28th September - 30th September 2008 | Vienna, Austria

Beyond Google Earth: Surfing Petabyte Maps in the Future Internet

Attention: open in a new window. PDFPrintE-mail

Tutorial 2
Beyond Google Earth: Surfing Petabyte Maps in the Future Internet
Sunday, September 28, half-day

Call for Participation

Peter Baumann I Jacobs University Bremen, DE

Abstract

Since the launch of Google Earth at the latest it is clear that online services for multi-Terabyte satellite imagery are becom­ing integral part of our Internet experience. Actually, 2-D imag­ery is but the tip of the iceberg - the general concept of multi-dim­ensional spatio-temporal raster data covers 1-D sensor time series, 2-D imagery, 3-D CAT scans (x/y/t) and ex­plor­at­ion data (x/y/z), 4-D climate, ocean, and cosmological simulation results as well as life science microarray data (x/y/z/t), and many more. The common data abstraction behind all these is that of a multi-dimensional array of some extent and cell (“pixel”, “voxel”) type. Sizes range well into multi-Terabyte object sizes, in future: multi-Petabyte. Today’s inexpensive high-capacity storage media indeed allow to give online access to such data. We face the appearance of navigational interfaces; the next step will consist of advancing from data stewardship to service stewardship based on open, flexible access interfaces for value-adding processing, analysis, and mining.

Examples are manifold: Sensor and streaming databases will allow ad-hoc summarization as well as standing queries for alerting. Hyperspectral satellite iamgery will not just be served as is, but derived products like vegetation index or snow index will be computed on the fly and without redundant storage. Human brain imaging will benefit from analysing thousands of brain activity maps simultaneously. Multi-Petabyte datacubes can be leveraged for online analysis. This poses new challenges on data manage­ment and, in particular, the design of open, inter­operable services.

In this tutorial we present the state of the art in semantic Web services for large-scale, multi-dimensional raster data. The tutorial will give an introduction to raster services, addressing modeling, architecture, optimization, and storage management. High emphasis will be devoted to applications in geo, life science, and the Grid; to this end, real-life use cases will be presented and discussed which stem from both projects carried out and our standardization work in the Open GeoSpatial Consortium (OGC, www.opengeospatial.org).

 

Fig.: collage of geo and life science raster applications
(origin: all but top-right are screenshots from our projects)

Outline and Schedule

09:00 - 09:20 Introduction: the data; the volumes; the application domains; the challenge; live demo
09:20 - 09:50 Modeling raster services: request languages; desirable properties of a language
09:50 - 10:20 Architecting raster services I: tiled array storage; storage management strategies
10:20 - 10:30 break
10:30 - 11:00 Architecting raster services II:
query evaluation; optimization; parallelization
11:00 - 11:30 A caleidoscope of scientific use cases:
-Open GeoSpatial Consortium (OGC) standards for 1D – 4D geo services: mapping, mining, and more
-Grid: Climate Modelling, Cosmologic Simulation
-Life Science: Genomic Data, Brain Imaging
11:30 - 12:00 Hands-on Web-based experimenting with 1-D to 4-D geo data sets
  Wrap-Up: research issues and an integrating vision; discussion

Format Description

The tutorial will start as a lecture. Later, participants with an Internet-connected laptop will be able to experiment with image services by writing their own requests.

Intended Audience

The tutorial is suitable for a large audience, including advanced users, beginning programmers, and system architects: while the overall presentation does not require deep pre-existing knowledge, at some points selected advanced issues will be discussed. The application domains, earth and life sciences, probably are not so common to the community participating, hence proper introduction will be given.

Prerequisites

General background in computer science, Internet, and Web services. Initial database knowledge is advantageous, but not necessary. Earth & life science background is not required. In particular, no programming know-how is necessary to for the hands-on experimenting session.

Presenter

Dr. Peter Baumann is Professor of Computer Science at Jacobs University Bremen where he teaches databases, Web services, and software engineering. Additionally, he is founder and CEO of rasdaman GmbH (www.rasdaman.com), a research spin-off specializing in software solutions for large-scale raster services. Previously he was lecturing at the University of the Armed Forces Munich, and Assistant Head of the Knowledge Bases Research Group at the Bavarian Research Center for Knowledge-Based Systems.

Peter Baumann's core research interest is large-scale multidimensional raster services and their application in geo, life science, Grid, and e-learning. He has published more than 70 book chapters and journal/conference articles in the areas of raster databases and their applications, graphics databases, CAD databases, and algebraic specification. He holds international patents on raster database technology and has received numerous national and international innovation awards for his work, such as the European IT Prize.

In the Open Geospatial Consortium (OGC, www.opengeospatial.org) standardization body he chairs the working groups on geo raster services standardization. Further, he is founding member and secretary of CODATA Germany, member of the Object Database Management Systems group (www.odbms.org), council member of the Commission for the Management and Application of Geoscience Information (a Commission of the International Union of Geological Sciences), and of national geo data bodies. Earlier he has been member of the OHBM (International Organization of Human Brain Mapping) Task Force on Neuroinformatics and of the EC-US Steering Committee on Neuroinformatics.

See www.faculty.jacobs-university.de/pbaumann for more information.

To contact please go:
Peter Baumann This e-mail address is being protected from spambots. You need JavaScript enabled to view it , Jacobs University Bremen, DE

Important Dates

Tutorial date: 28 September
Time: 09:00 – 13:00