Techniques for Information Extractions from Compressed GPS Traces

Nowadays, Global Positioning System (GPS) devices are routinely installed in motorized vehicles. These devices generate huge volumes of trace (or trajectory) data, with each trace giving the position (latitude and longitude) of a vehicle over time. GPS traces contain information that is valuable to many stakeholders such as transportation planners, policy analysts and business organizations (e.g. trucking industry and taxi companies). Such traces are often compressed to eliminate redundancy and reduce the amount of storage space. When additional data about vehicles (e.g. freight information and readings from on-board sensors for trucks, fare and occupancy information for taxicabs) is available, the combination of compressed trace data and vehicular data serves as a richer source of information that is useful in multiple application scenarios. The proposed research investigates methods to effectively extract information from large volumes of compressed GPS traces and other vehicular data. The specific tasks which will be carried out during this work are as follows: Task 1: Develop efficient techniques for retrieving from a database of compressed traces, a collection of traces that are similar to a given query trace. Such techniques should also have the capability to classify the retrieved collection of traces according to specified criteria. Task 2: Extend compression techniques for GPS traces so that traces that include other vehicular or sensor data (along with latitude, longitude and time) can also be effectively compressed. As in Task 1, such extensions should allow efficient retrievals of collections of traces that are similar to a given query trace. The proposed work falls under Focus Area 4 ("System modernization through implementation of advanced information technologies") of University Transportation Research Center (UTRC) Region 2. Potential long term benefits of the proposed research include the development of effective methods for extracting useful information from compressed representations of trajectories and other vehicular data. Such methods will be highly beneficial in processing complex queries that are of interest to transportation planners and other stakeholders. The deliverables of this project include software tools, research reports, papers in conferences/journals, a research brief suitable for distribution to policy makers and data sets generated as part of the work. These deliverables will be made available to the research community through an appropriate website.


  • English


  • Status: Completed
  • Funding: $103875.00
  • Contract Numbers:


  • Sponsor Organizations:

    Research and Innovative Technology Administration

    University Transportation Centers Program
    1200 New Jersey Avenue, SE
    Washington, DC  United States  20590

    University Transportation Research Center

    City College of New York
    Marshak Hall, Suite 910, 160 Convent Avenue
    New York, NY  United States  10031
  • Project Managers:

    Eickemeyer, Penny

  • Performing Organizations:

    State University of New York, Albany

    1400 Washington Avenue
    Albany, NY  United States  12222
  • Principal Investigators:

    Hwang, Jeong-Hyon

    Ravi, Sekharipuram

    Lawson, Catherine

  • Start Date: 20140301
  • Expected Completion Date: 0
  • Actual Completion Date: 20151231
  • Source Data: RiP Project 36320

Subject/Index Terms

Filing Info

  • Accession Number: 01564696
  • Record Type: Research project
  • Source Agency: University Transportation Research Center
  • Contract Numbers: 49997-39-25
  • Files: UTC, RIP
  • Created Date: May 27 2015 1:01AM