13
Submitter:

Institution/Business:
Phone Number:
E-mail Address:
WWW Address:
General Topic:
Topic Title:


George Edwards

Electrical Engineering
University of Denver
(303) 871-4252
gedwards@du.edu

Telecom
Diff-Serve and Cellular IP

Description

Diff-Serve is very important as one tries to scale the Internet to provide quality-of-service (QoS) in the IP environment. Equally, Cellular IP is very important as one tries to make web-browsing and multimedia capabilies available in a wireless, cellular environment. In the case of the former I am proposing to do research at the edge router to facilitate the the achievement of QoS inside the core routers. Specifically, I am proposing traffic shaping, etc. using via conventional as well as with the use of intelligent algorithms such as fuzzy or neural network. In the latter area I am would examine ways to make the network more efficient in respect to mobility and use shorter paths than in Mobile IP when a mobile migrates a foreign agent.

14
Submitter:

Institution/Business:
Phone Number:
E-mail Address:
WWW Address:
General Topic:
Topic Title:


William Martz

College of Business - Information Systems Department
University of Colorado, Colorado Springs
(719) 262-3414
wmartz@brain.uccs.edu
http://www.uccs.edu/~wmartz
Organizational Memory; Knowledge Management
Making Organization Memory Effective Description

Description

Companies are beginning to realize that simply storing data in warehouses and databases is not sufficient to ensure the usefulness of that data or information. Organizations must learn how to store and transfer knowledge over time and not rely on recreating it. Organizational Memories are the means by which organizations remember. They are both the software (database, knowledge base) and content (model, paradigm). With human memory as an analogy, organizational memories represent the long-term memory of organizations from which the organization wishes to extract accurately the information stored and use again, as with short-term memory, at a future date. We have preliminary research (1) that shows storing certain types of complementary information, comparative comments, influences the accuracy of the information retrieved.

The proposed research would expand the preliminary study to more formally identify what type of complementary information is best stored in an organizational memory in order to extract the original meaning or intent. A combination of lab experiments and on-site, field studies would test fundamental questions. For example: Does archived information on which a decision was based create the same decision when retrieved from the organizational memory? Do hypertext documents provide better information/knowledge transfer than plain text? What data/information do decision makers really use in making decisions? What information is best used to summarize decisions? Ultimately, the answers to these questions can help organizations understand what information is best stored to support ongoing information storage, decision making, and knowledge management in their organization.

(1) Martz, Wm. Benjamin, Jr. and Morgan Shepherd, "Getting More Out of Organizational Memory: Comparative Comments and Output Medium," Journal of Computer Information Systems, Vol. 41, No.4, pp90-94, 2001.

15
Submitter:

Institution/Business:
Phone Number:
E-mail Address:
WWW Address:
General Topic:
Topic Title:


Chris Debrunner

Division of Engineering
Colorado School of Mines
(970) 491-4608
ghosh@cs.colostate.edu
http://www.cs.colostate.edu/~ghosh/
Visual Perception
Near-field motion estimation for pose estimation

Description

In order to navigate to a desired location or to build a map, a robotic vehicle must know its position in its environment. GPS can be used for absolute position, but it is not always available and is not accurate enough for some navigation and mapping tasks. Inertial navigation systems, odometers, and compasses can accurately determine relative vehicle position over short times, but over longer times errors accumulate to significant levels. Navigation based on the near field motion of features of known or unknown location can complement these other position estimation methods to improve the overall reliability and accuracy of position estimates. Algorithms for estimating near field motion, for identifying known visual features, and for integrating the information from the multiple position information sources must be developed to realize a fully integrated robot navigation system.

16
Submitter:

Institution/Business:
Phone Number:
E-mail Address:
WWW Address:
General Topic:
Topic Title:


J. Thomas McKinnon

Department of Chemical Engineering
Colorado School of Mines
(303) 273-3098
jmckinno@mines.edu

Generic Workbench Architecture
Development of the Generic Workbench Architecture

Description
We present a plan for the development of a flexible, extensible framework for integrating engineering models under a common graphical user interface with a common relational database system. This tool, called the Generic Workbench, will allow engineering models developed under heterogeneous environments to be unified. The Generic Workbench is being written in Java to allow platform independence and is based on an industry-proven lightweight persistence framework object model. Distributed computing will be enabled using the XML/SOAP protocol. The initial demonstration of the viability of this tool will be made by porting it to the OpenChem Workbench. The general nature of the Generic Workbench will enable broad-based technology transfer to other fields.