Agent-Based Approach to Modeling Bus Bunching and Dispersion Under the Influence of Demand Variability
Gamal Eldeeb, University of British ColumbiaShow Abstract
Ahmed Idris, University of British Columbia
Lael Parrott, University of British Columbia
This paper presents an agent-based approach to understanding vehicle bunching under demand variability. A hypothetical bus route that consisting of 12 stops was designed. Using NetLogo 6.0, three different scenarios were modelled to analyze the bunching phenomenon under different conditions. Two new indices, a Bunching Index (BI) and a Dispersion Index (DI) were introduced to quantify vehicle bunching and dispersion by measuring the variance in passengers’ waiting time. According to the analysis, the best way to alleviate bunching is a combination of buses coordination, passengers’ variance accommodation, and using terminals as control points to prevent bunching propagation. Inadequate spacing between stops, traffic conditions, passengers’ variance, relatively small headway, and small terminal time may trigger bunching phenomenon.
Reliability-Oriented Transit Signal Priority System
Wenxin Qiao, Beijing Jiaotong UniversityShow Abstract
Ding Wang, Beijing Jiaotong University
Jia Hu, Tongji University
Transit signal priority (TSP) may sacrifice the right-of-way of vehicles on side streets, break progression, lead to substantial delay increase on general traffic. This paper addresses this problem by proposing a bi-level transit signal priority model whose goal is to optimize intersection reliability. The objective of this optimization model is to minimize the adverse effect of transit signal priority on the vehicles traveling in the non-priority direction while maintaining transit priority benefit and intersection reliability. Case study was conducted to show the performance of the proposed model: when the proposed model applied in passive transit signal priority, the average vehicle delay in the non-priority direction compared with the traditional model decreased 22.54 s (17.63%), meanwhile, the vehicle delay increases slightly in the priority direction (2.87 s) and the passenger delay for the entire intersection has small increase (0.26 s). In real-time transit signal priority case study, the result shows the proposed model can reduce 20s bus delay compare to no TSP plan.
A Network-Based Approach to Determining Optimal Operational Strategies for a Bus Transit Line
Chunyan Tang, Dalian University of TechnologyShow Abstract
Avishai Ceder, Technion - Israel Institute of Technology
Ying-En Ge, Shanghai Jiao Tong University
It is known that bi-directional bus-service passenger demand fluctuates across the day and it is neither uniform nor symmetric. However, in practice, there are no deployments of operational strategies to better match supply and demand. That is, mostly bus service is based on a full route operation (FRO) strategy, visiting each stop on the route. Thus, the FRO strategy may result in imbalanced passenger loads, i.e., empty seats or overcrowding, representing inefficiency of operation. Previous studies dealt with the introduction of different operational strategies using a given timetable, but with possible changes of vehicle scheduling in the process of optimal examination of different strategies. This study proposes a new network-based methodology for optimal determination of operational strategies using changes of departure times to accommodate the fluctuated passenger demand. Four possible strategies are considered: FRO, limited-stop, short-turn, and a combination of the latter two strategies, for a single-line bus transit operation. The methodology is based on a minimum-cost network-flow model constructed by adjusting vehicle departure times at determined stops. The network-flow problem, including special constraints, is solved by constructing an equivalent directed network of a particular pattern. The proposed methodology is firstly applied to a numerical example and secondly to a real-life bus line in Dalian, China. The results of the case study show that compared with the current FRO operation the implementation of a combination of different strategies reduces the total vehicle hours and empty-seat hours by 17.40% and 72.90%, respectively.
Model of a Semiflexible Transit Routing System Based on the Willingness to Pay
Peiqun Lin, South China University of TechnologyShow Abstract
Mingyang Pei, South China University of Technology
Yingying Ma, South China University of Technology
A Data-Driven Approach to Prioritizing Bus Schedule Revisions at New York City Transit
Lauren Tarte, MTA New York City TransitShow Abstract
Malik Coleman, MTA New York City Transit
Steve Chau, MTA New York City Transit
Brian Levine, MTA New York City Transit
Alla Reddy, MTA New York City Transit
Over two million trips are taken every weekday across the MTA New York City Transit (NYCT) bus network. Revising the schedules for each of these routes is a labor-intensive process, and due to limited resources, fewer than half of all routes are examined each year. Traditionally, schedules have been revised on a first-in, first-out basis, with most schedules rewritten once every two years. This approach leaves no room for reviewing routes that need more frequent changes, meaning service may not catch up to changes in demand or traffic patterns for several years. It also requires spending valuable staff time analyzing routes that may be stagnant.
In order to better address bus service in rapidly evolving corridors, NYCT developed a methodology to select routes for review based on need rather than a fixed cycle. This data-driven approach makes use of the Bus Time Automatic Vehicle Location (AVL) system and NYCT’s previously developed ridership model, which combines Automated Fare Collection (AFC) data with other sources to infer stop-by-stop boarding and alighting data at the individual trip level. Each route’s schedule is evaluated on the capacity of service supplied versus actual ridership, and scheduled versus actual running times. Routes that show the greatest discrepancies are then designated for later in-depth review. This methodology was applied to develop the 2018 list of schedules revisions. Because this process identifies routes that have too much capacity or running time, as well as those with too little, it allows resource-costly schedule adjustments to be offset with resource-saving ones. In addition, this methodology allows scheduling staff to react more quickly to changes in customer demand, road conditions, and new development, thereby improving bus performance and providing better service to passengers.
A New Approach to Precisely Evaluate the Operation of Bus Lanes
Yaning Xue, No OrganizationShow Abstract
Zhengyu Duan, Tongji University
Although more bus lanes have been constructed in Shanghai, the operating speed and passenger volume of urban bus system still keep dropping in recent years. It is necessary to evaluate the effect of new bus lanes. This paper proposes a method based on bus GPS data to identify the specific congested sections in bus lanes, which can get valuable insights into the operation of bus lanes. Firstly, the temporal and spatial distributions of GPS points are identified, at which the speed is less than 5km/h . Secondly, congested sections in bus lanes are identified and clustered by Hierarchical clustering method. Finally, several potential problems are put forward, with the consideration of bus lane realistic conditions. Through the case study of Xizang Road bus lane in Shanghai, China, it is found that: 1)78.2% congested sections are located in the bus lanes with bus stops; 2) Multiple and short congested sections locate between bus stops and upstream intersections, while single and long congested sections locate between bus stops and downstream intersections.
Study on Impacts of the Bus Bay on Traffic Flow
Xiaojian Hu, Southeast UniversityShow Abstract
The purpose of building a harbor-type bus station is to mitigate the impact of bus stops on traffic flow. This paper attempts to study the actual influence of the bus bay on its neighboring vehicles in China. Combining the traffic conflict technique and the ordered probit model, the paper records the intact operation of the station and explores the contributing factors. The results show that: 1) the violations of the bus and non-motor vehicles and bus-leaving condition have a significant impact on the drivers’ behaviors 2) the bus-leaving condition affects the severity of the lane-changing conflict. The paper also finds that violations around the bus bay, such as non-motor vehicles violation, slightly influence the conflict severity, which can be studied further to compare different bus stop type.
Why Are South African Cities Different? Comparing Johannesburg Rea Vaya Bus Rapid Transit System with Its Latin American Siblings
Harvey Scorcia, Development Bank for Latin America (CAF)Show Abstract
Ramon Munoz-Raskin, The World Bank
The objective of this paper is to shed light on the discussion about why Bus Rapid Transit (BRT) has structural operational indicator differences between South African cities and their siblings in Latin American.Under the understanding that BRT does (and will continue to) have a role in South African mobility, South African local and national governments are working together to better understand how does BRT work in the South African context and what type of subsidization would be needed as these systems expand. Inspired by the success of BRT systems in Latin American cities (LACs), Johannesburg pioneered implementing a BRT network in South Africa, along with Cape Town. As of 2016, 43.5 km of trunk BRT corridors were fully operational in Johannesburg and their results in terms of ridership and operating cost recovery from fares were approximately one-third of initial estimates. It is widely known that, as a result of apartheid, Johannesburg urban form is characterized by an average low density and spatial patterns where the poor are located very far from opportunities. This results into passenger travel demand patterns that have long, unidirectional trips with high peak-to-base ratios, which significantly differ to the travel patterns in dense LACs where there is high turnover and much shorter journeys. The paper provides a comparative analysis of Johannesburg’s BRT structural conditions and selected performance indicators with those BRTs in LACs that originally inspired their implementation in South Africa. The comparative analysis provides unequivocal evidence that Johannesburg’s (and South African cities’) BRT, as implemented, may not have the conditions to cover its operating costs from farebox revenue and therefore will typically present significantly different operation indicators to those of its LAC siblings.
Elimination of Weaving Bottleneck at Curbside Bus Stops with Dynamic Speed Guidance
Wenqing Chen, University of Wisconsin, MilwaukeeShow Abstract
Jie Yu, University of Wisconsin, Milwaukee
Jing Zhao, University of Shanghai for Science and Technology
Intersections often become the focus of various kinds of traffic contradiction, and thus have the traffic congestion more severe. A critical reason is that the near-side bus station may generate a weaving area, where the private vehicles and bus block each other. This paper proposes a dynamic speed control algorithm towards the intersection approaching lanes with a near-side bus station. The contribution of the algorithm is to utilize real time communication theory to change vehicles’ moving status, so as to avoid the weaving bottleneck caused by the near-side bus station. In addition, theoretically, the algorithm could minimize the total person travel time and idling number, simultaneously. An example is provided to validate the efficacy of the algorithm. This study also conducts the sensitivity analysis to provide operational guidelines for bus scheduling. Analysis results of the illustrative example indicate the validity and effectiveness of the proposed speed control algorithm.
Evaluation of Bus Rapid Transit Attributes in a Developing Country Related to Passenger Mobility and Operational Performance
Bilal Malik, University of Engineering and Technology, Lahore, PakistanShow Abstract
Bus rapid transit (BRT) systems can be particularly beneficial in developing countries as they can bring or extend many of the benefits of light and heavy rail transit systems while requiring substantially lesser capital investment. Despite rapid growth in BRTs worldwide, the dynamics of BRTs in developing countries still present unique challenges related to various underlying factors that may (or may not) positively or negatively affect commuter experience. This study focuses on identifying and evaluating such affects and their extent for the first BRT in Pakistan which started operating in 2013 in Lahore – the 2nd most populous city in the country and the 32nd in the world. This BRT has a 16.8 miles (27 km) long corridor and is a combination of both elevated and at-grade segments. A sample of 1282 commuters was surveyed throughout the corridor. Various attributes were identified and quantified that impact the satisfaction of BRT commuters using factor analysis and scoring of scales. Although majority of the factors suggested positive impacts on passenger mobility, it was determined that space for female passengers in the buses, facilities for the disabled and overall condition of seating/capacity in buses were among most critical factors in terms of commuter dissatisfaction. The findings can provide an insight in the user satisfaction related to BRT system performance in a developing country, and can be a useful platform for improving various planning aspects of future BRT systems.
Modeling Bus Bunching Using Cellular Automata
Fatemeh Enayatollahi, University of TehranShow Abstract
Ahmed Idris, University of British Columbia
Amiri Atashgah, University of Tehran
Bus bunching is a common problem in street-transit systems with high service frequency. Bunching can lower transit capacity and affect users’ satisfaction. In this paper, a bus route is modelled to investigate the effects of variations in demand on headway instability and bus bunching phenomenon. Besides, with the aim of comparing different bunching situations, an index is proposed based on bunching severity and intensity. Severity corresponds to the number of buses involved in bunching and intensity accounts for total deviations from designed headway. A one-dimensional Cellular Automata (CA) model is developed to simulate a designed route service under different demand conditions. Unlike previous researches, the focus of this study is on one design point with an operational approach, rather than obtaining a phase diagram from random demand and service frequencies. Using the proposed bunching index, it is shown that randomness in passenger arrival rate involves 65% of the service in bunching. Analyzing other factors such as road traffic, bus speed, and bus capacity, etc. on bunching formation, in addition to studying the effect of bunching on maintaining schedule and service reliability is suggested for future works.
Guidance for Identifying Corridor Conditions That Warrant Deploying Transit Signal Priority
MD Sultan Ali, Florida Atlantic UniversityShow Abstract
Evangelos Kaisar, Florida Atlantic University
Kyle Masters, Florida Department of Transportation
Transit Signal Priority (TSP) is an operational strategy that facilitates the movement of transit vehicles, either buses or streetcars, through traffic-signal controlled intersections. Nowadays, our streets and highways grow in traffic congestion as large number of cars enter the transportation system due to rise in population. The high volume of vehicles and numerous signalized intersections, traffic congestion is causing huge problems to schedule reliability. As transportation demand increases various road networks are facing increasing congestion. To mitigate the high-density congestion transit signal priority is a significant solution. Transit signal priority provides solutions according to many variables, and it pursues several valuable objectives such as: reduced transit travel times, better schedule adherence, better transit efficiency, and increased road network efficiency by car mobility. The objective is to compare and evaluate existing guidelines and form new guidelines for TSP. Furthermore, the aim of this research is to evaluate the effectiveness of the improvements made on the public transportation bus lines by applying transit signal priorities on the corridor. Improvements are reflected by comparing the results of total travel time and total delay on the same single corridor with and without TSP applied. Results show significant improvements in reducing the travel times and delays for the buses made by applying TSP. Cost of installing the Transit Signal Priorities were not taken in consideration in this study.
Dynamic Modeling and Optimization of Bus Bridging System During Metro Disruption
Hua HuShow Abstract
Jie Yu, University of Wisconsin, Milwaukee
Xiao-bing Ding, Shanghai University
Bus bridging refers to transporting affected passengers to their destination stations in case of metro operational disruption. It plays an important role in emergency transportation management and can significantly improve the level of service of the metro system if properly planned and operated. This paper develops a dynamic model for bus bridging system optimization, in which the interaction among time-varying locations of buses, passenger demand, the arrival and departure time of bridging bus fleet, and the number of debarking and boarding passengers at each disrupted metro station are modeled in details. A simulation-based genetic algorithm is developed to solve the proposed model. Case study results demonstrate the effectiveness of the proposed model and the solution algorithm in real-world application.
Bayesian Inference of Passenger Boarding Strategies at Express Stops with Real-Time Bus Arrival Information
Neema Nassir, Massachusetts Institute of Technology (MIT)Show Abstract
Jinhua Zhao, Massachusetts Institute of Technology (MIT)
John Attanucci, Massachusetts Institute of Technology (MIT)
Frederick Salvucci, Massachusetts Institute of Technology (MIT)
Nigel Wilson, Massachusetts Institute of Technology (MIT)
Efficient design of express and local bus services in urban corridors requires accurate understanding of the travel demand and heterogeneities in passengers’ preferences and needs. Public transit Automated Fare Collection (AFC) systems provide a high-coverage source of data that facilitates an unprecedented opportunity for understanding the demand patterns and passenger preferences for more efficient service designs.
In this paper, a Bayesian inference method is proposed to analyze the AFC repeated boarding records of passengers in the presence of real-time bus arrival information. A continuous representation of boarding strategies is introduced that can capture the behavior of passengers if they extend their waiting times to board a preferred route that is due shortly. The proposed method is tested in a case study on the Western Avenue corridor in Chicago, IL. The case study demonstrates the possibility of making confident inferences (95%) for thousands of the corridor passengers. The case study also confirms intuitive correlation of the inferred strategies with variables such as travel distance, egress distance, time of day, and availability of countdown timers at the stop. Potential biases of the inference sample and possible applications in service planning are discussed.
Incorporating the Impact of Spatiotemporal Interactions on Bus Ridership
Moshiur Rahman, University of Central FloridaShow Abstract
Shamsunnahar Yasmin, University of Central Florida
Ahmadreza Faghih-Imani, University of Toronto
Naveen Eluru, University of Central Florida
The over-reliance on the automobile in the US over last few decades resulted in traffic congestion and crashes, air pollution associated environmental and health concerns, and dependence on foreign fuel. To counter the problem, policy makers and transportation professionals are promoting public transit and non-motorist travel modes through infrastructure investments. An important tool to evaluate the influence of these public transit investments on transit ridership is the application of statistical models. Drawing on stop level boarding and alighting data for 6 four-month periods for Greater Orlando region from May 2013 to April 2015, the current study estimates a spatial panel model that accommodates for impact of spatial and temporal observed and unobserved factors. Two spatial models: 1) Spatial Error Model (SEM) and 2) Spatial Lag Model (SAR) are estimated for boarding and alighting separately by employing several exogenous variables including stop level attributes, transportation and transit infrastructure variables, built environment and land use attributes, sociodemographic and socioeconomic variables in the vicinity of the stop and spatial and spatio-temporal lagged variables. The model estimation results are further augmented by a validation exercise. These models are expected to provide feedback to agencies on the benefits of public transit investments while also providing lessons to improve the investment process.
Provision of Bus Real-Time Information: Making Passengers from the Contributor of Headway Irregularity to the Controller
Pengfei Wang, Southeast UniversityShow Abstract
Wendong Chen, Southeast University
Lele Yu, Southeast University
Da Lei, Southeast University
In frequently serviced bus routes passengers are more concerned about bus headway regularity. The buses tend to bunch naturally partly because of the participants of passengers in the system. Thus, the irregularity of headway and bunching of routes with high passenger demand are extremely serious. With the development of technique, bus real-time information including arrival times and passenger loads is possible to provide for passengers. A special ‘control way’ which is particular adaptive to the high-passenger-demand routes is proposed based on this background. It is thought that passengers would control the headway regularity themselves when provided bus real-time information. To test the effects of this special ‘control way’, bus-focused stochastic models are used for the simulation. The results show that providing real-time information does make a better headway regularity as expected. The impacts are significant in morning peak simulation. There is a reduction of 30.1% of headway irregularity after providing real-time information in morning peak, as well as, 14.3% in off-peak. Besides, through a comparison of line without passenger and normal line, it is confirmed that passengers themselves are a vital factor contributing to the headway irregularity. By providing real-time information the moderating effects of passengers basically offset the negative effects from them both in morning peak and off-peak. At last, as expected, passenger loads among buses become more even both in morning peak and in off-peak for a reduction of 49.3% and 39.5% of the standard deviation of passenger loads, respectively.
Identification of Control Group Sites with Spatial Clustering in Bus Rapid Transit Infrastructure Projects
Svitlana Pyrohova, University of QueenslandShow Abstract
Mark Hickman, University of Queensland
Kelly Bertolaccini, University of Queensland
Gavin Turrell, Catholic University, Australia
Studies that investigate land use impacts from transport infrastructure often rely on a quasi-experimental design approach. This method aims to mimic true experiments by drawing comparisons between test and control group samples. In transport and land use research, these groups usually correspond to areas affected by a change in transport infrastructure or land use and areas that are not affected by those changes. The selection of proper control sites is based on various criteria, such as demographic variables (income, age, household size), transport variables (accessibility, travel time), and/or land use variables (zoning, density, mix of land use). This paper aims to develop a methodological approach to the selection of control group sites on the grounds of socio-economic and geographical characteristics in a case study of the bus rapid transit (BRT) system in Brisbane, Australia. This work reviews the previous literature on criteria selection for control groups and further develops a statistical model to determine criteria for this specific study. On that basis, the model assigns a similarity index to every site in the dataset. It outputs cluster groups of similar values that occur to be the most suitable for validation across the full urban area.
A Comprehensive Approach to Transfer Synchronization in Public Transit
Zhongyun WuShow Abstract
Kun An, Monash University
Wei Wang, Southeast University
Graham Currie, Monash University
The dynamic, stochastic and uncertain characters of traffic often make bus schedules erratic and therefore the planned synchronized public transit transfers do not always materialize in practice. This research aims at reducing the operational deviation of buses, and thus increasing arrival reliability of vehicles as well as improving the occurrence possibility of the direct transfer. To achieve this objective, a mathematical model is developed to calculate the recommended parameters of real-time bus operation strategies according to real-time arriving time deviation data. A case study is conducted to demonstrate the effectiveness of the proposed approach in practice. Results illustrate that adopting the approach reduces the probability of bus bunching from 8.33% to 0, and increases the operational reliability of vehicles, although transit vehicle has a larger delay. Moreover, direct transfer can be guaranteed with our proposed approach.
Characteristics of Bus Transit Vehicles in the United States: A 30-Year National Trend Analysis
Li Tang, Florida International UniversityShow Abstract
Albert Gan, Florida International University
Fabian Cevallos, Florida International University
Priyanka Alluri, Florida International University
In 2015, transit agencies in the United States spent over 60% of their bus transit capital funds on their revenue vehicles. Using 30 years of data from the National Transit Database (NTD), this paper examines the national trends of seven major characteristics associated specifically with bus revenue vehicles. These trends can provide important information on where the market might be heading and aid in the planning decision on transit investments. The characteristics examined include number of vehicles, spare ratio, average age, average capacity, ADA accessibility, vehicle reliability, and vehicle operations and maintenance expenses. Some findings from the trend data include: (1) a steady increase in the privatization of bus services; (2) the average spare ratios have consistently exceeded the maximum of 20%, as suggested by the Federal Transit Administration for systems operating with more than 50 vehicles; (3) vehicles operated by contractors tended to be significantly newer than those operated directly by transit agencies, although the gap has narrowed in recent years; (4) vehicles operated directly by transit agencies tended to have higher seating and standing capacities than those operated by contractors; (5) there was a tendency among transit agencies to trade seats for more standing room; (6) by 2015, nearly all bus vehicles were ADA-compliant, and increasingly vehicles with lifts were converted to vehicles with ramps/low-floor; (7) reliability of vehicles in terms of number of mechanical failures per million revenue vehicle miles has significantly improved over the years; and (8) contractors spent significantly less on average than transit agencies in operating and maintaining their vehicles.
Accessibility Impacts of Bus Access to Managed Lanes
Kristin Carlson, University of Minnesota, Twin CitiesShow Abstract
Andrew Owen, University of Minnesota, Twin Cities
This paper introduces a method to measure changes in transit accessibility resulting from adjustments in bus-highway interactions. Operational differences between general purpose (GP) and managed lanes (ML) are measured using average travel time. Changes to transit travel time are systematically introduced to General Transit Feed Specification (GTFS) data through the use of the StopTimesEditor computer program developed for the purpose of this analysis. The methodology is tested on two express bus routes in the Minneapolis - St. Paul region (Twin Cities). The change in operating speed along portions of the selected transit routes is translated to changes in the job accessibility of the surrounding communities. The percent change in the worker-weighted average job accessibility for the area surrounding the transit routes and for the entire metropolitan region are 12% and 0.25% respectively. The methods introduced in this study can be used to evaluate the accessibility impacts of different highway operating environments for buses, or estimate the accessibility outcomes of different bus-highways scenarios.
Evaluation and Setting Condition of Bus Lane Based on Accessibility: A Case Study of Shanghai, China
Yi Jing, Tongji University
Hangfei Lin, Tongji University
Haipeng Cui, Tongji University