings: First, the unit cost increase in public transit sector is worse than what the Baumol disease predicts - i.e. more than the growth rate that would occur if transit wage rate increases were equal to those prevailing elsewhere in the economy. Second Contracting out tends to reduce unit costs, suggesting that the costs savings from private operations are partly explained by lower wages to workers. However, while overall costs are lower in contracted services, cost escalation in medium and large private bus operators is no different than in large public transit operators, and the cost savings are larger when the transit agency also directly operates a share of the overall transit service. Third, unique transit labor laws are a likely driver of the unit cost growth above inflation. Overall, these factors reflect inherent characteristics of the transit sector, such as the nature of low productivity growth and union bargaining power related to the need for public subsidy. They drive increase in both transit fares and public subsidy at rates higher than inflation, and play an important role in the deterioration of transit agencies' financial sustainability.

Access distance to public transport is an important metric for planning, modeling, and evaluating public transport networks and is often used in policy goals and statements. However, accurately measuring access (and egress) distance can be difficult. Estimates often rely either on aggregate inferences based on census data or on small samples of disaggregate data from travel diary surveys. When smart cards used for fare payment are also registered with home address information, they represent a new data source that can be used to infer access distances for a large sample of users, at a disaggregate level and at low cost, compared with travel diary surveys. This paper demonstrates the inference of access distance from smart card fare and transaction data for a large sample of London public transport journeys and compares the inferred access distributions to data from the London Travel Demand Survey, a travel diary survey. Possible instances of false inferences are considered and measures to eliminate false inferences are discussed. This access distance inference methodology allows for the analysis of variation in access distance across the network, and examples of this type of analysis are presented.

Access distance to public transport is an important metric for planning, modeling, and evaluating public transport networks and is often used in policy goals and statements. However, accurately measuring access (and egress) distance can be difficult. Estimates often rely either on aggregate inferences based on census data or on small samples of disaggregate data from travel diary surveys. When smart cards used for fare payment are also registered with home address information, they represent a new data source that can be used to infer access distances for a large sample of users, at a disaggregate level and at low cost, compared with travel diary surveys. This paper demonstrates the inference of access distance from smart card fare and transaction data for a large sample of London public transport journeys and compares the inferred access distributions to data from the London Travel Demand Survey, a travel diary survey. Possible instances of false inferences are considered and measures to eliminate false inferences are discussed. This access distance inference methodology allows for the analysis of variation in access distance across the network, and examples of this type of analysis are presented.

Dwell time constitutes a significant portion of transit vehicles' travel times and thereby directly affects the quality of service that customers experience. Although automated fare-collection (AFC) systems are purported to reduce dwell times on services that require onboard fare processing, the benefits of AFC systems (and by comparison, the costs of non-AFC transactions) have been difficult to quantify. This paper develops and applies a methodology for estimating the impact of various fare media upon bus and light-rail travel times using archived AFC and automated vehicle-location (AVL) data from the Massachusetts Bay Transportation Authority in Boston. From a baseline distribution of AFC-card transaction durations, marginal transaction durations are estimated for magnetic-stripe tickets and cash payments, yielding an excess farebox-interaction time metric. The metric is compared across services and time periods, and its impact upon each passenger is estimated by building upon previous research that infers individual passenger origins and destinations from AFC and AVL data. Cash fares are estimated to have a significantly higher marginal processing time than other media, yet magnetic-stripe tickets account for approximately one-third of system-wide excess fare-processing time because of their higher usage. Cash durations, however, are more variable, warranting possible further research into their impact upon headway variability. Card and ticket holders are classified by their inferred home locations, and the geographic distribution of riders enduring higher cumulative excess farebox-interaction times is compared to environmental justice data to explore potential equity implications.

Urban public transport systems must periodically close certain links for maintenance, which can have significant effects on the service provided to passengers. In practice, the effects of closures are mitigated by replacing the closed links with a simple shuttle service. However, alternative shuttle services could reduce inconvenience at a lower operating cost. This paper proposes a model to select shuttle lines and frequencies under budget constraints. We propose a new formulation that allows a minimal frequency restriction on any line that is operated and minimizes passenger inconvenience cost, which includes transfers and frequency-dependent waiting time costs. This model is applied to a shuttle design problem based on a real-world case study of the Massachusetts Bay Transportation Authority network of Boston, Massachusetts. The results show that additional shuttle routes can reduce passenger delay compared to the standard industry practice, while also distributing delay more equally over passengers, at the same operating budget. The results are robust under different assumptions about passenger route choice behavior. Computational experiments show that the proposed formulation, coupled with a preprocessing step, can be solved faster than prior formulations.

Increases in ridership are outpacing capacity expansions in a number of transit systems. By shifting their focus to demand management, agencies can instead influence how customers use the system, getting more out of the capacity they already have. This paper uses Hong Kong's MTR system as a case study to explore the effects of crowding-reduction strategies as well as methods to use automatically collected fare data to support these measures. MTR introduced a pre-peak discount in September 2014 to encourage users to travel before the peak hour and reduce on-board crowding. To understand the impacts of this intervention, existing congestion patterns were first reviewed and a clustering analysis was performed to reveal typical travel patterns among MTR users. Then changes to when users chose to travel were studied at three levels to evaluate the program’s effects. Patterns among all users were measured across both the whole system and for specific rail segments. The travel patterns of the user groups, who have more homogeneous usage characteristics, were also evaluated, revealing differing responses to the promotion among groups. The incentive was found to have small impacts on morning travel, particularly at the beginning of the peak hour and among users with commuter-like behavior. Aggregate and group-specific elasticities were developed to inform future promotions and the results were also used to suggest other potential incentive designs.

Arrival time predictions for the next available bus or train are a key component of modern Traveller Information Systems (TIS). A great deal of research has been conducted within the ITS community developing an assortment of different algorithms that seek to increase the accuracy of these predictions. However, the inherent stochastic and non-linear nature of these systems, particularly in the case of bus transport, means that these predictions suffer from variable sources of error, stemming from variations in weather conditions, bus bunching and numerous other sources. In this paper we tackle the issue of uncertainty in bus arrival time predictions using an alternative approach. Rather than endeavour to develop a superior method for prediction we take existing predictions from a TIS and treat the algorithm generating them as a black box. The presence of heteroscedasticity in the predictions is demonstrated and then a meta-model approach deployed that augments existing predictive systems using quantile regression to place bounds on the associated error. As a case study this approach is applied to data from a real-world TIS in Boston. This method allows bounds on the predicted arrival time to be estimated, which give a measure of the uncertainty associated with the individual predictions. This represents to the best of our knowledge the first application of methods to handle the uncertainty in bus arrival times that explicitly takes into account the inherent heteroscedasticity. The meta-model approach is agnostic to the process generating the predictions which ensures the methodology is implementable in any system.

Monitoring the service quality of high-frequency bus transit is important both to agencies running their own operations and those contracting out, where performance measures can be used to assess contract penalties or bonuses. The availability of automatically collected vehicle movement and demand data enables detecting changes in running times and demand, which may present opportunities to improve service quality and fleet utilization. This research develops a framework to maximize service performance in a set of high-frequency bus routes, given their planned headways and a total fleet size constraint. Using automatically collected data and simulation modelling to evaluate the performance of each route with varying fleet sizes, a greedy algorithm adjusts allocation toward optimality. A simplified case study involving morning peak service on nine bus routes in Boston demonstrates the feasibility and potential benefits of the approach. A potential application is automated detection of routes operating with insufficient or excessive resources.