Bus arrival time prediction at bus stop with multiple routes

Provision of accurate bus arrival information is vital to passengers for reducing their anxieties and waiting times at bus stop. This paper proposes models to predict bus arrival times at the same bus stop but with different routes. In the proposed models, bus running times of multiple routes are used for predicting the bus arrival time of each of these bus routes. Several methods, which include support vector machine (SVM), artificial neural network (ANN), k nearest neighbours algorithm (k-NN) and linear regression (LR), are adopted for the bus arrival time prediction. Observation surveys are conducted to collect bus running and arrival time data for validation of the proposed models. The results show that the proposed models are more accurate than the models based on the bus running times of single route. Moreover, it is found that the SVM model performs the best among the four proposed models for predicting the bus arrival times at bus stop with multiple routes.     

Estimating bus run times for new limited-stop service using archived AVL and APC data

In recent years, several transit agencies have been trying to be more competitive with the automobile to attract choice riders. Transit agencies can only be competitive if they can provide services that are reliable, have a short access and egress time, and have run times that are comparable to automobiles. Several transit agencies try to be competitive through offering faster service, such as limited-stop (express) bus service. This study uses AVL and APC data, in addition to a disaggregate data obtained from a travel behavior survey, to select stops and estimate run times for a new limited-stop service that will run parallel to a heavily used bus route (67 Saint-Michel) in Montréal, Canada. Three different scenarios are developed based on theory and practice to select stops to be incorporated in the new limited service. The time savings for each scenario are then evaluated as a range and a fourth scenario is developed. A limited-stop service is recommended based on selecting stops serving both directions of the route, major activity points and stop spacing. This study shows that implementing a limited-stop service would yield substantial time savings for both, the new limited service and the existing regular service running in parallel.     

Understanding the impacts of a combination of service improvement strategies on bus running time and passenger’s perception

Transit agencies implement many strategies in order to provide an attractive transportation service. This article aims to evaluate the impacts of implementing a combination of strategies, designed to improve the bus transit service, on running time and passenger satisfaction. These strategies include using smart card fare collection, introducing limited-stop bus service, implementing reserved bus lanes, using articulated buses, and implementing transit signal priority (TSP). This study uses stop-level data collected from the Société de transport de Montréal (STM)’s automatic vehicle location (AVL) and automatic passenger count (APC) systems, in Montréal, Canada. The combination of these strategies has lead to a 10.5% decline in running time along the limited stop service compared to the regular service. The regular route running time has increased by 1% on average compared to the initial time period. The study also shows that riders are generally satisfied with the service improvements. They tend to overestimate the savings associated with the implementation of this combination of strategies by 3.5-6.0 min and by 2.5-4.1 min for both the regular route and the limited stop service, respectively. This study helps transit planners and policy makers to better understand the effects of implementing a combination of strategies to improve running time and passenger’s perception of these changes in service.     

Bus congestion, optimal infrastructure investment and the choice of a fare collection system in dedicated bus corridors

Microeconomic optimisation of scheduled public transport operations has traditionally focused on finding optimal values for the frequency of service, capacity of vehicles, number of lines and distance between stops. In addition, however, there exist other elements in the system that present a trade-off between the interests of users and operators that have not received attention in the literature, such as the optimal selection of a fare payment system and a designed running speed (i.e., the cruising speed that buses maintain in between two consecutive stops). Alternative fare payment methods (e.g., on-board and off-board, payment by cash, magnetic strip or smart card) have different boarding times and capital costs, with the more efficient systems such as a contactless smart card imposing higher amounts of capital investment. Based on empirical data from several Bus Rapid Transit systems around the world, we also find that there is a positive relationship between infrastructure cost per kilometre and commercial speed (including stops), achieved by the buses, which we further postulate as a linear relationship between infrastructure investment and running speed. Given this context, we develop a microeconomic model for the operation of a bus corridor that minimises total cost (users and operator) and has five decision variables: frequency, capacity of vehicles, station spacing, fare payment system and running speed, thus extending the traditional framework. Congestion, induced by bus frequency, plays an important role in the design of the system, as queues develop behind high demand bus stops when the frequency is high. We show that (i) an off-board fare payment system is the most cost effective in the majority of circumstances; (ii) bus congestion results in decreased frequency while fare and bus capacity increase, and (iii) the optimal running speed grows with the logarithm of demand.     

Comparing static and dynamic threshold based control strategies

This research extends a static threshold based control strategy used to control headway variation to a dynamic threshold based control strategy. In the static strategy, buses are controlled by setting a threshold value that holds buses at a control point for a certain amount of time before allowing the bus to continue along the route. The threshold remains constant each time the bus stops at the control point. The dynamic strategy involves the same principle of holding buses at a bus stop; however, a different threshold value is chosen each time the bus holds at a control point. The results indicate that in cases where the static threshold is set equal to the scheduled headway, very low headway variation and passenger system times result; however, passengers on board the bus are penalized by extra delay on the bus while waiting at the control point. The dynamic strategy reduces the penalty to passengers delayed on-board the bus at a control point at the expense of a slight increase in overall passenger system time.The results indicate that in most cases, the tradeoff of the slight increase in waiting time for the significant decrease in on-board delay penalty makes the dynamic strategy an acceptable choice.     

A regression model for bus running times in suburban areas of winnipeg

To plan new bus routes in suburban areas, expected bus running times on these routes are needed. Using most readily available relevant variables, a regression model is developed for estimating bus running times. The model is conceptually reasonable and it was tested using data other than that used for estimation.     

Bus arrival time calculation model based on smart card data

Bus arrival time is usually estimated using the boarding time of the first passenger at each station. However, boarding time data are not recorded in certain double-ticket smart card systems. As many passengers usually swipe the card much before their alighting, the first or the average alighting time cannot represent the actual bus arrival time, either. This lack of data creates difficulties in correcting bus arrival times. This paper focused on developing a model to calculate bus arrival time that combined the alighting swiping time from smart card data with the actual bus arrival time by the manual survey data. The model was built on the basis of the frequency distribution and the regression analysis. The swiping time distribution, the occupancy and the seating capacity were considered as the key factors in creating a method to calculate bus arrival times. With 1011 groups of smart card data and 360 corresponding records from a manual survey of bus arrival times, the research data were divided into two parts stochastically, a training set and a test set. The training set was used for the parameter determination, and the test set was used to verify the model’s precision. Furthermore, the regularity of the time differences between the bus arrival times and the card swiping times was analyzed using the “trend line” of the last swiping time distribution. Results from the test set achieved mean and standard error rate deviations of 0.6% and 3.8%, respectively. The proposed model established in this study can improve bus arrival time calculations and potentially support state prediction and service level evaluations for bus operations.     

Evaluate bus emissions generated near far-side and near-side stops and potential reductions by ITS: An empirical study

In this paper, we use second-by-second automatic vehicle location data to estimate bus emissions near far-side and near-side stops. We classify the bus running state near a stop into approach, dwell, and departure. A vehicle specific power approach is used to estimate bus emissions for each state. We show that bus emissions generated near stops can be significantly reduced by using certain intelligent transportation systems techniques.     

Evaluating the added-value of online bus arrival prediction schemes

Online predictions of bus arrival times have the potential to reduce the uncertainty associated with bus operations. By better anticipating future conditions, online predictions can reduce perceived and actual passenger travel times as well as facilitate more proactive decision making by service providers. Even though considerable research efforts were devoted to the development of computationally expensive bus arrival prediction schemes, real-world real-time information (RTI) systems are typically based on very simple prediction rules. This paper narrows down the gap between the state-of-the-art and the state-of-the-practice in generating RTI for public transport systems by evaluating the added-value of schemes that integrate instantaneous data and dwell time predictions. The evaluation considers static information and a commonly deployed scheme as a benchmark. The RTI generation algorithms were applied and analyzed for a trunk bus network in Stockholm, Sweden. The schemes are assessed and compared based on their accuracy, reliability, robustness and potential waiting time savings. The impact of RTI on passengers waiting times are compared with those attained by service frequency and regularity improvements. A method which incorporates information on downstream travel conditions outperforms the commonly deployed scheme, leading to a 25% reduction in the mean absolute error. Furthermore, the incorporation of instantaneous travel times improves the prediction accuracy and reliability, and contributes to more robust predictions. The potential waiting time gains associated with the prediction scheme are equivalent to the gains expected when introducing a 60% increase in service frequency, and are not attainable by service regularity improvements.     

An improved headway-based holding strategy for bus transit

Abstract

This paper presents an improved headway-based holding strategy integrating bus transit travel and dwelling time prediction. A support vector machine-based (SVM) model is developed to predict the baseline travel and dwell times of buses based on recent data. In order to reduce prediction errors, an adaptive algorithm is used together with real-time bus operational information and estimated baseline times from SVM models. The objective of the improved holding strategy is to minimize the total waiting times of passengers at the current stop and at successive stops. Considering the time-varying features of bus running, a ‘forgetting factor’ is introduced to weight the most recent data and reduce the disturbance from unexpected incidents. Finally, the improved holding strategy proposed in this study is illustrated using the microscopic simulation model Paramics and some conclusions are drawn.     

An application-oriented model of passenger waiting time based on bus departure time intervals

This paper develops an application-oriented model to estimate waiting times as a function of bus departure time intervals. Bus stops are classified into Type A and B depending on whether they are connected with urban rail transit systems. Distributions of passenger arrival rates are analyzed based on field data for Beijing. The results indicate that the best fits for the distribution of passenger arrival rates for Type A and B bus stops are the lognormal distribution and gamma distribution, respectively. By analyzing relationships between passenger arrival rates and bus departure time intervals, it is demonstrated that parameters of the passenger arrival rate distribution can be expressed by the average and coefficient of variation of bus departure time intervals in functional relationships. The validation shows that the model provides a reliable estimation of the average passenger waiting time based on readily available bus departure time intervals.     

Rethinking bus punctuality by integrating Automatic Vehicle Location data and passenger patterns

This paper investigates punctuality at bus stops. Although it is typically evaluated from the point of view of bus operators, it must also account for users, as required in recent service quality norms. Therefore, evaluating punctuality at bus stops is highly important, but may also be a complex task, because data on both bus arrivals (or departures) and users must be taken into account and processed. Data on buses can be collected by Automatic Vehicle Location (AVL) systems, but several challenges must be addressed in order to use them effectively. Passengers data at bus stops cannot be derived from AVL, but they can be used to derive passenger patterns and need to be integrated into processed AVL data. This paper proposes a new punctuality measure defined as the fraction of passengers who will be served within an acceptably short interval after they arrive. A method is proposed to determine this measure: it provides (i) several rules to handle AVL collected data, (ii) a procedure integrating processed AVL data and potential passengers’ patterns and (iii) a hierarchical process to perform the punctuality measure on each bus route direction of a transit network, as well as for every bus stop and time period. The paper illustrates the experimentation of this method on more than 4,000,000 data of a real bus operator and represents outcomes by easy-to-read control dashboards.     

Optimizing coordinated transfer with probabilistic vehicle arrivals and passengers' walking time

Supporting efficient connections by synchronizing vehicle arrival time and passengers' walking time at a transfer hub may significantly improve service quality, stimulate demand, and increase productivity. However, vehicle travel times and walking times in urban settings often varies spatially and temporally due to a variety of factors. Nevertheless, the reservation of slack time and/or the justification of vehicle arrival time at the hub may substantially increase the success of transfer coordination. To this end, this paper develops a model that considers probabilistic vehicle arrivals and passengers walking speeds so that the slack time and the scheduled bus arrival time can be optimized by minimizing the total system cost. A case study is conducted in which the developed model is applied to optimize the coordination of multiple bus routes connecting at a transfer station in Xi'an, China. The relationship between decision variables and model parameters, including the mean and the standard deviation of walking time, is explored. It was found that the joint impact of probabilistic vehicle arrivals and passengers' walking time significantly affects the efficiency of coordinated transfer. The established methodology can essentially be applied to any distribution of bus arrival and passenger walking time. Copyright © 2017 John Wiley & Sons, Ltd.     

A model of bus bunching under reliability-based passenger arrival patterns

If bus service departure times are not completely unknown to the passengers, non-uniform passenger arrival patterns can be expected. We propose that passengers decide their arrival time at stops based on a continuous logit model that considers the risk of missing services. Expected passenger waiting times are derived in a bus system that allows also for overtaking between bus services. We then propose an algorithm to derive the dwell time of subsequent buses serving a stop in order to illustrate when bus bunching might occur. We show that non-uniform arrival patterns can significantly influence the bus bunching process. With case studies we find that, even without exogenous delay, bunching can arise when the boarding rate is insufficient given the level of overall demand. Further, in case of exogenous delay, non-uniform arrivals can either worsen or improve the bunching conditions, depending on the level of delay. We conclude that therefore such effects should be considered when service control measures are discussed.     

A theoretical travel-time model for flexible-route buses

A theoretical model for estimating the expectation and variance of buses' running times under a flexibly-routed mode of service is proposed. The model is based on a probabilistic concept that adequately accommodates the usual randomness in the number and location of passengers served during successive vehicle trips. A few simplifications are embodied in the model but it can serve as a basis for a more refined model such as a computer simulation model that can be used in designing real bus systems.The physical setting assumed in the model is a rectangular grid road network where all houses face the side streets as in some suburban regions. Because it oversimplifies real-life situations, this assumption would need to be relaxed to make the model applicable to more general cases.Both the partially flexible-route service where some passengers are captive to fixed-route service and others are served at their doorsteps, and the fully flexible-route service where all passengers are served at their doorsteps, are studied. In each case, a very simple routing convention that can be conveniently executed by the bus drivers is assumed. The proposed travel time model confirms the intuitively correct phenomenon that when the concentration of passenger trip-ends is very high, the vehicle-route degenerates into a fixed-route in which the buses visit all possible loading points within the service area.     

Bus ride times of exceptional school children

This paper examines bus ride times of exceptional school children in North Carolina. It uses levels of service data, which show excessive ride times for exceptional school children, to form clusters of local education agencies. From these clusters short term objectives or standards for maximum ride times are determined for each cluster. Among the findings are three distinct clusters of local education agencies (clusters 1, 2 and 3) whose maximum ride times are respectively 170 minutes, 105 minutes and 62 minutes.

The maximum ride times show that only one cluster (cluster 3) provides a service whose maximum ride time is comparable to the ideal maximum ride time (long term objective) of one hour revealed by the literature and also by responses of school superintendents and transportation directors. Methods of achieving these standards are examined. A comparison of the levels of service of regular and exceptional children transportation is also provided.

     

Models of bus boarding and alighting dynamics

Understanding the dynamics of boarding/alighting activities and its impact on bus dwell times is crucial to improving bus service levels. However, research is limited as conventional data collection methods are both time and labour intensive. In this paper, we present the first use of smart card data to study passenger boarding/alighting behaviour and its impact on bus dwell time. Given the nature of these data, we focus on passenger activity time and do not account for the time necessary to open and close doors. We study single decker, double decker and articulated buses and identify the specific effects of floor/entrance type, number of activities and occupancy on both boarding and alighting dynamics. A linear relationship between average boarding and alighting times and their respective standard deviations is also found, whereas the variability of boarding and alighting time decreases with the number of passengers boarding and alighting. After observing the cumulative boarding/alighting processes under different occupancy conditions, we propose a new model to estimate passenger activity time, by introducing critical occupancy – a parameter incorporating the friction between boarding/alighting and on-board passengers. We conduct regression analyses with the proposed and another popular model for simultaneous boarding/alighting processes, finding that the critical occupancy plays a significant role in determining the regime of boarding and alighting processes and the overall activity time. Our results provide potential implications for practice and policy, such as identifying optimal vehicle type for a particular route and modelling transit service reliability.     

Modeling and optimizing urban bus transit considering headway variation for cost and service reliability analysis

Due to the stochastic nature of traffic conditions and demand fluctuations, it is a challenging task for operators to maintain reliable services, and passengers often suffer from longer travel times. A failure to consider this issue while planning bus services may lead to undesirable results, such as higher costs and a deterioration in level of service. Considering headway variation at route stops, this paper develops a mathematical model to optimize bus stops and dispatching headways that minimize total cost, consisting of both user and operator costs. A Genetic Algorithm is applied to search for a cost-effective solution in a real-world case study of a bus transit system, which improves service reliability in terms of a reduced coefficient of variation of headway.     

Disruption management in public transit: the bee colony optimization approach

Disruptions in carrying out planned bus schedules occur daily in many public transit companies. Disturbances are often so large that it is necessary to perform re-planning of planned bus and crew activities. Dispatchers in charge of traffic operations must frequently find an answer to the following question in a very short period of time: How should available buses be distributed among bus routes in order to minimize total passengers' waiting time on the network? We propose a model for assigning buses to scheduled routes when there is a shortage of buses. The proposed model is based on the bee colony optimization (BCO) technique. It is a biologically inspired method that explores collective intelligence applied by honey bees during the nectar collecting process. It has been shown that this developed BCO approach can generate high-quality solutions within negligible processing times.     

Methods for pre-processing smartcard data to improve data quality

In recent years smartcards have been implemented in many transit systems around the world as a means by which passengers pay for travel. In addition to allowing speedier boardings there are many secondary benefits of smartcard systems including better understanding of travel patterns and behaviour of travellers. Such research is dependent on the smartcard correctly recording the boarding stop, and where available the alighting stop. It is also dependent on the smartcard system correctly aggregating individual rides into trips.This paper identifies causes for why smartcard systems may not correctly record such information. The first contribution of the paper is to propose a set of rules to aggregate individual rides into a single trip. This is critical in the research of activity based modelling as well as for correctly charging the passenger. The second contribution of the paper is to provide an approach to identify erroneous tap-out data, either caused by system problems or by the user. An approach to detecting this phenomenon is provided. The output from this analysis is then used to identify faulty vehicles or data supply using the “comparison against peers approach”. This third contribution of the paper identifies where transit agencies and operators should target resources to improve performance of their Automatic Vehicle Location systems. This method could also be used to identify users who appear to be tapping out too early.The approaches are tested using smartcard data from the Singapore public transport network from one week in April 2011. The results suggest that approximately 7.7% of all smartcard rides recorded the passenger as alighting one stop before the bus stop that they most probably alighted at. A further 0.7% of smartcard rides recorded the passenger as alighting more than one stop before the bus stop that they most probably alighted at. There was no evidence that smartcards overestimated the distance travelled by the passenger.