Automatic Passenger Counter Data Research Articles

Impact of a lower demand during the COVID-19 pandemic on the frequency of dwell time delays

This study investigates the relationship between changes in the frequency of dwell time delays and passenger flow characteristics and the frequency of railway service during the COVID-19 pandemic. To do so we make use of automatic passenger count data from commuter trains in Southern Sweden. Our findings show that the frequency of dwell time delays decreased under COVID-19 conditions compared to the two years prior, but dwell time delays did not cease to exist altogether. The improvement in dwell time performance is found to occur for delays of 30 s and larger, whereas the occurrence of delays up to 15 s increased. No decline in the service frequency under COVID-19 conditions was found. Overall an improvement in dwell time punctuality can be observed under COVID-19 conditions. The analysis, somewhat, points towards a reduction in peak passenger loads to be one of the factors associated with this improvement. This is, however, not the sole reason as dwell time punctuality is found to improve for all volumes of boarding passengers. The lower passenger demand thus improved dwell time punctuality under COVID-19 conditions, but it does not tell the whole story. When trying to improve dwell time punctuality it is thus not sufficient to only reduce the number of passengers boarding, but other measures are necessary as well.

Improving the Prediction of Passenger Numbers in Public Transit Networks by Combining Short-Term Forecasts With Real-Time Occupancy Data

Passengers of public transportation nowadays expect reliable and accurate travel information. The need for occupancy information is becoming more prevalent in intelligent public transport systems as people started avoiding overcrowded vehicles during the COVID-19 pandemic. Furthermore, public transportation companies require accurate occupancy forecasts to improve service quality. We present a novel approach to improve the prediction of passenger numbers that enhances a day-ahead prediction with real-time data. We first train a baseline predictor on historical automatic passenger counting data. Next, we train a realtime model on the deviations between baseline prediction and observed values, thus capturing events not addressed by the baseline. For the forecast, we attempt to detect emerging patterns in real time and adjust the baseline prediction with deviations from the patterns. Our experiments with data from Germany show that the proposed model improves the forecast of the baseline model and is only outperformed by artificial neural networks in some instances. If the training sets only cover a limited period of up to four months, our approach outperforms competing methods. For larger training sets, there are mixed results in the sense that for some test cases, certain types of neural networks yield slightly better results, but our method still performs well with less training effort, is explainable along the whole prediction process and can be applied to existing prediction methods.

Estimating on-board passenger comfort in public transport vehicles using incomplete automatic passenger counting data

The prevention of crowding inside buses, trams and trains is an important component of on-board passenger comfort and is central to the provision of good public transport services. In light of the COVID-19 pandemic and the associated significant reduction in public transport patronage and, more importantly, in passenger confidence, the avoidance of crowds by passengers and operators alike becomes even more critical. This is where the provision of information on on-board comfort becomes a necessity. The present study, therefore, proposes a new Kalman filter based estimation scheme for on-board comfort levels, employing historical and current (same-day) non-exhaustive Automatic Passenger Counting data, as well as Automatic Vehicle Locating measurements. The accuracy and reliability of the estimation is, then, evaluated through application to the tramway network of the French city of Nantes. The results suggest that the proposed method is able to deliver good estimation accuracy, both in terms of absolute passenger numbers, but also, more crucially, in terms of on-board comfort Levels of Service.

Transit Quality of Service Assessment Using Smart Data

In this paper we assess the transit quality of service (QoS) from a user’s standpoint, using smart data. A number of bus lines with different characteristics, operating in the Metropolitan Area of Athens, were chosen as a case study. The data used were gathered by an Automatic Passenger Counting (APC) system. APC technologies provide exact temporalized passenger counting along the line for each service, thus assisting to better understand causalities of delays and avoid operational problems. By employing archived APC data from buses running on crosstown routes between 15 January 2019 and 15 April 2019 we conducted a statistical analysis to explore occupancies and assess QoS, including under a social distancing scenario. The passenger distribution along the stops, the bus’s occupancy level, the stops that are maximum occupancy points and their rate of occurrence and, lastly, the passenger’s average trip length during the day and the week are examined.

Forecasting public transit passenger demand: With neural networks using APC data

The implementation of Intelligent Transportation Systems (ITS) as a part of smart mobility is crucial for solving the current problems of the transportation industry. The setting up and maintenance of ITS requires not only the current passenger demand but also the future passenger demand. The future passenger demand can be obtained with time-series forecasting carried out with different techniques. With the advancements in the technological field, modern and more advanced methods of time-series forecasting using deep learning are being preferred over traditional forecasting techniques. However, the research carried out in this regard is quite limited, particularly considering the Indian scenario. Hence this research work focuses on exploring the performance of deep learning forecasting techniques considering the aspects mentioned previously. Here, the forecasting of passenger demand was done with Long Short-Term Memory (LSTM) using the three months Automatic Passenger Counter (APC) data of the Hubballi-Dharwad Bus Rapid Transit System (HDBRTS) as part of a case study. Then the forecasting of passenger demand was also done with Seasonal Autoregressive Integrated Moving Average (SARIMA), and the comparison of the forecasting accuracy of both methods was made using Mean Absolute Error (MAE) and Root Mean Squared Error (RMSE). Furthermore, to validate the results, novel approach has been adopted for the process, by following some more time-series resampled with different time intervals. Study shows that LSTMs will be used satisfactorily in the traffic conditions of developing counties, for forecasting passenger demand using APC data. Study also provides detailed guiding methodologies of advanced methods of passenger forecasting along with conventional ones.

Estimation and Mitigation of Epidemic Risk on a Public Transit Route using Automatic Passenger Count Data

This paper studies the potential spread of infectious disease through passenger encounters in a public transit system using automatic passenger count (APC) data. An algorithmic procedure is proposed to evaluate three different measures to quantify these encounters. The first two measures quantify the increased possibility of disease spread from passenger interaction when traveling between different origin–destination pairs. The third measure evaluates an aggregate measure quantifying the relative risk of boarding at a particular stop of the transit route. For calculating these measures, compressed sensing is employed to estimate a sparse passenger flow matrix planted in the underdetermined system of equations obtained from the APC data. Using the APC data of Route 5 in Minneapolis/St. Paul region during the COVID-19 pandemic, it was found that all three measures grow abruptly with the number of passengers on board. The passenger contact network is densely connected, which further increases the potential risk of disease transmission. To reduce the relative risk, it is proposed to restrict the number of passengers on-board and analyze the effect of this using a simulation framework. It was found that a considerable reduction in the relative risk can be achieved when the maximum number of passengers on-board is restricted below 15. To account for the reduced capacity and still maintain reasonable passenger wait times, it would then be necessary to increase the frequency of the route.

A two‐stage method for bus passenger load prediction using automatic passenger counting data

In high-frequency transit, providing real-time crowding information (RTCI) is a potential way to promote passenger satisfaction and reduce negative crowding externalities, by assisting passengers in choosing less crowded vehicles. To make RTCI convincing and reliable, it is necessary to provide predictive RTCI, in which bus passenger load (BPL) prediction is the primary problem. This paper proposes a novel two-stage BPL prediction method using automatic passenger counting (APC) data. The first stage is to predict short-term passenger flows at stops based on an adaptive Kalman filter approach. Using the outputs from the first stage as well as other variables directly from APC data, the second stage is to predict BPL based on a support vector regression algorithm. Several methods from the existing literature are used as benchmarks to test the relative performance of the proposed method. An empirical study on bus line 1 in Suzhou, China shows that the proposed method outperforms all the benchmarks, and shows significant superiority over other methods for stops with sharp increases in BPL and for multi-step ahead prediction. This study contributes to the limited literature on BPL prediction and lays the foundation for providing accurate and reliable predictive RTCI in the future.

Understanding Human Perception of Bus Fullness: An Empirical Study of Crowdsourced Fullness Ratings and Automatic Passenger Count Data

Bus fullness, namely bus crowding level, is one of the most critical metrics used to quantify public transportation service quality and consumer satisfaction. Often it is simply represented by the head count within a bus vehicle that is measured using the prevailing automatic passenger counter (APC)., Little is known, however, about how the precise passenger count reflects the human perception of bus fullness. This study examines the linkage between APC data and crowdsourced fullness ratings data from riders in Pittsburgh, U.S.A., to understand how riders’ perception of bus fullness is related to spatial, temporal, and demographic factors in addition to the passenger counts. We found that human perception of bus fullness, matched with passenger counts, can vary substantially by time of day, bus vehicle seat capacity, and passengers’ income. Last but not least, we proposed a statistical model that estimates riders’ perception of bus fullness based on APC data and bus route characteristics. This model can be used by transit agencies who possess APC data to assess service quality better, and ultimately to enhance the design and operation of transit systems.

Identifying the Bias: Evaluating Effectiveness of Automatic Data Collection Methods in Estimating Details of Bus Dwell Time

Data from automated vehicle location (AVL) systems, automatic passenger counter (APC) systems, and fare box payments have been heavily used to generate dwell time models with the goal of recommending improvements in efficiency and reliability of bus transit systems. However, automatic data collection methods may result in a loss of detail with regard to the dynamics of passenger activity, which may bias the estimates associated with dwell or passenger activity time. The purpose of this study is to understand better any biases that might exist from using data from AVL–APC systems or fare box payments when estimating dwell time. Manually collected data from Montreal, Quebec, Canada, are used to estimate detailed dwell time models. This study compared those estimates to models generated by using data similar to what was reported by AVL–APC systems and fare boxes. The results reveal an overestimation in the passenger activity component of dwell time, which is mainly attributed to excess dwell time that AVL–APC data and fare box payments generally do not capture. While AVL–APC and fare box technologies provide transit agencies with rich data for analysis, adjustments to such data collection methods are warranted to reduce the overestimation of dwell time and to provide a more accurate picture of what is happening on the ground to generate better interventions that can reduce dwell times.

Use of Mobile Device Wireless Signals to Determine Transit Route-Level Passenger Origin–Destination Flows: Methodology and Empirical Evaluation

Directly observing stop-to-stop origin–destination (O-D) flows for transit passengers by onboard surveys is time-consuming and costly. While the use of automatic passenger count (APC) and automatic fare collection data to infer transit O-D flows has undergone important developments lately, the value of directly observing O-D movements remains of interest in light of inference errors. The increasing use of personal mobile devices and the corresponding reliance on wireless fidelity (Wi-Fi) communications in public spaces allow for the prospect of using the detected anonymous Wi-Fi signals of mobile devices carried by transit passengers to extract O-D movements. An empirical study is conducted when O-D movements determined directly from Wi-Fi signal data and from a combination of Wi-Fi and APC data are compared to extensive, directly observed high-fidelity ground truth data obtained from an onboard survey. While O-D flows determined directly from the Wi-Fi data for a specific bus trip do not correspond well to the ground truth flows, the results demonstrate the promise of using Wi-Fi signal data for O-D flow determination when the data are aggregated across multiple bus trips for a time-of-day period, especially when used in conjunction with APC data.

Perspectives on Transit: Potential Benefits of Visualizing Transit Data

Advancements in information and communication technologies have enabled transit agencies around the world to generate streams of data on a high-frequency basis. Increasingly, these agencies are interested in new methods of visualizing these data to communicate the results of their planning efforts, operational investments, and overall transit performance to decision makers and stakeholders. Most agencies today collect and provide numerous kinds of data, including Google’s general transit feed specification schedule data, automatic vehicle location data, and automatic passenger count data. This paper aims to demonstrate the untapped potential of these data sources; specifically, the paper uses transit data from Montreal, Quebec, Canada, to generate performance measures that are of interest to both transit planners and marketing professionals. Some of these measures can also help in communicating the positive attributes of public transportation to the community. Performance measures are generated at different scales, including transit system, neighborhood, route, and stop levels. This paper expands on previous research on transit performance research and visualization by adopting currently available resources for so-called big data.

Transit passenger origin–destination flow estimation: Efficiently combining onboard survey and large automatic passenger count datasets

As transit agencies increasingly adopt the use of Automatic Passenger Count (APC) technologies, a large amount of boarding and alighting data are being amassed on an ongoing basis. These datasets offer opportunities to infer good estimates of passenger origin–destination (OD) flows. In this study, a method is proposed to estimate transit route passenger OD flow matrices for time-of-day periods based on OD flow information derived from labor-intensive onboard surveys and the large quantities of APC data that are becoming available. The computational feasibility of the proposed method is established and its accuracy is empirically evaluated using differences between the estimated OD flows and ground-truth observations on an operational bus route. To interpret the empirical differences from the ground-truth estimates, differences are also computed when using the state-of-the-practice Iterative Proportional Fitting (IPF) method to estimate the OD flows. The empirical results show that when using sufficient quantities of boarding and alighting data that can be readily obtained from APC-equipped buses, the estimates determined by the proposed method are better than those determined by the IPF method when no or a small sample sized onboard OD flow survey dataset is available and of similar quality to those determined by the IPF method when a large sample sized onboard OD flow survey dataset is available. Therefore, the proposed method offers the opportunity to forgo conducting costly onboard surveys for the purpose of OD flow estimation.

Statistical inference on transit route‐level origin–destination flows using automatic passenger counter data

SummaryWe consider inferring transit route‐level origin–destination (OD) flows using large amounts of automatic passenger counter (APC) boarding and alighting data based on a statistical formulation. One critical problem is that we need to enumerate the OD flow matrices that are consistent with the APC data for each bus trip to evaluate the model likelihood function. The OD enumeration problem has not been addressed satisfactorily in the literature. Thus, we propose a novel sampler to avoid the need to enumerate OD flow matrices by generating them recursively from the first alighting stop to the last stop of the bus route of interest. A Markov chain Monte Carlo (MCMC) method that incorporates the proposed sampler is developed to simulate the posterior distributions of the OD flows. Numerical investigations on an operational bus route under a realistic OD structure demonstrate the superiority of the proposed MCMC method over an existing MCMC method and a state‐of‐the‐practice method. Copyright © 2015 John Wiley & Sons, Ltd.

Method for Assessing Bus Delay in Mixed Traffic to Identify Transit Priority Improvement Locations in Cambridge, Massachusetts

Urban transit services face a number of challenges from space constraints, congestion, and delays, among other issues. Implementing bus priority at traffic signals or providing exclusive operating space for buses can increase the attractiveness of taking the bus and thereby encourage ridership. The City of Cambridge, Massachusetts, was looking to pilot such interventions to demonstrate benefits of bus ridership, but needed a prioritized list of route segments with the largest levels of excess travel time to do so. Three metrics were used to evaluate delay: vehicle delay, overall passenger delay, and system reliability. These three metrics were combined into a single composite rating system for each segment and used to identify route segments along which buses experienced the most delay. The City of Cambridge analyzed high-ridership bus routes with automatic passenger counter data to identify segments along the route where buses experienced substantial delay. The next step for this project is to conduct an on-site field visit of targeted segments to develop potential bus prioritization proposals for each. This paper outlines the method developed to calculate bus delay by segment and presents results for one route analyzed.

Exploring Applications of Second-Generation Archived Transit Data for Estimating Performance Measures and Arterial Travel Speeds

Travel time and operating speed influence the attractiveness, operating cost, and system efficiency of transit service. As part of its bus dispatch system, the Tri-County Metropolitan Transportation District of Oregon has been archiving automatic vehicle location (AVL) and automatic passenger count data for all bus trips at the stop level since 1997. In 2013, a new and higher-resolution bus AVL data collection system was implemented. This new system provides stop-level data as well as 5-s resolution (5-SR) bus position data between stops. The objective of this paper is to explore potential applications of the new data for assessing transit performance and for estimating transportation system performance measures for urban arterials. Results suggest that the 5-SR data provide high-resolution information on time and position that can be used to determine bus travel speeds between bus stops, identify speed breakdowns, and estimate intersection signal and queuing delays.

An Offline Framework for Handling Automatic Passenger Counting Raw Data

Knowledge of ridership data on bus routes is pivotal for the quality and efficient operational planning of public transport companies. Automatic passenger counting (APC) can represent a powerful resource for supporting this activity, because it can provide a databank of accurate counts. However, relevant challenges, such as the matching of data to the bus stop, data validation, tackling anomalies, and building intelligible performance reports, must be faced in order to make APC data a mainstream source of information. This paper proposes an offline framework for addressing these challenges. In order to illustrate a possible application of the framework, its use for setting bus frequencies is investigated. The results are represented by easy-to-read control dashboards composed of tables and graphs. The methodology is experimentally tested with data records provided by the bus operator CTM in Cagliari, Italy. Finally, we discuss the implications on service rearrangement.

Estimating Transit Route OD Flow Matrices from APC Data on Multiple Bus Trips Using the IPF Method with an Iteratively Improved Base: Method and Empirical Evaluation

An iterative method is proposed to estimate bus route origin-destination (OD) passenger flow matrices from boarding and alighting data for time-of-day periods in the absence of good a priori estimates of the flows. The algorithm is based on the widely used iterative proportional fitting (IPF) method and takes advantage of the large quantities of boarding and alighting data that are routinely collected by transit agencies using automatic passenger count (APC) technologies. An arbitrarily chosen OD matrix can be used as the base matrix required to initialize the algorithm, and the IPF method is applied with bus trip-level boarding and alighting data and the base matrix to produce an estimate of the OD flow matrix for each bus trip. The trip-level OD flow matrices are then aggregated to produce an estimate of the period-level OD flow matrix, which in turn is used as the base matrix for the following iteration. The process is repeated until convergence. Empirical results are conducted on operational bus routes using APC data collected during multiple season-years, where directly observed OD passenger flows are available to represent the ground truth. In all cases in which APC data are available for even a reasonably small number of bus trips, the iteratively improved base method produces better estimates than the application of the traditional IPF method when using a null base matrix, which is commonly adopted in the absence of a priori information without updating. Moreover, the algorithm converges in minimal computational time to the same estimates regardless of the initializing matrices used.

All aboard at all Doors

The time that buses spend waiting for passengers to board can be a significant portion of a bus route's overall running time. A key determinant of boarding time is the number of doors through which passengers are permitted to board. Transit agencies that allow boarding through all doors, instead of just through the front door, typically enjoy decreased boarding times and decreased running times. This study focused on the feasibility of an all-door boarding policy for La Société de transport de Montréal (STM), the public transit agency of Montreal, Canada. The potential benefits of such a policy were assessed through three main steps. First, a selection methodology was developed to determine which of STM's bus routes would benefit most from various all-door boarding strategies. Second, a multivariate regression analysis was implemented with STM's archived automatic vehicle location and automatic passenger counter data to estimate the dwell and running-time savings that would result under various implementation scenarios. Third, a sensitivity analysis was developed to demonstrate the savings associated with implementing the policy. The findings showed that all-door boardings could yield substantial savings in running time, with morning peak savings as much as 15.8% on the best routes. In many cases, the running-time savings were enough to remove a bus from a route while still maintaining existing frequencies. The findings from this research may be beneficial for transit planners and operators since the presented methodologies show substantial savings from all-door boarding and can be adopted by other transit agencies.

Statistical Study of the Impact of Adaptive Traffic Signal Control on Traffic and Transit Performance

The Sydney Coordinated Adaptive Traffic System (SCATS) is designed to improve mobility on congested corridors with variable demand; SCATS is expected to improve transit performance. Because many urban corridors are heavily used by transit routes, it is important to determine how SCATS affects transit performance on congested corridors. However, there is limited research on evaluating the impact of SCATS on transit performance. To the best of the authors’ knowledge, this research presents the first field evaluation of SCATS with transit signal priority in the United States. The case study is based on before-and-after traffic and transit data along Powell Boulevard, one of the most congested urban arterials in the Portland, Oregon, metropolitan region. Using permanent traffic data collection stations, bus automated vehicle location and automatic passenger count data, and transit signal priority request data, this work presents the results of statistical tests and regressions to determine the impacts of SCATS. Statistically significant differences were observed in travel times and SCATS-related regression parameters. Overall, travel time changes or improvements related to SCATS seemed to depend greatly on the direction of travel and time of day.

Statistical Study of the Impact of Adaptive Traffic Signal Control on Traffic and Transit Performance

The Sydney Coordinated Adaptive Traffic System (SCATS) is designed to improve mobility on congested corridors with variable demand; SCATS is expected to improve transit performance. Because many urban corridors are heavily used by transit routes, it is important to determine how SCATS affects transit performance on congested corridors. However, there is limited research on evaluating the impact of SCATS on transit performance. To the best of the authors’ knowledge, this research presents the first field evaluation of SCATS with transit signal priority in the United States. The case study is based on before-and-after traffic and transit data along Powell Boulevard, one of the most congested urban arterials in the Portland, Oregon, metropolitan region. Using permanent traffic data collection stations, bus automated vehicle location and automatic passenger count data, and transit signal priority request data, this work presents the results of statistical tests and regressions to determine the impacts of SCATS. Statistically significant differences were observed in travel times and SCATS-related regression parameters. Overall, travel time changes or improvements related to SCATS seemed to depend greatly on the direction of travel and time of day.