Industry

Development of Web-Based Transit Trip-Planning System Based on Service-Oriented Architecture

Description
Development of Web-Based Transit Trip-Planning System Based on Service-Oriented Architecture
Categories
Published
of 17
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
    DEVELOPMENT OF WEB-BASED TRANSIT TRIP PLANNING SYSTEM BASED ON THE SERVICE ORIENTED ARCHITECTURE Daniel(Jian) Sun*, Ph.D Assistant Professor School of Transportation Engineering TongJi University 4800 Cao ’ An Road, Jia-Ding District Shanghai 201804, China Phone: (86-21) 6958-3695 Email:  jiansun@ufl.edu Zhong-Ren Peng, Ph.D, Chair and Professor Department of Urban and Regional Planning University of Florida P. O. Box 115706 Gainesville, FL 32611-5706 Email: zpeng@dcp.ufl.edu And Cheung Kong Chair Professor School of Transportation Engineering, Tongji University Shanghai, China  Weiya Chen, Ph.D Assistant Professor School of Traffic and Transportation Engineering Central South University 932 South Lu-shan Road, Changsha 410075, China Phone: (86-731) 8876739 Email: wei-ya.chen@163.com Xiaofang Shan, Ph.D Assistant Professor School of Economics and Management TongJi University Shanghai 20092, China Phone: (86) 133-1173-9866 Email: xiaofangsh@tongji.edu.cn Word count: 6,053 + 5 Figures + 1 Table = 7,553 Paper submitted for possible presentation and publication at the 90 th  Annual Meeting of the Transportation Research Board in response to the Call for Papers: Emerging and Innovative Public Transport and Technologies Call for Papers (AP020)  Nov 15 th , 2010   * Corresponding author TRB 2011 Annual Meeting Paper revised from srcinal submittal.  Sun, Peng, Chen and Shan, 2011 2 Abstract 1 2 The majority of transit trip planners currently exist as proprietary systems based on particular vendor 3  products. With more functional components incorporated, the system maintenance and regular transit 4 information updates become burdensome tasks for the transit agencies. Additionally, the proprietary 5 nature of the systems makes it difficult to take advantage of the rapid advancement of geospatial 6 information and Web technologies. This paper proposed an open and interoperable transit trip 7  planning system based on the service-oriented architecture (SOA), with the principle to reuse the 8 existing modular resources, while providing friendly interfaces for future functionality expansion. The 9 objective is to integrate geospatial services available online (such as Google Maps), open-source 10 geospatial database technologies, and path-finding algorithms in a loose-coupled manner. The 11  proposed system was developed with the spatial and temporal transit data from Waukesha Metro, WI. 12 The search results were validated by comparing the outputs from the existing South-East Wisconsin 13 Transit Trip Planner, and the route schedule matching. The comparison results show that the new 14 service-oriented architecture provides a flexible and efficient mechanism for transit trip planners that 15 takes advantage of rapidly changing online geospatial services, yet maintains the core functions of 16 itinerary search that may be unique to each transit agency. 17 18 Key Words : Internet GIS; Service-oriented architecture; Transit customer services; Transit trip planning 19 system. 20 21 22 23 TRB 2011 Annual Meeting Paper revised from srcinal submittal.  Sun, Peng, Chen and Shan, 2011 3 1. INTRODUCTION  1 Disseminating transit-related information appropriately and timely is a goal that most 2 transit agencies are striving for. Results from a survey on the effect of advanced transit 3 indicate that about 38% of non-transit users would consider transit if appropriate 4 information was available ( 1 ). Historically, transit agencies relied on call centers or route 5 and schedule brochures to provide passengers the schedule and route information. However, 6 this method is inefficient, especially in cases where the route network is complicated and in 7 a grid manner. Moreover, transit schedules are changing regularly, which brings large 8 redundant work in the brochure update and dissemination. This motivates the development 9 of online transit trip planning systems to assist transit users to schedule trip itinerary and 10 acquire instructive transfer and bus departure information before the trip begins. Web-based 11 online transit information systems facilitate transit information update and dissemination, 12 since the schedule and route information are typically stored in the data server, and only 13 require to be updated once in a while centrally. 14 15 Over the last twenty years, as technology changes, Web-based online transit trip planning 16 systems have been evolving in three major aspects: path-finding algorithms, 17 spatio-temporal data models and system architectures. Path-finding algorithms aim to 18 develop efficient procedures to search for trip itineraries or paths in a transit network given 19 the trip srcin, destination and time of travel. Spatio-temporal data models aim to construct 20 efficient and robust spatio-temporal data structure to better store and search transit 21 schedules and route/stop location data. System architecture research is to develop efficient 22 architecture to link end users, data, and path-finding algorithms together as a functional trip 23 itinerary system. A considerable amount of research has been conducted in the development 24 of path-finding algorithms and spatio-temporal data models ( 2-6  ), as well as on the system 25 architecture ( 7-9 ). However, the system architecture seems to be the least developed and 26 the most problematic one, especially in the environment of rapid advancement of Internet 27 technology and Web applications. The main problem within the system architecture is the 28  proprietary nature of the systems, which are based on particular vendor products, rather 29 than an open and interoperable system. The challenge is how to design an architecture 30 framework to largely utilize existing resources (i.e., proprietary components and online 31 resources), while provide expansible and extensible interfaces for future functionality 32 upgrade. 33 34 This paper documents and reports our recent work in advancing the state of the art in transit 35 trip planner architecture, which was built upon our work in this area over the last decade. 36 Specifically, the paper describes the efforts to develop an interoperable transit trip planning 37 system (Waukesha Metro Transit Trip Planner) based on a service-oriented architecture 38 (SOA) by taking advantage of existing geospatial services that are available online. These 39 include Google Maps, PostgreSql database management system, AJAX (Asynchronous 40 JavaScript and XML) technology, as well as the existing path-finding components that 41 were developed previously. The merit of this system architecture is in its interoperable and 42 non-proprietary nature, especially in taking advantage of many existing resources and free 43 services online. Even some services, such as Google Maps, are based on existing business 44 models and are subject to change in the future, but still the potential cost savings in the 45 TRB 2011 Annual Meeting Paper revised from srcinal submittal.  Sun, Peng, Chen and Shan, 2011 4 immediate near term are very real and thus worthy of consideration as proposed. The 1 general purpose of this research is to present an interoperable Internet GIS approach in 2 developing an online transit trip planner by adopting the service-oriented architecture and 3 computing environment. More specifically, three sub-objectives are: 4 5    To develop a new system framework and introduce SOA for the existing components 6 reuse and future functionality expansion; 7 8    To design an interactive graphic user interface (GUI) with embedded geospatial 9 analysis and map rendering components from PostgreSql database and Google Maps; 10 and 11 12    To improve and implement a schedule-based path-finding algorithm based on a 13  pattern-centered spatio-temporal transit network model 14 15 The remainder of this paper is organized as follows: Section 2 reviews the literature on the 16 existing transit trip planning systems and SOA. Next, descriptions of the system framework 17 and the basic functional components implemented are presented in Section 3, where the 18 feasibility and necessity of SOA are discussed, with a focus on implementation details. In 19 Section 4, the key functionalities, such as the start/end bus stops positioning, geospatial 20 operations and transit network analysis components are explained, followed by discussions 21 on encapsulating them as online services. Finally, conclusions and recommendations for 22 future work are provided in Section 5. 23 24 2. LITERATURE REVIEW 25 This section starts by reviewing several widely studied and used online trip planning 26 systems from the perspectives of path-finding algorithms, spatio-temporal data models and 27 system architectures. Next, Web service based SOA, as a new approach to address 28 distributed computing, was introduced as a potential solution to improve and organize the 29 existing systems. Finally, the application of SOA technologies in building a trip planner 30 system is proposed. 31 32 2.1 Web-based Transit Trip Planning Systems 33 Many Web-based trip planning systems, from simple static text schedule display to more 34 sophisticated real time trip estimation, have been built during the last twenty years ( 10, 11 ). 35 Peng and Huang ( 8 ) classified the online transit information systems from two dimensions: 36 information content and system functionality. The content information ranges from static 37 information to dynamic real time information: 38    Type A, basic information about the agency and services; 39    Type B, static information about transit routes, service schedule and fares; 40    Type C, trip planning and itinerary information; and 41    Type D, real time information about bus locations and the expected arrivals 42 The system functionality ranges from information dissemination to interactive 43 communication and online transaction. By this classification, many existing online trip 44 TRB 2011 Annual Meeting Paper revised from srcinal submittal.  Sun, Peng, Chen and Shan, 2011 5  planners, including the trip planning systems to be reviewed, focus on disseminating or 1  providing interactive static transit information or trip itinerary information. 2 3 Google Transit Trip Planner (http://www.google.com/transit) has been used in many cities 4 to access public transit schedules and routes, and to plan trips ( 12 ). Transit information 5  provided by this system includes: 1) the upcoming departures, step by step directions for 6 the entire itinerary, and travel time and transfer information; and 2) the recommended bus 7 route drawn along the streets on Google Maps, which provide free quality maps with 8 automatic Geo-spatial features update. In view of path finding, Google Transit requires 9 users to input the srcin and destination locations, together with the scheduled 10 departure/arrival times to obtain the trip route(s) and schedules. The spatio-temporal data 11 structures of Google Transit are largely dependent on Google Maps, for both search results 12 display and users ’  point-and-click on the map to define the srcin and destination. One 13 disadvantage is that the transfer points of search results are not marked out on the map. 14 Moreover, the system architecture and path-finding algorithms of Google Transit are sealed 15 in a black box, and hidden from the users with no detailed technical “white paper” 16 available. The system is not designed to connect with existing transit trip planning systems 17 that many transit agencies have already. Other limitations include that Google Transit only 18 supports “shortest -  path” searches with specific departure time or arrival time, with no other 19 search options (e.g. minimal transfers) provided. 20 21 Cherry et al. ( 13 ) designed an interactive online trip planner for the Sun Tran bus network 22 in Tucson, AZ. In view of path finding, users of this system can prepare the trip srcin and 23 destination from three options: manually typing in text addresses, selecting from landmark 24  pull-down menus, or clicking the geospatial locations directly on the map. A 25 forward-searching algorithm was implemented and executed either from the srcin to the 26 destination, or from the destination back to the srcin ( 14 ). With each O-D bus stop pair, 27 the algorithm traces a minimal possible time path from node (bus stop) to node along the 28 transit network. If the system cannot find a path to the destination bus stop on the same 29 route as the srcin, it then looks for transfer nodes, which are defined as time points in the 30 schedule serving multiple routes. Once a transfer node is found, the algorithm begins to 31  perform a similar search along the new route toward the destination. In this algorithm, all 32  possible routes are considered and the one with the shortest travel time is selected. Finally, 33 the users have the option to determine the optimal path from the existing schedule or from 34 historical bus arrival time data, which is presented in both text instructions and vector route 35 overlapped on the base map. As indicated, commercial GIS software - ArcGIS and ArcIMS 36 were adopted to provide map-based GUI. This means the transit agency has to prepare the 37 spatio-temporal data in the form of both schedule files and the GIS map layers, such as the 38 city streets, bus routes and bus stops, and maintain these files accordingly. The trip planner 39 is strictly dependent on ArcIMS for the three stages of site development: map composition, 40 Web service creation and publishing, which is not flexible and interoperable for the regular 41 maintenance and system upgrade. Lastly, the proposed GIS functionalities were not 42 available online at the time of writing this paper. By going through the website 43 (http://infoweb.suntran.com/hiwire?.a=iTripPlanning), it was found that the existing system 44 only generated text-based instructions without map renderings. 45 46 TRB 2011 Annual Meeting Paper revised from srcinal submittal.
Search
Similar documents
View more...
Tags
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks