• DocumentCode
    2857571
  • Title

    On-Line Analytical Process Tools for the VISION Engineering Performance Data Mart

  • Author

    Yetzer, Ken ; Warnock, Andrew ; Krishnamurthy, Ashok

  • Author_Institution
    Ohio Supercomput. Center, Columbus
  • fYear
    2007
  • fDate
    18-21 June 2007
  • Firstpage
    277
  • Lastpage
    282
  • Abstract
    VISION (versatile information system - integrated, on-line) represents a comprehensive, holistic, top down approach to information collection, management, and ultimate transformation into knowledge. VISION employs a Web based approach that combines a modular instrumentation suite and a digital library (VDLS) with modern communications technology. VDLS is a Web based collaborative knowledge management system that incorporates associated data marts and additional test information sources such as reports, analyses and documents. One of these data marts contains engineering performance data. A Web based online analytical processing (OLAP) toolbox has been developed to allow querying of this data mart via a Java graphical user interface (GUI). Currently, the OLAP toolbox contains functions to perform metadata searches, view a Global Positioning System (GPS) map of the location of the item under test, view time series traces of all the parameters, generate custom plots of any parameter or download raw data files. The work described in this paper adds additional functions to this toolbox that allows test engineers, data analysts or mechanical engineers to quickly find the data of interest. Tools were developed that use wavelet transforms for both data validation purposes and for data de-noising. Other tools were created that use Google Earth to plot GPS coordinates as markers where each marker contains a balloon of information (e.g., time, date, latitude, longitude, speed, direction). In addition, Google Earth was used in a spatial data mining application where the GPS coordinates of extreme values (i.e., values exceeding a given threshold) are plotted along the track. All of the algorithms developed in this paper were tested using a dataset from the 10 ton dump truck of the family of medium tactical vehicles (FMTV). Up to 35 sensors were used in these tests measuring everything from GPS position to engine speed to strain rate. The tests were conducted at the Aberdeen Test- Center on a variety of courses covering everything from smooth paved surfaces to block gravel to dirt.
  • Keywords
    Global Positioning System; Internet; Java; data mining; data warehouses; graphical user interfaces; information management; knowledge management; meta data; query processing; visual databases; wavelet transforms; Global Positioning System; Google Earth; Java graphical user interface; VISION engineering performance; Web based collaborative knowledge management system; World Wide Web; communications technology; data denoising; data mart; data validation; digital library; information collection; information knowledge; information management; metadata search; online analytical processing tool; querying; spatial data mining; versatile information system-integrated online; wavelet transform; Data engineering; Earth; Global Positioning System; Graphical user interfaces; Knowledge management; Machine vision; Management information systems; Performance analysis; Strain measurement; System testing;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    DoD High Performance Computing Modernization Program Users Group Conference, 2007
  • Conference_Location
    Pittsburgh, PA
  • Print_ISBN
    978-0-7695-3088-5
  • Type

    conf

  • DOI
    10.1109/HPCMP-UGC.2007.57
  • Filename
    4437997