Conference Paper

GIS-based Map-matching: Development and Demonstration of a Postprocessing Map-matching Algorithm for Transportation Research

Authors:
To read the full-text of this research, you can request a copy directly from the authors.

Abstract

This paper presents a GIS-based map-matching algorithm that makes use of geometric, buffer, and network functions in a GIS – to illustrate the suitability of a GIS platform in developing a postprocessing mapmatching algorithm for transportation research applications such as route choice analysis. This algorithm was tested using a GPS-assisted time-use survey that involved nearly 2,000 households in Halifax, Nova Scotia, Canada. Actual routes taken by household members who travelled to work by car were extracted using the GPS data and the GIS-based map-matching algorithm. The algorithm produced accurate results in a reasonable amount of time. The algorithm also generated relevant route attributes such as travel time, travel distance, and number of left and right turns that serve as explanatory variables in route choice models. The ease and flexibility of the Python scripting language used in developing the GIS-based mapmatching algorithm make this tool easy to develop and implement. It can be improved to suit data inputs and specific fields of application in transportation research.

No full-text available

Request Full-text Paper PDF

To read the full-text of this research,
you can request a copy directly from the authors.

... However GPS data collected may consists of positional errors due multi-path signals, atmospheric disturbances, orientation and number of Satellites in view. [2] Provides one of the objectives of post processing of GPS trajectory data is to accurately determine the Links traversed considering the topology and complex junctions in the Road Network. Previously various approaches have been utilized for map matching namely geometric, probabilistic, fuzzy logic, particle filtering, kalman filtering, AI theory etc. Regarding the use of GIS functionalities for Map Matching very few works has been done, the works mainly used GIS environment for visualization and storing of digitized data without giving much importance to spatial analysis tools [3][4] [5]. ...
... Previously various approaches have been utilized for map matching namely geometric, probabilistic, fuzzy logic, particle filtering, kalman filtering, AI theory etc. Regarding the use of GIS functionalities for Map Matching very few works has been done, the works mainly used GIS environment for visualization and storing of digitized data without giving much importance to spatial analysis tools [3][4] [5]. In [2] a purely GIS based map matching algorithm for post processing of GPS trajectory data using GIS functions mainly buffer analysis and network analysis. From the perspective of GIS functionalities for map matching geometric approaches like point-to-point [6] and point-tocurve [7] need some consideration. ...
... One of the issues for GPS data logging is the setting of appropriate sampling interval. In previous works researchers have used different sampling intervals, [2] processed data having 3 recordings in 2 seconds, [8] selected datasets with sampling intervals of 2 and 10 seconds, [9] collected data with interval of 300 seconds, [10] used datasets with 2,3,4 and 5 minutes interval. In [10] an algorithm is provided which accurately map matches GPS points having large sampling rates, however the main objective of the present research is not map matching but to determine average link speed. ...
... The real-time map-matching captures the location of a traveler in the road network with a real-time feed of GPS locations. Post-processing map-matching takes GPS data recorded from a travel and matches it to the road network to trace the routes taken by travelers [5]. In this paper, we have used post-processing map matching in order to project the inaccurate trajectories on the road network. ...
... Based on the discovered rule, trajectory classiication model is built to predict the class label of new trajectory. However, Dalumpines in [5] ooered post processing based approach for the map matching algorithm in order to project the inaccurate GPS trajectories on the road network using the geometric, buuer, and network functions in a GIS software. The algorithm also generated relevant route attributes such as travel time, travel distance, and number of left and right turns that serve as explanatory variables in route choice models. ...
... In this section, the pre-processing of the trajectory data is carried out based on the existing map matching and interpolation techniques [2, 5]. The trajectories and given set of POIs are stored in the spatial grid data structures after pre-processing. ...
Article
Full-text available
Identifying the interesting places through GPS trajectory mining has been well studied based on the visitor’s frequency. However, the places popularity estimation based on the trajectory analysis has not been explored yet. The limitation in the majority of the traditional popularity estimation and place user-rating based methods is that all the participants are given the same importance. In reality, it heavily depends on the visitor’s category, for example, international visitors make distinct impact on popularity. The proposed method maintains a registry to keep the information about the visited users, their stay time and the travel distance from their home location. Depending on the travel nature the visitors are labeled as native, regional and tourist for each place in question. It considers the fact that the higher stay in a place is an implicit measure of the greater likings. Theweighted frequency is eventually fuzzified and applied rule based fuzzy inference system (FIS) to compute popularity of the places in terms of the ratings ∈ [0, 5]. We have evaluated the proposed method using a large real road GPS trajectory of 182 users for identifying the ratings for the collected 26807 point of interests (POI) in Beijing (China).
... This is commonly known as the map-matching process, and it has been tackled by researchers following different procedures (Schuessler and Axhausen, 2009). For this study, based on the map-matching algorithm created by Dalumpines & Scott (2011), a new version was developed, improving an aspect that was relevant for the purpose of our research. ...
... Much better map-matching results are obtained when combining geometric and topological procedures, since they "consider the connectivity of the network in assessing the feasibility of a route" (Hudson et al., 2012). Between these hybrid approaches, the algorithm developed by (Dalumpines and Scott, 2011) was especially interesting for our research, because of its easy integration into a GIS environment (using ArcGIS's Network Analyst tools) and because of the results obtained by (Larsen et al., 2013) in the map-matching process applied to the routes collected in Texas, with 88% success. For this study, based on the map-matching algorithm created by Dalumpines & Scott (2011), a new version was developed, improving an aspect relevant to the purpose of our research. ...
... Between these hybrid approaches, the algorithm developed by (Dalumpines and Scott, 2011) was especially interesting for our research, because of its easy integration into a GIS environment (using ArcGIS's Network Analyst tools) and because of the results obtained by (Larsen et al., 2013) in the map-matching process applied to the routes collected in Texas, with 88% success. For this study, based on the map-matching algorithm created by Dalumpines & Scott (2011), a new version was developed, improving an aspect relevant to the purpose of our research. The procedure basically creates a buffer around the GPS track-line that constrains the estimation of the shortest path between the origin and the destination, by using Dijkstra's algorithm. ...
Article
Full-text available
Given the growing interest in promoting more sustainable urban transport, over the past years, many researchers have analyzed cycling mobility from different perspectives. However, some important aspects remain underexplored. One of them is the study of cycling speeds. The first goal of this research is to analyze the impact of a wide range of factor on cyclists’ speeds. Based on the examination of thousands of GPS routes, the investigation conducts diverse OLS regressions in order to analyze cyclists’ speed according to the diverse local factors that affect cyclists along their journey, such as the slope, the existence – and type – of bike infrastructures, the average traffic speed or the density of traffic lights or intersections. Cycling speed is also analyzed according to cyclists’ gender or age, the purpose of the journey, or even the weather conditions. The research includes the analysis of regular cyclists’ trips, as well as the analysis of bike-messengers’ routes. The results obtained shed light on the influence of these factors on cyclists’ speed by quantifying their specific impact, and diverse models predict cyclists’ travel times in the current scenario but also in future ones that may correspond to the implementation of new infrastructure or policies. In addition, the models allowed us to pursue the second goal of this study: to conduct a comparative analysis of accessibility for different transport modes, and then evaluate the competitiveness between them. The results evidence that cycling is not only a sustainable transport mode, but the most competitive for small-medium distances.
... We have proposed to use map matching techniques to preprocess the GPS trajectory before actually estimating the popularity of the locations. Map Matching, is the process of projecting the GPS fixes on the road network graph G = (V, E) [6]. In this paper, we have used map matching in order to project the inaccurate trajectories on the road network. ...
... It is hence difficult to calculate the travel distance only based on the trajectories. In this section, the preprocessing of the trajectory data is carried out based on the existing map matching and interpolation techniques [5, 6]. The trajectory is converted into the road network graph which is aware of the connectivity of the roads and hence it becomes more accurate to compute the reachability among consecutive points in the trajectory. ...
... The input to the map marching is the road network, trajectory database; however the output is the accurate GPS sequence that guarantees the point being exactly on the road if moving or in any geographical location (if there is a stay). For the sake of completion, we have discussed the map matching in short, that is inspired from the work in [6]. ...
Article
Full-text available
The mining of the user GPS trajectories and identifying the interesting places have been well studied based on the visitor's frequency. However, every user is given the same importance in the majority of the trajectory mining methods. In reality, the popularity of the place also depends on the category of the visitor i.e. international vs local visitors etc. We are proposing user category based location popularity estimation using the trajectories databases. It includes mainly three steps. First, pre-processing – the error correction and the graph connection establishment in the road network in order to be able to carry the graph based computations. Second, find the stay regions where the travelers spent some time off-the-road. The visitors can be easily categorized for each POI based on the travel distance from the home location. Finally, normalization and popularity estimation – measure the frequency and stay time of the visitors of each category in the places in question. The weighted sum of the frequency and stay time for each category of the visitors is calculated. The final popularity of the places is computed with values of the pre-configured range. We have implemented and evaluated the proposed method using a large real road GPS trajectory of 182 users that was collected in a period of over three years by Microsoft Asia Research group.
... The algorithms need to be executed usually for a wide range of parameter values, resulting in extensive experimentation and resource consumption. Traditionally, the parameter values are calibrated for data collected in a specific environmental setting such as dense urban or rural areas [4]. Hence, the reported performance and parameters values of MMAs may be conditioned and biased according to the employed test dataset. ...
... Post-processing MMAs determine the correct road on which a user or vehicle is traveling once the complete or part of the GPS dataset is available. In this study, a post-processing MMA is applied to an instance that is comprised of a set of GPS data points that represents a complete vehicle route [4], [9]. For example, post-processing MMAs are needed to obtain correct paths when collecting multi-modal trip data with GPS receivers for route choice modeling in a GIS environment [1], [14], [17]- [19]. ...
... Road matching algorithms, based on the premise that the ego vehicle's position will always be restricted to the road network, aim to integrate previous positioning data with spatial road network data to identify the actual road link on which the vehicle is traveling. Different approaches have been made to solve this issue, and the map-matching algorithms are usually divided into three categories: geometric [16], topological [17] and advanced map matching [14]. ...
... On the other hand, road-matching techniques are usually based on simplified maps and are less costly in computational terms. Even in the case of not having a digital map of the driving environment, most of the time this can be easily built using geographic information systems (GIS) [16]. Thus, if the maps are accurate, the latter approach can provide global positioning of the vehicle, making it an interesting complement to GPS-based positioning [10]. ...
Article
Full-text available
Accurate localization for autonomous vehicle operations is essential in dense urban areas. In order to ensure safety, positioning algorithms should implement fault detection and fallback strategies. While many strategies stop the vehicle once a failure is detected, in this work a new framework is proposed that includes an improved reconfiguration module to evaluate the failure scenario and offer alternative positioning strategies, allowing continued driving in degraded mode until a critical failure is detected. Furthermore, as many failures in sensors can be temporary, such as GPS signal interruption, the proposed approach allows the return to a non-fault state while resetting the alternative algorithms used in the temporary failure scenario. The proposed localization framework is validated in a series of experiments carried out in a simulation environment. Results demonstrate proper localization for the driving task even in the presence of sensor failure, only stopping the vehicle when a fully degraded state is achieved. Moreover, reconfiguration strategies have proven to consistently reset the accumulated drift of the alternative positioning algorithms, improving the overall performance and bounding the mean error.
... The algorithms need to be executed usually for a wide range of parameter values, resulting in extensive experimentation and resource consumption. Traditionally, the parameter values are calibrated for data collected in a specific environmental setting such as dense urban or rural areas [4]. Hence, the reported performance and parameters values of MMAs may be conditioned and biased according to the employed test dataset. ...
... Post-processing MMAs determine the correct road on which a user or vehicle is traveling once the complete or part of the GPS dataset is available. In this study, a post-processing MMA is applied to an instance that is comprised of a set of GPS data points that represents a complete vehicle route [4], [9]. For example, post-processing MMAs are needed to obtain correct paths when collecting multi-modal trip data with GPS receivers for route choice modeling in a GIS environment [1], [14], [17]- [19]. ...
Article
Map Matching Algorithms (MMAs) are developed to solve spatial ambiguities that arise in the process of assigning GPS measurements onto a digital roadway network. Scarce systematic parameter tuning approaches exist in the literature for optimizing MMA performance. Thus, a novel framework is proposed for a systematic calibration of the parameters of a post-processing MMA. The calibration approach consists of an Instance-specific Parameter Tuning Strategy (IPTS) that employs Fuzzy Logic principles. The proposed fuzzy IPTS tool determines algorithm-specific parameter values based on instance-specific information a priori to the execution of the MMA. Finally, the proposed IPTS tool is able to adjust to two particular decision maker preferences on algorithm performance, namely solution quality and computational time.
... Moreover, an adequate understanding and modeling of travel decision variability may improve transportation management and information systems. Travel behavior research may address trip purpose, route and parking choices, travel mode selection, travel modeling, congestion management, transportation system performance, and forecast demand models, which require the identification of accurate track movements of vehicles, individuals and objects along trip segments and routes on a digital roadway map (Dalumpines & Scott, 2011), usually embedded in GIS or AVL systems. Other common applications in this field are studies of behavioral responses to advanced traffic and incident management, advanced traveler information systems, and dynamic traffic assignment. ...
... In this case, trips are correctly referenced to the roadway network and travel mode, and trip time and distance are identified for each trip segment. Hence, travel behavior analysis is enhanced by providing a more accurate description of observed routes (Dalumpines & Scott, 2011). ...
Article
Full-text available
The map matching problem arises when GPS measurements are incorrectly assigned to the roadway network in a GIS environment. This chapter presents a real-time topological decision rule-based methodology that detects and solves spatial mismatches as GPS measurements are collected. A real-time map matching methodology is required in several applications, such as fleet management, transit control and management, and travel behavior studies, in which decision-making must be performed simultaneously with the movement of vehicles, individuals, or objects. A computational implementation in a real case scenario in Chile indicates that the algorithm successfully resolves over 96% of the spatial mismatches encountered in real time. Various algorithmic parameter values were employed to test the performance of the algorithm for data collected every 5 and 10 seconds. Overall, the algorithm requires larger buffer sizes and speed ranges to obtain better results with lower spatial data qualities. 2014 by IGI Global. All rights reserved.
... Although a few studies have developed a post-processing map-matching algorithm in a GIS-environment, they still adapt real-time map-matching procedures (Chung and Shalaby 2005). Dalumpines and Scott (2011) have tried to fill this void by proposing the first post-processing map-matching algorithm fully integrated into a GIS-environment. They set out to convert the stream of GPS-measurement into a polyline feature. ...
... The first alternative method suggested here is the 'Connected Subset' assignment procedure, somewhat similar to the method described to make sense of locational data derived from cell phone records (Krygsman, Nel, and De Jong 2008) and to the real-time algorithm proposed by Miwa et al. (2012). Whereas Dalumpines and Scott (2011) use a buffer with a predefined threshold value to limit the network through which the shortest path algorithm finds a solution, the 'Connected Subset' assignment takes a different approach by using the full network to connect consecutive GPS-Points by a series of 'mini' shortest path assignments, assuming that on such a short time interval the shortest path principle always applies as it is highly unlikely that detours are made between two GPS-measurements. The combination of all these 'mini' shortest path assignments results in a fully connected subnetwork. ...
Article
The exact distance and routes travelled on an individual level are essential variables in determin- ing the effectiveness of ‘soft’ transport demand management strategies. The ability to track individuals in great spatial detail by means of Location Aware Technologies such as GPS has opened avenues for gathering these data with great precision. Route reconstruction with posi- tional data is typically done by a process referred to as map-matching; however, despite the large number of real-time map-matching algorithms developed, few studies have developed a post- processing map-matching algorithm in a Geographic Information Systems (GIS)-environment. This paper presents two GIS-based map-matching methods that predominantly use a digital road network with speed and directionality attributes for route reconstruction of raw GPS-trajectories. The methodologies were tested for a data set in which actual routes travelled were known. Both explored procedures, the ‘Connected Subset’ assignment procedure based on network subset selection and the ‘Impedance Reduction’ assignment procedure based on attri- bute adjustment, provide accurate results. In addition, both procedures effectively deal with commonly GPS-induced problems such as measurement gaps and positional drift.
... Our literature search only found one previous study examining route choice using bikeshare GPS data (based on the Phoenix system) [11]. Khatri employed a map-matching system developed by Dalumpines & Scott [24] using Network Analyst in ArcGIS, and also used by Hudson et al. [13] in a non-bikeshare based study. This method involves creating a buffer around each GPS point transmitted by the bicycles and then calculating the shortest route between origin and destination of GPS points within the buffer. ...
... This is especially true in cities, where "urban canyons" can block satellite signals and create inaccurate tracks. The aforementioned map-matching method by Dalumpines & Scott [24] has been shown to be useful for cleaning direct routes. This data was also not accurate enough to differentiate between sidewalk and roadway riding. ...
Article
Full-text available
Bikeshare systems with docking stations have gained popularity in cities throughout the United States—increasing from six programs with 2,300 bikes in 2010 to 74 systems with 32,200 bikes in 2016. Even though bikeshare systems generate a wealth of data about bicycle check-out and check-in locations and times at docking stations, virtually nothing is known about routes taken and activities undertaken between check-out and check-in of the bikes. This information could greatly enhance future expansion of bikeshare systems, placement of new docking stations, as well as the location for new bike lanes and paths. To study this, the District Department of Transportation (DDOT) placed GPS trackers on 94 Capital Bikeshare (CaBi) bikes in the spring of 2015. Based on these data this GIS analysis distinguishes riders by type of CaBi membership, identifies popular routes, analyzes bicycle infrastructure usage, and examines stopping and dwelling times at places of interest. Results show strong differences in trip attributes between short-term users (24h or 3-day memberships) and monthly or annual CaBi members. Trips taken by short-term users were longer in distance, slower, and concentrated off-road on parkland in and around the National Mall. Trips by CaBi members were shorter, faster, and were concentrated away from the National Mall in popular mixed-use neighborhoods. Short-term members rode 12 percent of miles on dedicated bicycle infrastructure, 61 percent in parks, and 27 percent on roadways with motorized traffic, while the shares for members were 33, 17, and 50 percent respectively. Based on routes taken this study also recommends potential locations for bicycle infrastructure improvements and new bikeshare stations.
... Start and end times of drive to work and shop trips were used to extract travel trajectories (sequences of GPS points representing trip segments) from the STAR GPS data (Fig. 1). Work and shop travel trajectories were extracted then input into a map-matching algorithm (Dalumpines and Scott, 2011) to extract observed routes. In turn, the observed routes were used to generate the route choice data using a route choice set generation algorithm (discussed in Section 3.3), and then route attributes were generated. ...
... 1. In generating route choice sets, PPAG creates an initial list of traversable links by defining a PPA around the observed route-the path generated by the map-matching algorithm (Dalumpines and Scott, 2011) for the GPS trip segment as superimposed on digital road network. The PPA is implemented using service area analysis in ArcGIS® based on maximum travel distance or time depending on the unit of travel impedance used (meters or minutes). ...
Article
This paper presents an in-depth comparison of route choice models for work and shop vehicle trips—with emphasis on the interactions between route attributes and individual characteristics—to better understand the route choice determinants that are assumed to vary by trip purpose. Insights into the route choice behavior involving two dominant vehicle trip purposes—work and shop trips—will help in the design of traffic facilities and implementation of measures to influence route choice in the desired direction. In this study, we show that the utility and scale parameters for separate models of work and shop trips differ by direct comparison using a sequential scaling estimation method and likelihood ratio tests, and highlight the differences in route choice behavior by considering the interaction of route attributes and individual characteristics using Path-Size Logit modeling. In the process, we used Potential Path Area - Gateway (PPAG) algorithm—that generates feasible route choice sets for route choice modeling from GPS trajectories of observed routes. The results show that, indeed, route choice behavior varies by trip, which suggests that drivers attach value to route choice determinants relative to trip purpose. The inclusion of interaction terms in model specifications further indicates that work route choice behavior tends to be restrictive compared to the nonrestrictive route choice for shop trips—a generalization consistent with the mandatory and discretionary nature of work and shop trips, respectively. Specifically, individual characteristics such as personal income, age, gender, tenure, household size, and access to public transit affect route choice behavior.
... Un des inconvénients de cette méthode est l'absence de l'attribut vitesse dans la (Dalumpines, 2011). ...
... Cette méthode consiste à tracer la trace GPS, à créer une zone tampon autour et à chercher le chemin le plus court dans cette zone tampon entre les deux extrémités de la trace GPS. (Dalumpines, 2011) Il suffit ensuite d'accrocher les points de la trace GPS à l'itinéraire généré précédemment. Pour ce projet, le choix a été de proposer un modèle permettant l'automatisation du processus. ...
Experiment Findings
Full-text available
Ce projet a permis d'imaginer et d'appliquer une méthode visant à qualifier la structure et le fonctionnement de réseaux de transport en partant de la technologie GPS. Cette méthode prévoit la collecte des traces GPS par téléphone, le traitement SIG des données et la représentation des données obtenues sous formes de cartes et de grilles horaires.
... The GPS sometimes fails to locate mobile phone devices when they are heavily shielded, and the positioning is not always accurate when the signal is weak or when there is signal interference (Kim et al., 2010;Kim et al., 2012). Some of these errors are rectifiable and can be corrected by map matching (Dalumpines and Scott, 2011;Zheng, 2015). However, some errors are un-rectifiable and need to be filtered out using some thresholds, such as distance and speed thresholds between two consecutive positioning points (Oksanen et al., 2015). ...
Article
Full-text available
Travel behaviour has been studied for decades to guide transportation development and management, with the support of traditional data collected by travel surveys. Recently, with the development of information and communication technologies (ICT), we have entered an era of big data, and many sources of novel data, including mobile phone data, have emerged and been applied to travel behaviour research. Compared with traditional travel data, mobile phone data have many unique features and advantages, which attract scholars in various fields to apply them to travel behaviour research, and a certain amount of progress has been made to date. However, this is only the beginning, and mobile phone data still have great potential that needs to be exploited to further advance human mobility studies. This paper provides a review of existing travel behaviour studies that have applied mobile phone data, and presents the progress that has been achieved to date, and then discusses the potential of mobile phone data in advancing travel behaviour research and raises some challenges that need to be dealt with in this process.
... The GPS tracks were map-matched to the street network using a geometrical, postprocessing approach as described by Dalumpines and Scott (2011). The approach involves geometric and topological GIS functionalities. ...
Chapter
Full-text available
A high share of bicycle traffic in urban areas can be advantageous in order to tackle traffic related problems such as congestion, over-crowded public transportation or air pollution. Through an increased dissemination of e-bikes in recent years, cycling has become a viable transportation alternative for an even broader audience. The consequences of this trend on urban mobility are not yet clear. In order to get a clearer picture, one first needs to understand the major usage differences between e-bikers and cyclists. In this paper we demonstrate how a first insight into these differences can be gained by analysing GPS tracking data, recorded within the context of a field study. E-bikers as well as conventional cyclists prefer riding on any kind of bike trail whilst e-bikers rather choose bike trail types with a larger exposure to vehicular traffic. Taking a minimal distance route was the most important route choice factor for both cyclists and e-bikers. E-bikers perceived their rides to be slightly more safe and convenient as compared to conventional cyclists.
... If the discrepancy between historical average speed and the vehicle's average speed is high, one might consider other possible paths rather than the shortest path between the two GPS samples. To the best of our knowledge, other algorithms that combine spatial analysis only consider the shortest path (Chen and Bierlaire 2015;Dalumpines and Scott 2011;Lou et al. 2009;Oshyaniv et al. 2014; for travel path inference problems. ...
Article
Full-text available
In this study, we propose a novel method for a travel path inference problem from sparse GPS trajectory data. This problem involves localization of GPS samples on a road network and reconstruction of the path that a driver might have been following from a low rate of sampled GPS observations. Particularly, we model travel path inference as an optimization problem in both the spatial and temporal domains and propose a novel hybrid hidden Markov model (HMM) that uses a uniform cost search (UCS)-like novel combinational algorithm. We provide the following improvements over the previous studies that use HMM-based methods: (1) for travel path inference between matched GPS positions, the proposed hybrid HMM algorithm evaluates all candidate paths to find the most likely path for both the temporal and spatial domains. In contrast, previous studies either create interpolated trajectories or connect matched GPS positions using the shortest path assumption, which might not be true, especially in urban road networks (Goh et al. 2012; Lou et al. 2009). (2) The proposed algorithm uses legal speed limits for the evaluation of discrepancy in the temporal domain as in Goh et al. (2012), and Lou et al. (2009) only if there is not sufficient historical average speed data; otherwise, we use historical average speed computed from data. Our experiments with real datasets show that our algorithm performs better than the state of the art VTrack algorithm (Thiagarajan et al. 2009), especially for cases where GPS data is sampled infrequently.
... This final stage consists of three main modules: Choice Set Generator Module (CSGM), RCA Variables Generator (RVGM), and Activity Locations Identification Module (ALIM). CSGM and RVGM generate data for route choice modeling; both modules use the map-matching algorithm (Dalumpines and Scott, 2011) that snaps walk and non-walk trip segments to a digital road/pedestrian network to derive the actual travel (observed) routes. CSGM uses a modified potential path area -gateway algorithm (PPAG) to generate alternative routes for trip segments produced by TGEM or TSEM (Dalumpines and Scott, 2017a). ...
Article
The widespread use of global positioning systems (GPS) has prompted transportation researchers to develop tools and methods to extract information from person-based GPS data. However, most of these procedures suffer from specific data requirements and complexity that limit their transferability. Further, they have a limited set of modules to extract all necessary information (e.g., route attributes), and were not specifically designed to handle huge GPS datasets. To deal effectively with these problems, this paper presents a framework based on three design principles (transferability, modularity, and scalability), along with the geographic information system (GIS)-based episode reconstruction toolkit (GERT) based on this framework, for automated extraction of activity episodes from GPS data.
... The procedures presented were developed and implemented in Python®, a free scripting language (www.python.org). The scripting language facilitates fast development and allows easy integration with the map-matching algorithm (Dalumpines and Scott 2011), which was also written in Python®. These procedures and algorithms are integral part of GERT, which was developed to automatically extract activity episodes from GPS data and derive associated information for each extracted episode. ...
Article
The increasing popularity of global positioning systems (GPSs) has prompted transportation researchers to develop methods that can automatically extract and classify episodes from GPS data. This paper presents a transferable and efficient method of extracting and classifying activity episodes from GPS data, without additional information. The proposed method, developed using Python®, introduces the use of the multinomial logit (MNL) model in classifying extracted episodes into different types: stop, car, walk, bus, and other (travel) episodes. The proposed method is demonstrated using a GPS dataset from the Space-Time Activity Research project in Halifax, Canada. The GPS data consisted of 5127 person-days (about 47 million points). With input requirements directly derived from GPS data and the efficiency provided by the MNL model, the proposed method looks promising as a transferable and efficient method of extracting activity and travel episodes from GPS data.
... Second, the other part in the map-matching framework is the complex topological relationships among these shapes in traffic context, particularly when real trajectories and calculated trajectories are related to the context. For mobile phone positioning, which exploits the data from a cellular network system, the positioning results are highly sensitive to the local context [5][6][7]. The preliminary positioning result from mobile positioning are more complex than that the preliminary positioning results from other positioning methods [4,8]. ...
Article
Full-text available
Numerous map-matching techniques have been developed to improve positioning, using Global Positioning System (GPS) data and other sensors. However, most existing map-matching algorithms process GPS data with high sampling rates, to achieve a higher correct rate and strong universality. This paper introduces a novel map-matching algorithm based on a hidden Markov model (HMM) for GPS positioning and mobile phone positioning with a low sampling rate. The HMM is a statistical model well known for providing solutions to temporal recognition applications such as text and speech recognition. In this work, the hidden Markov chain model was built to establish a map-matching process, using the geometric data, the topologies matrix of road links in road network and refined quad-tree data structure. HMM-based map-matching exploits the Viterbi algorithm to find the optimized road link sequence. The sequence consists of hidden states in the HMM model. The HMM-based map-matching algorithm is validated on a vehicle trajectory using GPS and mobile phone data. The results show a significant improvement in mobile phone positioning and high and low sampling of GPS data.
... Topological approaches, other then the sole distance, also take into consideration the sequence of GPS points and the connectivity of network elements. Different procedures have been tested [10] [11] [12] [13]. Most of them consist of a combination of geometrical approaches: first, the initial edgeor nodeis found based on some geometrical criteria; then a set of likely edges is defined, for example using a buffer around the GPS trace; finally, the route is developed by choosing the most likely edge out of the set, again using different geometrical criteria. ...
Article
A novel map-matching algorithm is proposed, implemented and applied to global positioning system (GPS) traces which have been recorded by cyclists in Bologna, a medium-sized city in the North of Italy. The algorithm has been developed to match geo-referenced traces to a sequence of edges of a given road network model. Map-matching for bike trips is particularly challenging as cyclists often use footpath or parks which are not necessarily represented by the road network model. The matching algorithm should smartly tolerate the lack of network information. The algorithm should also be fast and capable of processing thousands of GPS traces in a reasonable time. The proposed probability-based method, which also exploits information on various network attributes, allows a reliable and fast map matching, even in dense street networks and with interrupted GPS data streams. In fact, one serious issue is to find a reliability measure which allows to verify the matched routes, without the knowledge of the real routes, as the available cyclist traces are anonymous. In addition to the reliability check, a sensitivity analyses with respect to the most relevant parameters has been conducted. The advantages of the proposed map-matching algorithm are quantified through a direct comparison with a topology-based map-matching algorithm from literature. In this work a novel, buffer-based map-matching algorithm has been devised and applied to GPS traces collected by cyclists, and the road network of Bologna, Italy. Approximately 3250 (55%) traces could be reliably matched to the road network. The creation of buffers around network edges and the determination of the probability of GPS points inside the buffer, in combination with length-and type-specific edge attributes, has shown promising results, in particular in dense street grids and for different road types. The reliability of the matched routes has been verified by comparing the length of the matched route to the length of the GPS trace. The ratio of both length measures has been termed length index, which should be ∼1 for a good match. Sensitivity analyses for all important algorithm parameters have been conducted. The resulting distance errors are in the range of the GPS recording precision. The algorithm generally satisfies the set objectives: the computing time is 1.6 ms per metre, a 3 km trace can be matched in <5 s; the reliability of the routes has been verified even in the presence of complex street grids and junctions; the ‘closing’ of gaps caused by GPS signals blackouts is part of the algorithm itself and does not need to be performed externally; the proposed algorithm does also use edges with reserved bikeways, when available. The performance of the proposed buffer-based matching algorithm has been compared with a topology-based algorithm, using the same traces, network and computer. The topology-based algorithm has seven times faster execution times, but showed a length index of 1.9 due to unrealistic detours at junctions. Furthermore the investigated topology-based algorithm requires continuous GPS points, while gaps in the trace must be closed by an external algorithm. Matching errors can also occur due to imperfections in the connectivity of the network. The level of details in modelling cycling networks should follow the increasing level of details that GPS data provides to analysts. There is still ample room for improvement of map-matching algorithms. In particular, an alternative definition of the edge weights could further improve matching reliability. PLEASE CONTACT ME FOR THE FULL ARTICLE
... Following the data filtering and cleaning phases, the route traces could be mapmatched. There exist many approaches to perform map-matching, however, few are user-friendly for the non-programmer (Dalumpines & Scott, 2011;Schuessler & Axhausen, 2009;. The chosen approach was based on GIS, using ArcMAP 10.6 to conduct a shortest path search for each origin-destination (OD) pair on the transport network contained within a 50m buffer around the user-drawn route. ...
Thesis
Full-text available
University download page: http://hdl.handle.net/11250/2612495 The prioritisation of bicycle-friendly infrastructure is now on the agenda of many policymakers seeking to capitalise on the advantages of cycling for transport. This thesis focusses upon how the improved availability, quality, and connectivity of infrastructure suitable for cycling can influence cycling behaviour at the city and neighbourhood level. Two key elements are necessary to understand the local-scale impact of bicycle infrastructure: the decision to bicycle in preference of other transport modes and the choice of route on the transport network. This thesis first addresses bicycle mode and route choice independently of each other before analysing the interaction between these elements in the context of bicycle infrastructure interventions. This article-based thesis is comprised of five research papers: four empirical studies and one literature review. Three of the empirical cases are based in the Norwegian city of Trondheim and the fourth is based in Oslo. Paper I addresses the modal shift of employees following a workplace relocation. Papers II and III are focused on bicycle route choice – firstly as a review of methods and then in connection with student route preferences. The two final papers, Papers IV and V, integrate both mode and route choice elements for the detailed analysis of neighbourhood scale effects resulting from the installation of bicycle lanes in Trondheim and Oslo respectively. The research uses a mixed methods approach, with a focus on empirical data to address the objectives of the thesis. Before and after travel surveys, web-based maps and GPS are the main means of data collection. Comparative analyses are performed using a Geographic Information System (GIS). Findings suggest that the decision to bicycle is to a significant extent determined by trip and spatial characteristics of the destination (Paper I). Route substitution is witnessed in both intervention studies (Papers IV and V), whilst significant changes (p < .05) in the modal share of cyclists is only witnessed in one (Paper IV), suggesting that it is mostly changes of route rather than mode that contribute to an individual intervention street’s change in bicycle volumes. Bicycle-specific infrastructure appears to be generally valued by all types of road users, however, the evidence suggests that public transport users and pedestrians are more willing to change their mode of transport assuming the only changes made are to the bicycle infrastructure (Papers I and IV). This suggests that much of the increase in the use of new bicycle infrastructure is the result of a reduction in the use of other sustainable transport modes. Many of the benefits associated with increased cycling are the result of reduced private car use, but for this to be achieved, it appears that initiatives beneficial for cyclists alone are insufficient. The means by which different transport infrastructure attributes can be researched and are valued by users are discussed by Papers II and III respectively. Paper II is a systematic review summarising the means through which revealed preference bicycle route choice data can be collected whilst Paper III evaluates four different Bicycle Level of Service (BLOS) methods for determining bicycle route choice. The latter study reveals that empirically founded BLOS methods with the most explanatory infrastructural attributes correspond best with actual route choices of university students. Of the tested BLOS methods, the Bicycle Compatibility Index is found to correspond best with actual route choice. Developing an understanding of the impacts of bicycle infrastructure can assist the prioritisation of limited city budgets towards the promotion of sustainable mobility behaviour. This research attempts to advance the state of the art for bicycle route choice research whilst also addressing the decision to bicycle for transportation purposes.
... Uncertainties arise e.g. because of GNSS errors, in cities especially in urban canyons (Thiagarajan et al., 2009). Map matching can remove some of these errors (Dalumpines & Scott, 2011), but at the same time, mismatches introduce new uncertainties. On the other hand, given the volume of the dataset the impact of small errors in single trajectories is expected to be negligible, in contrast to the bias that the data carries. ...
Article
Full-text available
Mobile activity tracking data, i.e. data collected by mobile applications that enable activity tracking based on the use of the Global Navigation Satellite Systems (GNSS), contains information on cycling in urban areas at an unprecedented spatial and temporal extent and resolution. It can be a valuable source of information about the quality of bicycling in the city. Required is a notion of quality that is derivable from plain GNSS trajectories. In this article, we quantify urban cycling quality by estimating the fluency of cycling traffic using a large set of GNSS trajectories recorded with a mobile tracking application. Earlier studies have shown that cyclists prefer to travel continuously and without halting, i.e. fluently. Our method extracts trajectory properties that describe the stopping behaviour and dynamics of cyclists. It aggregates these properties to segments of a street network and combines them in a descriptive index. The suitability of the data to describe the cyclists' behaviour with street-level detail is evaluated by comparison with various data from independent sources. Our approach to characterizing cycling traffic fluency offers a novel view on the cyclability of a city that could be valuable for urban planners, application providers, and cyclists alike. We find clear indications for the data's ability to estimate characteristics of city cycling quality correctly, despite behaviour patterns of cyclists not caused by external circumstances and the data's inherent bias. The proposed quality measure is adaptable for different applications, e.g. as an infrastructure quality measure or as a routing criterion.
... Wolf et al., 2004;Stopher et al., 2005;Schssler, 2010;Moiseeva et al., 2010;Marchal et al., 2011), due to the sheer amount of information available but also in relation to mapping the data onto a road network (map matching) (e.g. Pyo et al., 2001;Marchal et al., 2005;Schssler, 2010;Dalumpines and Scott, 2011) and addressing data errors. This section describes some of these steps for the present study, looking at how the data was prepared and cleaned with a view to making it suitable for choice set generation and the estimation of route choice models. ...
Article
This paper presents a novel application in route choice modelling using Global Positioning System (GPS) data, focussing on heavy goods vehicles which typically make longer journeys with decisions potentially underpinned by different priorities from those used by car drivers. The scope of the study is larger than many previous ones, using the entire road network of England. Making use of the error components model put forward for route choice by Frejinger and Bierlaire (2007), the work reveals low elasticities in response to changes in travel time, reflecting the limited opportunity for avoiding specific roads on long distance journeys by heavy goods vehicles.
... Bicycle trips were mapped onto specific segments of the transportation network using ArcGIS. A map matching algorithm based upon the method by Dalumpines and Scott (2011) was used: bicycle trips were matched to the roadway without knowledge of where bicyclists were riding on the roadway. In total, 3615 trips were recorded by 317 users. ...
Article
This paper develops an approach that uses GIS and an unlabeled multinomial logit (MNL) model to estimate the impact bridge facility attributes have on bicycle travel behavior. The data used to estimate the model were collected from May to October 2011 in Austin, TX via a GPS-based smartphone application that allowed trips to be tracked in real time. Demographic (age, gender, and cycle frequency) and trip purpose information was also collected. Three attributes are analyzed in the model: bridge accessibility to the bicycle network, vehicular volume, and bicycle separation from traffic. Accessibility and bicycle separation significantly impacted bicyclists' behavior, especially for female and infrequent bicyclists as well as for trips where travel time is not a significant issue. Distance was also analyzed and found to be the most significant factor, particularly for timeconstrained trips (trips during the peak period and commute trips). Distance was less important for recreational trips as well as for female bicyclists.
... In this type of analysis, there is some uncertainty about a vehicle's spatial position at various times, and the analyst aims to determine the most likely routes or stops made by the driver. Methods of trip reconstruction are commonly applied to travel survey data, where the analyst knows the set of origins and destinations but not the routes travelled between locations (Dalumpines and Scott, 2011). In practice, these pathways are sometimes estimated as shortest or simplest paths or using other approaches (Duckham and Kulik, 2003). ...
Article
Vehicle tracking data are often used to explore human travel behavior and activity patterns. Time geography is a useful approach for analyzing such datasets, as it provides a means for identifying the set of possible routes and stops for a vehicle between known locations, which is termed a potential path tree. This research extends the utility of the time-geographic approach by developing a method to generate probabilistic potential path trees that represent the space–time potential of a vehicle’s movements. First, this research provides the mathematical formulation of the new technique, network-based time-geographic density estimation (TGDE), and demonstrates the computation using a hypothetical tracking dataset and road network. Its formulation operates as a network adaptation of classical TGDE, which has been previously employed to analyze the movements of objects travelling in continuous, Euclidean space. Second, network-based TGDE is applied in the context of analyzing vehicle tracking data collected by GPS and filtered to protect an individual’s privacy. The method was used to map and quantify the vehicle’s most likely routes, origins, intermediate stops, and final destinations. The results indicate network-based time-geographic density estimation provides a powerful approach for both geovisualizing and analyzing vehicle tracking data.
Article
To understand a bicyclist’s route choice is difficult, given the many factors that influence the attractiveness of different routes. The advent of low-cost GPS devices has made route choice analysis more precise. Bikeshare, with instrumented bikes, allows for better assessment of revealed route preference of a large subpopulation of cyclists. This study used GPS data obtained from 9,101 trips made by 1,866 users of Grid Bikeshare, Phoenix, Arizona. This unique bikeshare system relied on Social Bicycles’ onboard telematics, which allowed nonstation origins and destinations, and operated on a grid street network. The system enabled unique route choice analysis. The trips studied included only direct utilitarian trips. Circuitous trips that could have included multiple destinations or could have been recreational trips were removed. The analysis focused on facility use assessment and route choice behavior. The results were compared between two categories of bikeshare users: registered users and casual users....
Conference Paper
We address the problem of efficient spatio-temporal clustering of speed data in road segments with multiple lanes. We postulate that the navigation/route plans typically reported by different providers as a single-value need not be accurate in multi-lane networks. Our methodology generates lane-aware distribution of speed from GPS data and agglomerates the basic space and time units into larger clusters. Thus, we achieve a compact description of speed variations which can be subsequently used for more accurate trips planning. We provide experiments that demonstrate the benefits of our proposed approaches.
Conference Paper
Full-text available
Dear reader, You are holding in your hands a volume of the series „Reports of the DLR-Institute of Transportation Systems“. We are publishing in this series fascinating, scientific topics from the Institute of Transportation Systems of the German Aerospace Center (Deutsches Zentrum für Luft- und Raumfahrt e.V. - DLR) and from his environment. We are providing libraries with a part of the circulation. Outstanding scientific contributions and dissertations are here published as well as projects reports and proceedings of conferences in our house with different contributors from science, economy and politics. With this series we are pursuing the objective to enable a broad access to scientific works and results. We are using the series as well as to promote practically young researchers by the publication of the dissertation of our staff and external doctoral candidates, too. Publications are important milestones on the academic career path. With the series „Reports of the DLR-Institute of Transportation Systems / Berichte aus dem DLR-Institut für Verkehrssystem¬technik“ we are widening the spectrum of possible publications with a bulding block. Beyond that we understand the communication of our scientific fields of research as a contribution to the national and international research landscape in the fiels of automotive, railway systems and traffic management. This volume contains the proceedings of the SUMO2014 – Modeling Mobility with Open Data, which was held from 15th to 16th May 2014 in Berlin-Adlershof, Germany. SUMO is a well established microscopic traffic simulation suite which has been available since 2002 and provides a wide range of traffic planning and simulation tools. The conference proceedings give a good overview of the applicability and usefulness of simulation tools like SUMO ranging from new methods in traffic control and vehicular communication to the simulation of complete cities. Another aspect of the tool suite, its universal extensibility due to the availability of the source code, is reflected in contributions covering parallelization and interfacing improvements to govern microscopic traffic simulation results. The major topic of this second edition of the SUMO conference is open data. Several articles cover the acquisition and refinement of traffic networks as one of the fundamental data sources. Subsequent specialized issues such as data models for emissions and Bluetooth simulation are targeted as well. The conference’s aim was bringing together the large international user community and exchanging experience in using SUMO, while presenting results or solutions obtained using the software or modeling mobility with open data. Let you inspire to try your next project with the SUMO suite. There are many new applications in your environment. Prof. Dr.-Ing. Karsten Lemmer
Article
Full-text available
Abstract— The design in this project showcase the implementation of arduino base intelligent multitasking system for milk tanker. This system consisting of the Arduino , GPS & GSM Module, LCD display, Magnetic reed relay sensor & PT100 sensor. It performs the function like temperature sensing and real time path tracking. It also includes system for safety messages, warning messages and traffic conjunction messages By using agnetic reed relay and limit switch we can detect if any fault occur at knob side. This system operates with the aid of Arduino, sensors , GPS & GSM module which will control the whole function of the system.
Article
Bicycling has become increasingly popular in the United States in recent years for both recreation and utilitarian purposes. Yet, attributes of the bicycle riding experience and riding differences between adults and children and males and females are not well documented. Most existing data on bicycling trip characteristics are based on self-reported interviews or surveys, which are prone to recall bias. The purpose of this exploratory study was to capture naturalistic bicycling data to examine trip characteristics and compare exposure classification accuracy between GPS and video data. We enrolled 10 children and 10 adults and captured their bicycling trips for one week each using PedalPortal, a GPS-enabled helmet camera data capture and coding system developed by the authors and a team of engineering students. Overall, 261 trips, 57 h, and 670 miles of bicycling were captured. The video data allowed for correct classification of riding location (sidewalk, bicycle lane, street, etc.), an advantage over GPS data alone. Child trips were significantly shorter in both time and distance than adult trips (p<0.01). The majority of male trips were commutes (69.8% child, 60.5% adult), while female trips were more evenly distributed among commute, non-commute utilitarian, and recreation. Adults primarily chose paved streets with no bicycle facilities, but also sought out on-road bicycle facilities (bike lanes and shared lane markings). Children rode most frequently on sidewalks. Results from this study demonstrate that the addition of video data can improve classification of bicycling exposure and differences by age and gender that can help planners and engineers better understand bicyclist behavior variations and increase safety by selecting appropriate and targeted countermeasures.
Chapter
The map matching problem arises when GPS measurements are incorrectly assigned to the roadway network in a GIS environment. This chapter presents a real-time topological decision rule-based methodology that detects and solves spatial mismatches as GPS measurements are collected. A real-time map matching methodology is required in several applications, such as fleet management, transit control and management, and travel behavior studies, in which decision-making must be performed simultaneously with the movement of vehicles, individuals, or objects. A computational implementation in a real case scenario in Chile indicates that the algorithm successfully resolves over 96% of the spatial mismatches encountered in real time. Various algorithmic parameter values were employed to test the performance of the algorithm for data collected every 5 and 10 seconds. Overall, the algorithm requires larger buffer sizes and speed ranges to obtain better results with lower spatial data qualities.
Article
To identify the determinants of bike share users' route choices, this research collects 132,397 hub-to-hub global positioning system (GPS) trajectories over a 12-month period between April 1, 2015 and March 31, 2016 from 750 bicycles provided by Hamilton Bike Share (HBS). A GIS-based map-matching algorithm is used to derive users' routes along the cycling network within Hamilton, Ontario and generate multiple attributes for each route, such as route distance, route directness, average distance between intersections, and the number of turns, intersections, and unique road segments. Concerning route choice analysis, the origin and destination pair should be the same for all routes within a choice set, thus HBS users' trips are grouped by origin-destination hub pairs. Since trips taken by different users between a hub pair can follow the same route, unique routes are extracted using a link signature extraction tool. Following this, a normalized Gini (Gn) coefficient is calculated for each hub pair to evaluate users' preferences among all the unique hub-to-hub route choices. A Gn closer to 0 indicates that routes between a hub pair are more evenly used, while a value closer to 1 implies a higher preference toward one dominant route. Three route choice models, a global model, a medium Gn model, and a high Gn model, are estimated using Path-Size Logit to determine how route choice is affected by the presence of dominant routes. These models suggest that HBS users are willing to detour for some attributes, such as bicycle facilities, but tend to avoid circuitous routes, turns, steep slopes, and roads with high traffic volume.
Article
Full-text available
Detection and correction of errors in map data based on spatial reasoning may be used to improve their quality. However, the majority of current spatial reasoning approaches are based on binary spatial relations and are not able to perform analyses involving more than two objects. This article proposes building accessibility analysis with the ternary ray intersection model to detect potential map errors. Where buildings are not accessible from the road network, this may indicate potential errors in map data such as roads that are not mapped. The plausibility of the proposed method was tested in a case study on OpenStreetMap data. The results have been published in an online mapping challenge where volunteering mappers have used them to correct errors in map data, and have provided feedback on the analysis. The results show that the proposed method can detect errors in map data that are caused by incorrect classification of buildings, incorrect mapping of multi‐part buildings, and missing road data.
Article
GPS-equipped bike-share fleets are a source of rich data that can be used to estimate cycling volumes to assist infrastructure investment decisions aimed at increasing ridership. Using global positioning system (GPS) trajectories collected between January 1st, 2018 and December 31st, 2018 by Hamilton Bike Share (HBS), the volume of bike share trips on every traveled link in the HBS service area is modeled. A map-matching toolkit is used to generate users’ routes to derive the number of observed bike share trips on every traveled link. To model annual bike share traffic volumes, several variables were created at the link level including accessibility measures, distances to important locations in the city, proximity to transportation infrastructure, and bike infrastructure. A linear regression model was estimated, incorporating eigenvector spatial filtering to remove spatial autocorrelation. The results suggest that the largest positive predictors of bike share traffic volumes in terms of cycling infrastructure are those that are physically separated from automobiles by a space or barrier. Additionally, hub-trip distance accessibility, a novel measure, was significant in the model, outperforming other accessibility metrics. A demonstration of how the model can be used for planning cycling infrastructure upgrades is presented.
Conference Paper
Full-text available
Bus link speed is an essential input for accurate estimation of arrival times for providing reliable service information to passengers. The speed data generated by the Global Positioning System (GPS) units affixed on buses have created an opportunity to derive reliable bus link speeds. To use GPS data, a critical requirement is to match the related positional data precisely on the respective link of its route. This research focuses on a post-processing technique for map matching for GPS data on a digital map with the assistance of Geographic Information Systems (GIS). The results from the GIS-based proposed model shows its ability to perform the map matching within a reasonably short processing time of 0.387 milliseconds per GPS observation. Moreover, the average hourly speeds of buses operating on a defined link were calculated with the developed algorithm and compared for verification purposes.
Article
To track a vehicle using GPS data, map matching is not only to select the map feature that the vehicle was at a GPS point, but also to find out the route it traveled. For GPS data at high polling rate, it is very obvious that the vehicle stays on the same map feature or a map feature directly connected with the previous GPS point. For low polling rates, a map feature where a vehicle is at a GPS point, is not always directly connected with the map feature where the vehicle is at the previous GPS point. In this paper, an improved model is proposed. The candidate set generation considers the difference of the width of links, and the matched map feature selection considers the length of the route, the difference between the azimuth of the GPS point and the map feature and the deviation of the raw GPS position to the candidate map feature. A field test shows that the improved model performs better than the previous model.
Article
Full-text available
GPS based navigation and route guidance systems are becoming increasingly popular among bus operators, fleet managers and travelers. To provide this functionality, one has to have a GPS receiver, a digital map of the traveled network and software that can associate (match) the user's position with a location on the digital map. Matching the user's location has to be done even when the GPS location and the underlying digital map have inaccuracies and errors. There are several approaches for solving this map matching task. Some only match the user's location to the nearest street node while others are able to locate the user at specific location on the traveled street segment. In this paper a topologically based matching procedure is presented. The procedure was tested with low quality GPS data to assess its robustness. The performance of the algorithms was found to produce outstanding results.
Article
Full-text available
Global Positioning System (GPS) technology can continuously monitor the time and location of vehicle usage. By recording and analyzing detailed vehicle activity data, researchers can analyze the safety and environmental implications of driver behavior and trip-making patterns. In 2000, NHTSA awarded the Georgia Institute of Technology a contract to equip 1,100 vehicles with a GPS-enhanced device to collect speed and location data. The objective was to acquire more accurate information on the role of excessive speed on crash frequency and severity. GPS technology allows the researcher to continuously measure driver speed, acceleration, and location. When merged with roadway characteristics within a geographic information system (GIS) environment, determinations of driver risk-taking behavior can be made. Second, continuous logging of GPS data allows researchers to capture high-resolution vehicle activity immediately before a crash event, reducing the potential error and bias introduced during determination of precrash speed estimates. Until May 1, 2000, the military degraded the position accuracy of GPS signals for commercial use, known as selective availability. For researchers, life without selective availability is a great improvement. Travel routes can clearly be discerned without the addition of differential correction units. The accuracy of speed, acceleration, and position data obtained from GPS signals for use in determining driver performance parameters without selective availability were tested. The test included four GPS packages, both corrected and uncorrected, simultaneously validated against a distance-measuring instrument. Equipment configuration, data collection methods, and sources of error are reported. The results suggested that noncorrected data can be used to obtain data within a reasonable range of the application requirements. Even without selective availability, GPS accuracy is still problematic in urban canyons and under heavy tree canopies. Although filtering for urban canyon outliers is labor intensive in a continuous monitoring situation, improvements in GIS hold promise for automation of this task.
Article
Full-text available
Since the first Global Positioning System (GPS) studies in the mid-1990s, this method of surveying individual travel behavior has gained attention in transport research. Compared with classic travel survey methods, GPS studies offer researchers benefits of more accurate and reliable informa- tion. At the same time, the participants' burden is reduced substantially if the GPS data collection does not involve time-consuming questions. However, without additional information, such as modes and trip pur- poses, extensive postprocessing is required to derive data that can be used for analysis and model estimation. The corresponding procedures are an ongoing research issue. This paper describes a postprocessing procedure needing no input other than the most basic GPS raw data: three-dimensional positions and the corresponding time stamps. First, the data are thoroughly cleaned and smoothed. Second, trips and activities are determined. Third, the trips are segmented into single-mode stages, and the transport mode for each of the stages is identified. The procedure is applied to GPS records collected in the Swiss cities of Zurich, Winterthur, and Geneva. A total of 4,882 participants carried an on-person GPS receiver for an average of 6.65 days. The results are compared with the Swiss Microcensus 2005 to demonstrate that derived data are ready for further applications, such as discrete choice model estimations.
Article
Full-text available
This article reports on the development of a trip reconstruction software tool for use in GPS-based personal travel surveys. Specifically, the tool enables the automatic processing of GPS traces of individual survey respondents in order to identify the road links traveled and modes used by each respondent for individual trips. Identifying the links is based on a conventional GIS-based map-matching algorithm and identifying the modes is a rule-based algorithm using attributes of four modes (walk, bicycle, bus and passenger-car). The tool was evaluated using GPS travel data collected for the study and a multi-modal transportation network model of downtown Toronto. The results show that the tool correctly detected about 79% of all links traveled and 92% of all trip modes.
Conference Paper
Full-text available
Map-matching is the process of aligning a sequence of observed user positions with the road network on a digital map. It is a fundamental pre-processing step for many applications, such as moving object management, traffic flow analysis, and driving directions. In practice there exists huge amount of low-sampling- rate (e.g., one point every 2-5 minutes) GPS trajectories. Unfortunately, most current map-matching approaches only deal with high-sampling-rate (typically one point every 10-30s) GPS data, and become less effective for low-sampling-rate points as the uncertainty in data increases. In this paper, we propose a novel global map-matching algorithm called ST-Matching for low- sampling-rate GPS trajectories. ST-Matching considers (1) the spatial geometric and topological structures of the road network and (2) the temporal/speed constraints of the trajectories. Based on spatio-temporal analysis, a candidate graph is constructed from which the best matching path sequence is identified. We compare ST-Matching with the incremental algorithm and Average-Fré chet-Distance (AFD) based global map-matching algorithm. The experiments are performed both on synthetic and real dataset. The results show that our ST-matching algorithm significantly outperform incremental algorithm in terms of matching accuracy for low-sampling trajectories. Meanwhile, when compared with AFD-based global algorithm, ST-Matching also improves accuracy as well as running time.
Conference Paper
Full-text available
The traditional way of representing motion in 3D space-time uses a trajectory, i.e. a sequence of (x,y,t) points. Such a trajectory may be produced by periodic sampling of a Global Positioning System (GPS) receiver. The are two problems with this representation of motion. First, imprecision due to errors (e.g. GPS receivers often produce off-the-road locations), and second, space complexity due to a large number of sam- plings. We examine an alternative representation, called a nonmaterial- ized trajectory, which addresses both problems by taking advantage of the a priori knowledge that the motion occurs on a transport network.
Conference Paper
Full-text available
Vehicle tracking data is an essential "raw" material for a broad range of applications such as traffic management and control, routing, and navigation. An important issue with this data is its accuracy. The method of sampling vehicular movement using GPS is affected by two error sources and consequently produces inaccurate trajectory data. To become useful, the data has to be related to the underlying road network by means of algorithms. We present three such algorithms that consider especially the trajectory nature of the data rather than simply the current position as in the typical map-matching case. An incremental algorithm is proposed that matches consecutive portions of the trajectory to the road network, effectively trading accuracy for speed of computation. In contrast, the two global algorithms compare the entire trajectory to candidate paths in the road network. The algorithms are evaluated in terms of (i) their running time and (ii) the quality of their matching result. Two novel quality measures utilizing the Fréchet distance are introduced and subsequently used in an experimental evaluation to assess the quality of matching real tracking data to a road network.
Article
Full-text available
Although the acquisition and maintenance of spatial data is very costly and time consuming very often the same objects of the real world are captured in many different data models, at different acquisition times, with different quality characteristics or at different scales. This situation will become intensified when more and more digital spatial data are offered by using internet technologies. Integration methods are needed to take advantage of the characteristics of more than one data set. These advantages could be, for example, new applications for which the data models had not been originally designed, higher reusability, improvement of the quality, or cost minimization of data acquisition. In this paper a relational matching approach for integration of spatial data from different sources is introduced. The research work was performed on street centrelines which were captured in different data models. The approach is based on statistical investigations between the data of two data models, and can be used in the same way for other data models. The matching problem is mapped onto a communication system, and measures, derived from information theory, are used to find an optimal solution. These measures can also be used for calculating local and global quality characteristics of the matching result.
Article
Full-text available
This article was published in the journal, GPS Solutions [© Springer-Verlag] and the definitive version is available at: http://dx.doi.org/10.1007/s10291-003-0069-z This paper describes a map-matching algorithm designed to support the navigational functions of a real-time vehicle performance and emissions monitoring system currently under development, and other transport telematics applications. The algorithm is used together with the outputs of an extended Kalman filter formulation for the integration of GPS and dead reckoning data, and a spatial digital database of the road network, to provide continuous, accurate and reliable vehicle location on a given road segment. This is irrespective of the constraints of the operational environment, thus alleviating outage and accuracy problems associated with the use of stand-alone location sensors. The map-matching algorithm has been tested using real field data and has been found to be superior to existing algorithms, particularly in how it performs at road intersections. Accepted for publication
Article
Speed and location observations from Global Positioning System (GPS) loggers are quickly becoming an important source of data for travel behavior researchers. Postprocessing these data requires identifying the location of the GPS data points on a coded map of the transportation network. The output of the map-matching process is the identification of the routes that were actually taken. This paper presents an innovative map-matching algorithm that relies only on the GPS coordinates and the network topology. Examples are provided on a large data set for the Zürich area. The paper demonstrates the efficiency of the algorithm in regard to accuracy and computational speed.
Article
Speed and location observations from Global Positioning System (GPS) loggers are quickly becoming an important source of data for travel behavior researchers. Postprocessing these data requires identifying the location of the GPS data points on a coded map of the transportation network. The output of the map-matching process is the identification of the routes that were actually taken. This paper presents an innovative map-matching algorithm that relies only on the GPS coordinates and the network topology. Examples are provided on a large data set for the Zürich area. The paper demonstrates the efficiency of the algorithm in regard to accuracy and computational speed.
Article
Advancements in global positioning system (GPS) technology now make GPS route choice data collection for travel diary studies and other transportation applications a reality. Opportunities abound for increased quantities of data, for improved quality of data, and for new data elements that were once considered too burdensome or expensive to capture. For example, automated travel diaries can electronically capture trip purpose, origin and destination location names, and driver and passenger names at the push of a button. An accompanying GPS receiver can accurately capture origin and destination locations, departure and arrival times, as well as trip lengths and travel routes. This wealth of data can be used to validate or calibrate travel demand models, for in-vehicle information systems analysis, and for modeling mobile source emissions across a given network. These data collection and processing advancements do have their costs, however. In fact, care and caution should be exercised when GPS technologies are selected and used to collect route choice data. The focus of this paper is on the accuracy issues related to route choice data collection and processing using GPS technology. Vendor specifications, observation techniques, data collection procedures, data postprocessing, and the importance of using a reliable and accurate geographic information system (GIS) database are examined in detail. Critical issues in the calculation of GPS accuracy are reviewed. Finally, recent experience in Atlanta is reported, and recommendations designed to reduce the introduction of error into automated route choice data collection are provided.
Article
An evaluation is presented of the effectiveness of Global Positioning System (GPS) data in capturing complex travel patterns characterized by multistop trip chains. This evaluation is made by analyzing trip-chaining patterns for a sample of individuals from whom travel data were collected using an innovative GPS device that was fitted to one of their household vehicles. By tracking the movements of household vehicles and automatically recording their locations, it is potentially possible to capture short, intermediate, and infrequent trips that would have been missed in a traditional travel diary survey. The primary objective of the study was to evaluate the effectiveness of GPS in capturing such trips. A secondary objective was to examine whether GPS trip data are suitable for analyzing trip-chaining behavior in the context of typical activities that people pursue outside the home. Descriptive statistical analysis techniques are used to explore trip-chaining patterns of individuals and model the propensity of linking various typical out-of-home activities in home-based chains. The results indicated that GPS data collection is more effective in capturing activity-linking behavior but merits some refinement to minimize the amount of missing data.
Article
The overlay process is currently one of the main computational solutions used to integrate several data layers from different sources. Unfortunately, it is problematic when trying to overlay many layers. This leads to several geometric problems such as the management of sliver polygons. This paper proposes a new merging process to complement the vector overlay for data integration of several layers. This process, based on measures derived from the Fréchet distance, matches common points (either lines or polygons). It also merges an ordered set of pairs of matching points (vertices) into a single geometry.
Article
The technology of Global Positioning Systems (GPS) provides new ways for collecting information about travel behavior. When it is used in combination with an electronic travel diary, valuable information of high quality about travel behavior becomes available. To learn about the possibilities of this new technology, a pilot study involving 151 people was performed in the Netherlands. What makes this pilot unique is the monitoring of all modes of travel, not just travel by motorized vehicles. The findings suggest that although it is possible to monitor all travel modes, data quality differs among them. The GPS device registered nearly all car driver trips, compared to half of the tram and train trips. When trips are registered, public transport trips are registered with a higher accuracy level than car trips. For specific trips, respondents did not use the equipment because of the burden placed on them or because of a lack of time. This was especially the case for walking, cycling, public transport, shopping trips, and visits. Concerning route, distance, and trip segments, it can be concluded that although valuable information is collected, there remain many difficulties to be overcome.
Article
Geographic integration using overlay operations remains one of the key features of GIS and has received much attention from geographers. In promotional materials integration is presented as the universal ability to combine heterogeneous sources of data into meaningful information based on location. While all data can be merged based on location, the significance and meaning of the resulting data is often dubious. Theoretically and practically, integration remains very complex. Cartographers have paid little attention to geographic integration, but, as this paper argues, distinctions between attributes and geometry in work on automated generalization helps develop approaches that can take this complexity into account. This paper examines the different meanings of integration and considers two types of integration (aligning and matching) based on differences in the handling of geometry and attributes. Integration is generally conceived of as a function, an operation, or a procedure. The application of GIS-based operations combines or amalgamates geometry and/or attributes without consideration of generalization operations. Automated generalization work has developed methods to mathematically model the relationships between geometry and attributes. The key distinction for GIS-based geographic integration is attribute handling. Aligning accounts either explicitly or implicitly for both geometry and attributes. Matching is the operation that merges geometry and data without any consideration of attributes. The distinction between geometry and attribute is important to assure the results of merging data according to location correspond to intents and constraints.
Article
Despite its application in many fields, map matching in studies of travel/transport geography is unique in two aspects: 1) The correct road links traversed by the traveler need to be unambiguously identified; 2) All the identified links should form a meaningful travel route. This paper discusses the application of map matching methodologies in the context of deriving people's travel behavior from GPS-traced multi-modal trip data. In recognition of the disadvantages associated with the existing algorithms, this research proposed and implemented a heterogeneous map matching approach suitable for travel/activity research needs which is uniquely characterized by: 1) data preprocessing with point cluster reduction and density leverage; 2) offering the candidate solution within a pool of "the best"; 3) the balancing of matching results from multiple matching factors with rank aggregation; 4) Utilizing the network constraint attributes to increase the matching accuracy; and 5) Use of the Dempster belief test to discern the noise and off-road travel.
Article
A test-bed application, called Map Matched GPS (MMGPS) processes raw GPS output data, from RINEX files, or GPS derived coordinates. This developed method uses absolute GPS positioning, map matched, to locate the vehicle on a road centre-line, when GPS is known to be sufficiently accurate. MMGPS software has now been adapted to incorporate positioning based on odometer derived distances (OMMGPS), when GPS positions are not available. Relative GPS positions are used to calibrate the odometer. If a GPS position is detected to be inaccurate, it is not used for positioning, or for calibrating the odometer correction factor. In OMMGPS, GPS pseudorange observations are combined with DTM height information and odometer positions to provide a vehicle position at ‘1 s’ epochs. The described experiment used GPS and odometer observations taken on a London bus on a predefined route in central of London. Therefore, map matching techniques are used to test GPS positioning accuracy, and to identify grossly inaccurate GPS positions. In total, over 15,000 vehicle positions were computed and tested using OMMGPS.In general, the position quality provided by GPS alone was extremely poor, due to multipath effects caused by the urban canyons of central London, so that odometer positioning was used much more often to position the vehicle than GPS. Typically, the ratio is 7:3 odometer positions to GPS positions. In the case of one particular trip, OMMGPS provides a mean error of position of 8.8 m compared with 53.7 m for raw GPS alone.
Article
In the United States, information about daily travel patterns is generally captured using self-reported information using a written diary and telephone retrieval (or mail-back of diary forms). Problems with these methods include lack of reporting for short trips, poor data quality on travel start and end times, total trip times and destination locations. This project combined a hand-held computer (Personal Digital Assistant or PDA) with a Global Positioning System (GPS) receiver to capture vehicle-based, daily travel information. The vehicle driver uses a menu to enter variables such as trip purpose and vehicle occupancy, but other data such as date, start time, end time, and vehicle position (latitude and longitude) are collected automatically at frequent intervals. The field test was conducted in Lexington, Kentucky in fall, 1996, with 100 households to use the equipment for six days. Respondents also completed a telephone survey for one day of travel (attempted for day 5). The field test was a test of equipment and willingness of the general public to participate, rather than to obtain a statistically valid travel behavior dataset for the Lexington area. One improvement to the hardware would be for the equipment to turn on automatically. There are limitations to the dataset and analyses that are discussed where appropriate. Although the dataset is small, this paper compares the results of the machine-recorded trips to self-reported trips captured by telephone interview. Self-reported distances are much longer than distances recorded by the PDA/GPS. A recalled distance of 10 miles was, on average, only 6.5 miles when the GPS points are matched to a positionally accurate base file. Similarly, recalled times generally exceed median measured values, but the differences are much smaller than for distances. Respondents reported that data entry of 1 min at the beginning of each trip over the six-day survey period was not burdensome.
Article
Third-generation personal navigation assistants (PNAs) (i.e., those that provide a map, the user's current location, and directions) must be able to reconcile the user's location with the underlying map. This process is known as map matching. Most existing research has focused on map matching when both the user's location and the map are known with a high degree of accuracy. However, there are many situations in which this is unlikely to be the case. Hence, this paper considers map matching algorithms that can be used to reconcile inaccurate locational data with an inaccurate map/network.
Article
Map-matching algorithms integrate positioning data with spatial road network data (roadway centrelines) to identify the correct link on which a vehicle is travelling and to determine the location of a vehicle on a link. A map-matching algorithm could be used as a key component to improve the performance of systems that support the navigation function of intelligent transport systems (ITS). The required horizontal positioning accuracy of such ITS applications is in the range of 1 m to 40 m (95%) with relatively stringent requirements placed on integrity (quality), continuity and system availability. A number of map-matching algorithms have been developed by researchers around the world using different techniques such as topological analysis of spatial road network data, probabilistic theory, Kalman filter, fuzzy logic, and belief theory. The performances of these algorithms have improved over the years due to the application of advanced techniques in the map matching processes and improvements in the quality of both positioning and spatial road network data. However, these algorithms are not always capable of supporting ITS applications with high required navigation performance, especially in difficult and complex environments such as dense urban areas. This suggests that research should be directed at identifying any constraints and limitations of existing map matching algorithms as a prerequisite for the formulation of algorithm improvements. The objectives of this paper are thus to uncover the constraints and limitations by an in-depth literature review and to recommend ideas to address them. This paper also highlights the potential impacts of the forthcoming European Galileo system and the European Geostationary Overlay Service (EGNOS) on the performance of map matching algorithms. Although not addressed in detail, the paper also presents some ideas for monitoring the integrity of map-matching algorithms. The map-matching algorithms considered in this paper are generic and do not assume knowledge of ‘future’ information (i.e. based on either cost or time). Clearly, such data would result in relatively simple map-matching algorithms.
Non-Response Challenges in GPS-Based Surveys
  • S Bricka
Bricka, S. (2008) Non-Response Challenges in GPS-Based Surveys, paper at the 8 th International Conference on Travel Survey Methods, May 2006, Annecy, France.
Lexington Area Travel Data Collection Test: GPS for Personal Travel Surveys. Final Report. Office of Highway Policy Information and Office of Technology Applications, Battelle Transport Division, FHWA
  • D P Wagner
Wagner, D. P. (1997) Lexington Area Travel Data Collection Test: GPS for Personal Travel Surveys. Final Report. Office of Highway Policy Information and Office of Technology Applications, Battelle Transport Division, FHWA, Sept. 1997, Columbus, Ohio.
Defining unmoveable nodes/segments as part of vector overlay: The alignment overlay
  • F Harvey
Harvey, F. (1994) Defining unmoveable nodes/segments as part of vector overlay: The alignment overlay. In Advances in GIS Research, T. C. Waugh and R. C. Healey. London, Taylor and Francis, 1, pp. 159-176.
Global Positioning System (GPS) and Travel Surveys: Results from the 1997 Austin Household Survey
  • D Pearson
Pearson, D. (2001) Global Positioning System (GPS) and Travel Surveys: Results from the 1997 Austin Household Survey. Presented at 8th Conference on the Application of Transportation Planning Methods, April 2001, Corpus Christi, Texas.
Geometric matching of polygonal surfaces in GISs
  • F Vauglin
  • A Bel Hadj Ali
Vauglin, F. and Bel Hadj Ali, A. (1998) Geometric matching of polygonal surfaces in GISs. ASPRS Annual Meeting, Tampa, FL, ASPRS.
Geographic data matching: First investigations for a generic tool
  • C Lemarié
  • L Raynal
Lemarié, C. and Raynal, L. (1996) Geographic data matching: First investigations for a generic tool. GIS/LIS '96, Denver, Co, ASPRS/AAG/URISA/AM-FM.
Trip Reporting in Household Travel Diaries: A Comparison to GPS-Collected Data
  • J Casas
  • C H Arce
Casas, J., and Arce, C. H. (1999) Trip Reporting in Household Travel Diaries: A Comparison to GPS-Collected Data. Presented at 78th Annual Meeting of the Transportation Research Board, Washington, D.C.
Appariement geometrique des objets géographiques et étude des indicateurs de qualité
  • A Bel Hadj Ali
Bel Hadj Ali, A. (1997) Appariement geometrique des objets géographiques et étude des indicateurs de qualité. Saint-Mandé (Paris), Laboratoire COGIT.
Defining unmoveable nodes/segments as part of vector overlay: The alignment overlay
  • F Harvey
  • T C Gis Research
  • R C Waugh
  • Healey
Analysis of Global Positioning System-Based Data Collection Methods for Capturing Multistop Trip-Chaining Behavior
  • L Yalamanchili
  • R M Pendyala
  • N Prabaharan
  • P Chakravarty