Automatic Vehicle Location Systems Research Articles

Use of High-Resolution Signal Controller Data to Measure Transit Signal Priority Performance: A Case Study in the Boston Region

Automated traffic signal performance measures (ATSPMs) have been developed to organize data within a traffic signal, which will then be used to evaluate and monitor signal timing strategies. Past research has summarized the use of ATSPMs for performance-based management of a traffic signal system. The National Cooperative Highway Research Program Research Report 954 focused on many measures useful for general traffic, but did not consider transit agencies as specific stakeholders. ATSPMs do provide insights into measuring the number of preemption events, and the percentage of transit vehicles arriving on green, but additional information is needed to determine the effectiveness of transit signal priority or other transit preferential treatment strategies. This paper proposes a method for evaluating transit performance at signalized intersections more fully using the high-resolution data available within modern traffic signal controllers. The addition of transit-specific inputs within the ATSPM enumerations is used to produce transit-specific signal performance measures designed to be useful to practitioners. The methodology can be followed by other agencies seeking to enhance the capabilities of existing automatic vehicle location systems to further describe the performance at each signalized intersection. By incorporating transit into the ATSPM system, an agency can better assess the delay of buses, which can contribute to schedule adherence, on-time performance, and overall bus service. The methodology is presented as a prototype for further development and adaptation to individual agency objectives and data sources.

Methodology to Detect Bus Stop Influence Zones Utilizing Facebook Prophet Changepoint Detection Method

Travel time of buses, a major part of the urban public transit system, is affected by various factors and foremost are the bus stops. In addition to the dwell time at the stop, the deceleration and acceleration zones reduce average speed, particularly in mixed traffic, and increase travel time. Several approaches are used in estimating the link-based travel time of public transit systems in transportation planning, but Bus Stop Influence Zones (BSIZs) are ignored. Moreover, fuel consumption and pollutant emissions increase in BSIZs, and this demonstrates the importance of BSIZs in measuring the performance of public transit systems as well as planning. Data obtained from the Automatic Vehicle Location system was used and visualized on Geographic Information System and data preprocessing steps were performed. Finally, changepoint detection method of Facebook Prophet (FBP-CDM) was exploited to identify changepoints in location-speed data on the selected route. Results were validated with real-life data and expert opinion, and then compared with the findings acquired by the K-means clustering. Based on the conclusions, FBP-CDM was found quite effective in detecting and predicting BSIZs accurately and the proposed methodology is useful for studies in transportation planning and operations.

Dynamic Modeling for Metro Passenger Flows on Congested Transfer Routes

With the rapid development of urbanization, the metro becomes more and more important for people’s travel in big cities. To quantitatively describe metro passenger flows on congested transfer routes, this paper introduces a dynamic model based on automated data from the automatic fare collection (AFC) and automatic vehicle location (AVL) systems. An expectation maximization (EM) algorithm is proposed to compute the maximum likelihood estimates of unknown parameters in our model. Our model can yield a systematic analysis of one-transfer passenger flows on both population and individual aspects. Important characteristics, including transfer time, boarding probabilities, walking time, passenger-to-train assignment probabilities, and total travel time, can be inferred using only the AFC and AVL data. We provide a case study on the Beijing metro. Detailed analysis results based on our model are given. We also present a cross-validation method to validate our model with real data.

On the prediction of intermediate-to-long term bus section travel time with the Burr mixture autoregressive model

Travel time is an essential indicator for trip planning, transportation service planning, and operation. This study aims to propose a novel Burr mixture autoregressive (BMAR) model for the intermediate-to-long term period of bus section travel time prediction, which is useful for bus service and schedule planning. The BMAR model exhibits greater flexibility, allowing it to effectively capture the multi-peak and non-peak, non-linear with heteroscedasticity characteristics of travel time. The model is trained and well-validated with 6-month bus section travel time data collected via the automatic vehicle location system. Results show that the BMAR model gives promising results in travel time point and interval prediction, especially for the higher degree of variability and irregular pattern of travel time observed on urban roads and highways. The reliability ratio index derived from the BMAR model could be used to measure the bus service reliability and aid in the bus scheduling of maintenance activities.

Automatic Vehicle Location and Computer-Aided Dispatch Systems: Design and Application Considerations

Automatic Vehicle Location (AVL) systems can track transit vehicles in real time. Computer-Aided Dispatch (CAD) software is used to monitor transit operations and assist management of transit operations. Together with AVL systems, CAD software can be used to replace a disabled vehicle by dispatching another vehicle, or meet fluctuating travel demand by adjusting transit headways, schedules, and routes. AVL and CAD technologies can vitalize transit by directly improving on-time performance, increasing transit efficiency through providing dispatchers with location, direction and status information, and reducing operating costs through reducing dependence on transit field supervisory personnel. Direct benefits to travelers can include higher reliability of travel times and reduced stress in dealing with transit unreliability. This study explores the development, availability, and impacts of AVL/CAD technologies as reported by AVL vendors and transit implementers. The study defines the key features, functions, and performance characteristics of AVL/CAD technologies that can influence the level of benefits realized. The AVL/CAD implementation context is explored by examining where, when, and for what users these systems are being implemented. The results of two surveys are reported. To explore the availability of AVU CAD systems, technology suppliers were surveyed. Suppliers identified the features, functions, and performance of available AVUCAD technologies. To determine the extent of AVL/CAD deployment, transit operators were surveyed regarding their experiences with AVL/CAD technologies and the subsequent impacts on travelers and transit agency performance. This research provides a systematic method for evaluation of AVL/CAD systems and reports the perceptions of AVL/CAD vendors and transit implementers regarding available products and their impacts. The results suggest a need for better tools to characterize and quantify the impacts and benefits of AVU CAD systems.

Evaluation of Automatic Vehicle Location Technologies for Paratransit in Small and Medium-Sized Urban Areas

This paper presents an evaluation framework designed for small and medium-sized city automatic vehicle location (AVL)-based paratransit systems and describes the application of that framework to one such paratransit system in North Carolina. Issues addressed include the benefits of implementing AVL, whether AVL for paratransit can be used effectively in small to medium-sized cities, and the problems associated with implementing such technologies. The paper is meant to provide guidance for other similar AVL evaluation efforts. A controlled before and after study approach is used to evaluate the AVL system in the City of Winston-Salem, North Carolina (population 265,878). Key findings are that the paratransit system’s efficiency improved slightly with AVL but that demand variables did not change significantly relative to non-AVL equipped vehicles. Reasons for this outcome are examined.

Transforming Bus Service Planning Using Integrated Electronic Data Sources at NYC Transit

The installation of an Automatic Vehicle Location (AVL) system alongside existing Automated Fare Collection (AFC) data spurred development of an inferred bus boarding and alighting ridership model at New York City Transit (NYCT), allowing for 100% passenger origin-destination (O-D) data citywide. Analysis techniques that relied primarily on professional judgment due to lack of data were replaced by more sophisticated statistical techniques. This paper describes two case studies and the resulting service planning potential from having access to fully-integrated big data sources: a neighborhood-wide analysis of performance and ridership, where 100% data allowed planners to pinpoint specific, low-cost reroutes and stop changes to better serve riders, and identification of an optimal route split location for a long route with poor performance by minimizing passenger impact using modeled O-D data. In both examples, new data sources allowed for novel analysis throughout problem investigation as well as forecasting ridership and cost impacts of proposed service adjustments. As the agency’s ability to leverage these data improves, it will support Title VI obligations as well as performance monitoring.

A general dynamic sequential learning framework for vehicle trajectory reconstruction using automatic vehicle location or identification data

Vehicle trajectory data derived from automatic vehicle location (AVL) and automatic vehicle identification (AVI) systems provide critical support for intelligent transportation systems. However, the field-obtained vehicle trajectories are usually incomplete due to sensor malfunction or communication issues. To recover the incomplete data, the existing reconstruction methods have to impose strong assumptions on driver route choice behaviors (network level) and/or traffic dynamics (link level). With the tremendous data available, leveraging data-driven approaches to address the vehicle trajectory reconstruction problem with minimal assumptions is promising. This paper proposes a general dynamic sequential learning framework to reconstruct vehicle trajectory points for both AVL and AVI data. First, an Isolation Forest based ensemble learning model is developed to extract trajectory sequences attributed to different trips in an entire trip chain of a vehicle. Second, the dynamic recurrent neural network (dynamic RNN) is tailored to learn the underlying patterns from the complete AVL and AVI trajectories, respectively. Third, a sequential prediction scheme is customized to reconstruct AVL and AVI trajectories based on the trained networks. To validate the proposed method, two experiments are conducted. One is a simulation experiment with AVL data gathered from a well-calibrated simulation model. The other is a field experiment with AVI data collected from a real-world automatic license plate recognition (ALPR) system. The results show that the proposed method achieves superior performance for both AVL and AVI data compared with the traditional methods. The impacts of different sampling rates and traffic conditions on the model performance are also discussed.

Extracting Bus Transit Boarding and Alighting Information Using Smart Card Transaction Data

The smart card–based automated fare collection (AFC) technologies provide new opportunities for transportation data collection since smart card data (SCD) contain a significant amount of archived information that can be gathered and leveraged to help estimate public transit origin-destination (OD) matrices. Both boarding and alighting location detection are important particularly when there is no automatic vehicle location (AVL) system or other information in the database. With the analysis of raw data using limited information in this paper, an algorithm for boarding stop identification is presented after the trip direction and boarding cluster identification. Alighting stops are then identified based on the relationship between records. Finally, the feasibility and practicality of the methodology are tested using the bus transit SCD collected in Guangzhou, China.

Statistical estimation in passenger‐to‐train assignment models based on automated data

AbstractWith the rapid development of metro systems, it has become increasingly important to study phenomena such as passenger flow distribution and passenger boarding behavior. It is difficult for existing methods to accurately describe actual situations and to extend to the whole metro system due to the limitations from parameter uncertainties in their mathematical models. In this article, we propose a passenger‐to‐train assignment model to evaluate the probabilities of individual passengers boarding each feasible train for both no‐transfer and one‐transfer situations. This model can be used to understand passenger flows and crowdedness. The input parameters of the model include the probabilities that the passengers take each train and the probability distribution of egress time, which is the time to walk to the tap‐out fare gate after alighting from the train. We present the likelihood method to estimate these parameters based on data from the automatic fare collection and automatic vehicle location systems. This method can construct several nonparametric density estimates without assuming the parametric form of the distribution of egress time. The EM algorithm is used to compute the maximum likelihood estimates. Simulation results indicate that the proposed estimates perform well. By applying our method to real data in Beijing metro system, we can identify different passenger flow patterns between peak and off‐peak hours.

Comparison of Door-to-Door Transit Travel Time Estimation Using Schedules, Real-Time Vehicle Arrivals, and Smartcard Inference Methods

Estimating passengers’ door-to-door travel time, for journeys that combine walking and public transit, can be complex for large networks with many available path alternatives. Additional complications arise in tap-on only transit systems, where passengers alightings are not recorded. For one such system, the Chicago Transit Authority, this study compares three methods for estimating door-to-door travel time: assuming optimal path choice given scheduled service, as represented in the General Transit Feed Specification (GTFS); assuming optimal path choice given actually operated bus service, as recorded by automatic vehicle location systems; and using inferred path choices based on automated fare collection smartcard records, as processed with an origin-destination-interchange (ODX) inference algorithm. As expected, ODX-derived travel times are found to be longer than those derived from GTFS, indicating that purely schedule-based travel times underestimate the travel times that are actually available and experienced, which can be attributed primarily to suboptimal passenger route choice. These discrepancies additionally manifest in significant spatial variations, raising concerns about potential biases in travel time estimates that do not account for reliability. The findings bring about a more comprehensive understanding of the interactions between transit reliability and passenger behavior in transportation research. Furthermore, these discrepancies suggest areas of future research into the implications of systematic and behavioral assumptions implied by using conventional schedule-based travel time estimates.

The use of the genetic algorithms for optimizing public transport schedules in congested urban areas

Transit operators face with travel time variability issues related to the design of new transit lines or related to the improvement of the performances of existing lines, especially when transit vehicles without exclusive line. Nowadays, the availability of large data through automated monitoring allows more in-depth this phenomenon of variability in timetables to be pointed out with recorded data. The lack of schedule stability confuses the traveling public and thus the degree of dissatisfaction increases. This leads to a decrease of the public transportation users, especially because of those users having travel alternatives to public transportation. The reliability of the travel time and the vehicle operation according to a stable timetable are the most relevant attributes for the users fidelisation. At the same time, it is well known that the demand for public transportation in congested networks has a large variability over space and time. In this paper, we analyse the deviations from the planned (declared and publicly) schedule for a tram line in Bucharest city in different stations, in diferrent times of the day and in different days of the week, in order to build an optimisation model for needed adjustments to the planned schedule. The data are collected with the Automatic Vehicle Location (AVL) system installed on trams’ board. The number of the adjustments solutions is very large and the genetic algorithm is engaged for the optimisation model solving.

Passenger itinerary inference model for congested urban rail networks

The paper develops a data-driven approach for the inference of passenger itineraries in urban heavy rail systems, the Passenger Itinerary Inference Model (PIIM). In light of increasing demand and crowding, the model can be used to assess the impact of near capacity operations on customers, evaluate system performance, and understand passenger behavior when choosing alternative routes. PIIM uses data from Automatic Fare Collection (AFC) and Automatic Vehicle Location (AVL) systems and is applicable to general urban rail networks with entry and exit fare transactions. PIIM consists of three main modules: the left behind model, the route fractions inference model, and the inference model. The left behind model estimates the probability of any passenger being left behind by station and time interval. The route fractions inference model estimates the route choice fractions given the left behind probabilities. The inference model maps each pair of fare transaction records (entry and exit) to a set of feasible itineraries and infers the train(s) boarded by passengers based on the left behind probabilities and the route choice fractions. Synthetic data using AFC and AVL data from a major, congested urban rail system were generated to validate the model. The results show that the model accurately predicts the itineraries used by passengers, the left behind probabilities, route choice fractions, train loads, and other performance metrics of interest. PIIM is also used with actual data to illustrate its applicability.

Vehicle Routing and Scheduling of Flex-Route Transit under a Dynamic Operating Environment

To improve the reliability, responsiveness, and productivity of the flex-route transit service, this paper investigates the vehicle scheduling and routing problem under a dynamic operating environment. First, we discuss the new operating polices after the introduction of intelligent transportation systems (ITSs), including automatic vehicle location (AVL) system, mobile data terminal (MDT), and computer-aided dispatch (CAD) system. Second, a mixed integer programming (MIP) formulation is employed to solve the offline routing problem. Third, an online scheduling scheme is presented to tackle different dynamic events, such as dynamic requests, travel time fluctuations, cancellations of requests, and customer no-shows. Finally, simulation experiments based on a real-life flex-route transit service are conducted to assess the influence of different dynamic events. The results demonstrate that the proposed scheduling scheme is reliable for coping with various dynamic events, and our findings can be used to guide the policy making of flex-route transit services.

Impact of stops for bus delays on routes

In this paper bus travel time and delay is significant measures in public transportation. This research aims to explore impact of stops for the selected bus route and to find some simple suitable solutions to reduce the travel time and minimize delays for scheduling preferences of travellers. Delays data were collected using automatic vehicle location system (GPS technology) for the bus route #51 in Tashkent. Based on the collected GPS data were proposed the developed a linear and a logistic multivariate regression model. Travel time reliability were estimated regarding to a multivariate regression models and performed. The result shows that very good statistical suitable as expected and thus can be used in the public transportation to determine any travel time using real-time and available offline data. These outcomes, in the future it is expected that will be applicable and more efficient bus public transportation system in Tashkent.

Analysis of global positioning system based bus travel time data and its use for advanced public transportation system applications

The rapid advancements in sensor technologies has resulted in the increased use of Automatic Vehicle Location (AVL) systems for traffic data collection. Global Position System (GPS) sensors are the most commonly used AVL system, majorly because of it being a time-tested technology and being relatively cheap. Also, many of the transportation agencies have their vehicles equipped with GPS sensors. One of the interesting challenges in the field of Intelligent Transportation Systems (ITS) is to effectively mine useful information from such large-scale database accumulated over time. The current study analyses travel time data obtained from buses fitted with GPS devices in Chennai, India to understand its variation over time and space to find the spatial and temporal points of criticality. For this, Cumulative Frequency Distribution (CFD) curves, bar charts and boxplots were used. Inter-Quartile Range (IQR) was used as a measure to quantify the variations in travel time. Analysis showed that both travel time and its variation increased approximately 10% and 40%, respectively, from 2014 to 2016. This increase was observed to be primarily concentrated in six critical intersections during morning and evening peak hours. The findings from the study were further used in demonstrating possible user applications that can improve the efficiency of public transportation systems. As part of this, a real-time bus travel time prediction method was developed using a deep learning approach, Long and Short-Term Memory (LSTM) networks. Along with this, a robust fleet management system was also developed to check the adequacy of buses along the study corridor for different time of the day.

Assessing operation and customer perception characteristics of high frequency local and limited-stop bus service in Vancouver, Canada

Public transport agencies implement different strategies aimed at improving the operation of transit service and to increase satisfaction among its riders. One service strategy employed by transit agencies is a limited-stop bus service that runs in parallel to a heavily used route to decrease travel times for existing riders and to reduce pressure on the local route. Using bus operations data obtained from automatic vehicle location (AVL) and automatic passenger counter (APC) systems and customer satisfaction data collected in Vancouver, Canada, the present study evaluates levels of satisfaction among users of a local and limited-stop bus service while controlling for the service characteristics these users have experienced in the past seven days. Our results reveal that after controlling for characteristics related to the conditions of the service experienced by users, namely passenger activity levels, patrons of the express route service were more likely to be satisfied with the transit service compared to users of the local service. This finding indicates that the operational characteristics of a limited-stop service, including in-vehicle time savings and higher route frequency, are highly valued by its users. Results of this study demonstrate how operations data can provide a greater context for customer satisfaction analyses. Finally, this study provides transit planners and policy makers with a better understanding of how customers perceive local and limited-stop service.

Empirical Performance Analysis of Bus Speed and Delay at Intersections for Emerging Spot Improvement Programs

Many North American cities are increasingly interested in implementing small-scale localized spot treatments to surface routes as a simpler approach than top-down, disruptive route change, or redesign. This research seeks to support the identification of effective spot treatments at intersections using a systematic, data-driven approach. By analyzing key bus performance indicators in Toronto, this study developed insights into factors affecting peak-period bus speeds and delays at the segment and intersection levels for a wide variety of route and intersection configurations across eight high-frequency routes. Candidate treatments were then identified to improve bus performance. Data were sourced from the automatic vehicle location system, general transit feed specification, and a specialized ride check and GPS survey. Features of the approaches of 100 signalized intersections along the study routes were analyzed using K-means clustering, ordinary least squares regression, and regression trees, with target variables as their morning and evening peak operating speeds, segment-level delays, and signal delays. The results showed that long signal split is a significant contributor to higher operating speeds and lower delays, suggesting signal timing adjustments are an effective treatment. Clustering analysis suggested turning restrictions, particularly for right turns at intersections with near-side stops, could be effective, since turning volumes of similarly configured intersections were lower at locations with better transit performance. Regression analyses showed that queue jump lanes are an effective treatment if signal timing plans cannot be adjusted. The results from this study are intended to assist in informing transit authorities wishing to implement future spot improvement programs.

Understanding the factors that influence the probability and time to streetcar bunching incidents

Bunching is a well-known operational problem for transit agencies and it has negative impacts on service quality and users’ perception. While there has been a substantial amount of literature about understanding the factors associated with bus bunching and strategies used to mitigate the effects of this problem, there has been little research on streetcar bunching. Although bus and streetcar systems share many similarities, one major difference between the two is that streetcars cannot overtake each other. This makes bunching in streetcar networks more critical to the reliability of the system and an important topic that requires more in-depth understanding. This research aims at understanding the factors that are associated with the likelihood of streetcar bunching and to investigate in greater detail the external and internal factors that relate to the time to the initial bunching incident from terminal. To achieve the first goal, the study uses a binary logistic regression model, while it uses an accelerated failure time model to address the second goal. The study utilizes automatic vehicle location system data acquired from the Toronto Transit Commission, the transit provider for the City of Toronto. The models’ results show that headway deviations at terminals are related to both an increase in the probability of bunching and an acceleration of the time to bunching. The discrepancy in vehicle types between two successive streetcars also has the same relationship as headway deviations at terminals. This study offers a better understanding of the factors that are associated with streetcar service bunching, which is an important component of transit service reliability.

MODELLING TRAVEL TIME DISTRIBUTION AND ITS INFLUENCE OVER STOCHASTIC VEHICLE SCHEDULING

Due to the paucity of well-established modelling approaches or well-accepted travel time distributions, the existing travel time models are often assumed to follow certain popular distributions, such as normal or lognormal, which may lead to results deviating from actual ones. This paper proposes a modelling approach for travel times using distribution fitting methods based on the data collected by Automatic Vehicle Location (AVL) systems. By this proposed approach, a compound travel time model can be built, which consists of the best distribution models for the travel times in each period of a day. Applying to stochastic vehicle scheduling, the influence of different travel time models is further studied. Results show that the compound model can fit more precisely to the actual travel times under various traffic situations, whilst the on-time performance of resulting vehicle schedules can be improved. The research findings have also potential benefit for the other research based on travel time models in public transport including timetabling, service planning and reliability measurement.