Track 5: Systems
Engineering: Abstracts and Biographies
Track 5, Session 1: 8:30-9:30am
Engineering of Complex
Systems: A Focus on Advanced Agent-Based Systems
by
Dr. Suzanne Barber,
UT Austin
Abstract: Automation is an effective mechanism for managing
complex and dynamic environments. Thus, end-users are investigating new
methodologies, technologies, and architectures for leveraging the benefits
of automated systems. The Laboratory for Intelligent Processes and Systems
is conducting research in two focus areas currently receiving critical
attention: formal engineering methodologies and distributed agent-based
systems.
Requirements posed by domain applications (e.g. manufacturing planning
and scheduling, combat system command and control, training) demand more
intelligent systems, re-usable system components, extendible system components
and decreased maintenance. This research program gives critical attention
to formal analysis and design issues highlighting the contributions of
systems theory, software engineering, and the integration of engineering
disciplines. Regarding the development of these component-based systems,
distributed autonomous agents (i.e. components) can provide more responsive
intelligent automated systems. Mechanisms to promote the flexible interaction
and execution of these agent-based systems is key.
The Laboratory organization consists of two programs addressing research
essential to successful development and deployment of modular, reusable,
and flexible systems:
Formal Engineering Processes to Support System Life Cycle Approaches. This
program investigates methodologies, modeling techniques, and tools for
formal requirements modeling, component-based system design, and evaluation
of system designs using metric constraints imposed by requirements and
specification models. Example tools include the Requirements Integration
and Verification Tool (RIVT) and Hybrid Domain Representation Architecture
(HyDRA). RIVT employs a formal representation (developed within the Laboratory)
of system components which describes declarative knowledge (attributes
characterizing the component), behavioral knowledge (component states,
events), and integration constraints imposed by the domain application
but independent of implementation. Users are allowed to retrieve components
by posing requirements-based queries based on: 1) targeted constraints
for individual components or 2) integration constraints aimed at identifying
the optimal configuration of system components. HyDRA allows for the modeling
and management of heterogeneous domain models representing various views
(information, behavior, decisions, task sequence) employed to define domain
requirements.
Autonomous Sensible Agents. The practical deployment of distributed agent-based
systems mandates that each agent behave sensibly, incorporating an understanding
of both global system goals and their own local goals. A critical consideration
for this behavior is the agent's level of autonomy. The term level of autonomy
refers to the types of roles an agent plays in its interactions with other
agents. Specifically, this research seeks to prove the following hypothesis:
The operational level of agent autonomy is key to an agent=92s ability
to respond to dynamic situational context, (i.e. the states, events, and
goals that exist in a multi-agent system), conflicting goals, and constraints
on behavior. Levels of autonomy are defined along a spectrum ranging from
command-driven (agent executes commands from another agent), to consensus
(agents work together to meet goals), to locally autonomous (agent can
initiate its own thread of execution), to master (agent controls other
agents). These descriptive autonomy levels are tied to the responsibility
an agent assumes when planning to solve its goals. Thus, the program addresses
both 1) controlling the autonomy of sensible agents and 2) distribution
of intelligence across agents comprising a system.
Acting as a test bed, the Virtual Decision Environment (VDE) project aims
to simulate the execution and propagation of decisions. Visualizations
driven by actual systems component models and other agent-based planning
software provide a "what-if" environment for users to gain insight
regarding how agent planning decisions (executed in the simulation) and/or
and systems design decisions (made during the design effort) are related
and impact productivity, efficiency, and quality.
Biography: Dr. Suzanne Barber is an Assistant Professor in
the Electrical and Computer Engineering Department at The University of
Texas at Austin. She began her career at The Robotics Institute at Carnegie
Mellon University as a Research Scientist working on symbolic languages
and user interfaces for automated robotic assembly systems. She later joined
The University of Texas Automation and Robotics Research Institute (ARRI)
where she began her research in the area of knowledge-based representations
and planning systems for CAD/CAM integration. She received her B.S. in
Engineering Science at Trinity University and her Ph.D. in Electrical Engineering
from The University of Texas at Arlington in 1992.
Dr. Barber is currently the Director of The Laboratory for Intelligent
Processes and Systems where on-going research projects address 1) formal
software engineering approaches, modeling techniques, and tools, and 2)
distributed, knowledge-based planning and control. Her research focuses
on the development of distributed, autonomous agent-based systems. Dr.
Barber has taught Robotics & Automation, Control Theory and developed
two courses: Manufacturing Systems Automation and Domain Specific System
Architectures. She has also participated in a number of federal programs
including the USAF Next Generation Controller program to develop a Specification
of an Open System Architecture Standard (SOSAS) permitting interchangeable
and interoperable system components for real-time machine and workstation
controllers. As a researcher and technical reviewer, Suzanne has collaborated
with the National Center for Manufacturing Sciences (NCMS), National Institute
of Standards and Technology (NIST), National Science Foundation (NSF),
and Department of Defense (DoD) research programs. Industrial partners
have included TRW, IBM, Eastman Kodak, Texas Instruments, Apple, SAIC,
Lockheed Martin, SEMATECH, and General Motors.
Return to TechCon Schedule
Track
5, Session 2: 9:45-10:45am
Neural Networks and
Fuzzy Logic Systems for Control of Industrial Motion Systems
by
Dr. Frank L. Lewis,
P.E., Automation and Robotics Research Institute
Abstract: Several approaches are given to design "Intelligent
Controllers" for industrial motion systems using neural networks (NN)
and fuzzy logic (FL) systems. Design techniques are given that are easy
to use and give guaranteed and repeatable results. Tuning algorithms are
given for the NN weights and/or the FL membership functions that do not
require any off-line training phase, guarantee stability, and are performed
on-line in real-time so that the intelligent controller learns the unknown
system dynamics. If full state feedback measurements are available, the
NN controller consists of a static feedforward net in the feedback loop
plus an outer PD tracking loop. If full state measurements are not available,
then an additional dynamic NN is required to estimate the unmeasured states.
Applications are to actuators with deadzones, backlash, and friction, and
to systems with vibratory modes and high-frequency actuator dynamics, including
robotic systems, DoD systems, and commercial systems.
Biography: Dr. Lewis was born in Wurzburg, Germany, subsequently
studying in Chile and Scotland. He obtained the Bachelor's Degree in Physics/Electrical
Engineering and the Master's of Electrical Engineering Degree at Rice University
in 1971. He spent six years in the U.S. Navy, serving as Navigator aboard
the frigate USS Trippe (FF-1075), and Executive Officer and Acting Commanding
Officer aboard USS Salinan (ATF- 161). In 1977 he received the Master's
of Science in Aeronautical Engineering from the University of West Florida.
In 1981 he obtained the Ph.D. degree at The Georgia Institute of Technology
in Atlanta, where he was employed from 1981 to 1990 and is currently an
Adjunct Professor. He was awarded the Moncrief-O'Donnell Endowed Chair
in 1990 at the Automation and Robotics Research Institute of The University
of Texas at Arlington.
Dr. Lewis has studied the geometric properties of the Riccati equation
and implicit systems; his current interests include robotics, intelligent
control, neural and fuzzy systems, nonlinear systems, and manufacturing
process control. He is the author/co-author of 109 journal papers, 190
refereed conference papers, five books: Optimal Control, Optimal Estimation,
Applied Optimal Control and Estimation, Aircraft Control and Simulation,
Control of Robot Manipulators, and the IEEE reprint volume Robot Control.
Dr. Lewis is a registered Professional Engineer in the State of Texas and
serves on the Editorial Board of International Journal of Control, Neural
Computing and Applications, and Int. J. Intelligent Control Systems. He
is the recipient of an NSF Research Initiation Grant and has been continuously
funded by NSF since 1982. He received a Fulbright Research Award, the American
Society of Engineering Education F.E. Terman Award, three Sigma Xi Research
Awards, the UTA Halliburton Engineering Research Award, the UTA University-Wide
Distinguished Research Award, the ARRI Patent Award, and the IEEE Control
Systems Society Best Chapter Award (as Founding Chairman). He was selected
as Engineer of the year in 1994 by the Ft. Worth IEEE Section and is a
Fellow of the IEEE. He was appointed to the NAE Committee on Space Station
in 1995 and to the IEEE Control Systems Society Board of Governors in 1996.
Return to TechCon Schedule
Track
5, Session 3: 11:00-12:00pm
Hardware/Software
Codesign of Embedded Systems
by
Dr. Margarida F. Jacome,
UT Austin
Abstract: Embedded system architectures comprising of
software programmable components (e.g., DSP, ASIP. and microcontroller
cores) and dedicated hardware processing modules, integrated into a single
cost-effective VLSI chip, are emerging as a key solution to today's microelectronics
design problems. This trend towards heterogeneous (H/S) architectures is
being driven by new emerging applications in the areas of wireless communication,
multimedia, broad-band networks, and industrial and automotive control.
While design methods and tools exist for designing software and hardware
separately, only recently the codesign of mix hardware/software solutions
started being addressed. This talk provides an overview of the state-of-the-art
in the discipline of hardware/software codesign. Particular emphasis is
given to the initial codesign phases, including architecture selection,
performance and power estimation, and hardware/software partitioning.
Biography: MARGARIDA F. JACOME (S'92-M'94) received the
B.S. and the M.S. degrees from the Technical University of Lisbon, in 1981
and 1988, respectively, and the Ph.D. degree in electrical and computer
engineering from Carnegie Mellon University, in 1993. She is an Assistant
Professor in the Department of Electrical and Computer Engineering at the
University of Texas at Austin. Her research interests include system-level
design and H/S codesign and design reuse.
Dr. Jacome received the National Science Foundation Career Award in
1996. Her seminal work defining the set of advanced planning and management
services needed for supporting complex design processes was recognized
with the 1992 ACM/IEEE Design Automation Conference (DAC) Best Paper Award.
She was the guest editor of a 1996 special issue (on design process management)
published by the journal "Computers in Industry." She has been
a member of the Electronic Design Processes Sub-Committee (EDPS) of the
IEEE Computer Society Design Automation Technical Committee (DACT) since
1992. She was Co-chairperson of the 1995 ACM/IEEE Workshop on Electronic
Design Processes, organized by the EDPS. She served as member of Technical
Program Committee of the ACM/IEEE International Conference on CAD (ICCAD)
from 1993 to 1995.
Return to TechCon Schedule
Track 5, Session 4: 1:30-2:30pm
Systems Performance
Modeling and Measurement
by
Dr. George Kondraske,
Human Peformance Institute, UTA
Abstract: Many contemporary system designs result from
the integration of multiple sophisticated modular subsystems. Procedures
used in systems integration design decision-making of such systems are
frequently rather non-quantitative compared to more quantitative and systematic
methods employed elsewhere in engineering. Experimental findings associated
with verification and validation (V&V) are often applicable only to
a specific situation.
A common theme that is inextricably linked to such engineering tasks is
that of performance. In fact, the concept of "performance" pervades
nearly all aspects of life and is associated with all types of living and
artificial systems. Yet, it is asserted that it is not well understood
theoretically and techniques for its modeling and measurement in all fields
have been ad hoc at best. Although a considerable body of material known
as general systems theory exists, the concept of performance has not been
incorporated in it nor has performance been addressed in a general sense
elsewhere. Most knowledge that does exist about performance and its quantitative
treatment has evolved within specific applications, where generalizations
can easily be elusive. Despite the unavoidable and growing relevance, formalized
treatment of systems performance in educational settings is virtually nonexistent.
Performance is multi-faceted, pertaining to how well a given system executes
an intended function and the various factors that contribute to this. It
differs from "behavior" in that "the best of something"
is implied.
In this presentation, motivations leading to the development of a General
Systems Performance Theory (GSPT) are summarized, followed by a brief introduction
to key aspects of the theory. GSPT provides a comprehensive modeling/measurement
strategy applicable to complex systems including both human and artificial
components. Example applications, drawn from a wide range of system types,
will be presented to illustrate basic concepts and demonstrate the types
of benefits possible when they are employed.
Biography: George V. Kondraske received a doctorate in
biomedical engineering from the University of Texas at Arlington/University
of Texas Health Science Center at Dallas (Joint program) in 1982 and a
BSEE from the University of Rochester in 1978. He is currently Professor
of Electrical Engineering and founding director of the Human Performance
Institute at UT Arlington (1986) and is widely recognized for work in the
modeling and measurement of system performance, with particular emphasis
on human performance.
The author of more than 100 papers and book chapters, he and his research
team have developed a modular human performance measurement system that
was subsequently transferred to the private sector (Human Performance Measurement,
Inc.). These instruments and methods are now used in nine countries. In
1986, he introduced General Systems Performance Theory (GSPT) and a unifying
conceptual model (the elemental resource model or ERM) to explain the interface
of the human system to tasks. He has extended the applications of GSPT
to other systems such as artificial vision, telerobotic systems, and most
recently to information technology-based training systems.
Return to TechCon Schedule
Track
5, Session 5: 2:45-3:45am
Safety and Computers
by
Dr. Nancy Leveson, U.
of Washington
Abstract: Computers are being introduced into the control
of virtually every dangerous system --- defense, transportation, aerospace,
medical, chemical and nuclear. Few engineering techniques exist to provide
assurance that safety is not degraded by the introduction of computer control.
At the same time, nothing is absolutely safe, and computers provide important
advantages over the human operators, social systems, and engineered devices
that they are replacing. This talk will describe the state-of-the-art in
assuring safety of computer-controlled systems along with the important
open questions that still need to be resolved. These questions include
whether we are putting too much trust in new technology and what directions
appear most promising for providing greater assurance.
Biography: Dr. Nancy Leveson is Boeing Professor of Computer
Science and Engineering at the University of Washington. She is a Fellow
of the ACM, an elected member of the Board of Directors of the Computing
Research Association and also the Board of Directors of the International
Council on Systems Engineering, a member of the ACM Committee on Computers
and Public Policy, and a member of the National Research Council Commission
on Engineering and Technical Systems. She received the 1995 AIAA Information
Systems Award for "developing the field of software safety and for
promoting responsible software and system engineering practices where life
and property are at stake." She is author of a new book, "Safeware:
Systems Safety and Computers," published by Addison-Wesley.
Return to TechCon Schedule
Track 5, Session 6: 4:00-5:00am
Digital Image Processing
and Photogrammetry in Flight Simulation
by
Dr. Venkat Devarajan,
UTA
Abstract: Several different aspects of digital image
processing and >photogrammetric technology are used in the realization
of a high performance visual system for flight simulation. In this tutorial,
we will provide an overview of the elements of a visual system. Each of
these elements will be discussed in detail and the use of digital image
processing and photogrammetric techniques will be presented. A video tape
presentation will demonstrate the various elements discussed in the tutorial.
A related area called virtual prototyping will also be briefly mentioned.
Biography: Dr. Venkat Devarajan was the chief architect
of TOPSCENE, presently US Navy's primary mission rehearsal system. TOPSCENE
was the earliest system to use real world photo texture in a real-time
rendering system. He also developed an elaborate data base generation system
that implemented for the first time, digital photogrametric techniques
to create country-sized texture data bases. Dr. Devarajan has been a faculty
memeber at UTA since 1990 and his research interests are in collaborative
virtual prototyping and computer vision.
Return to TechCon Schedule