|
1. Suboptimal multiscale
representation and denoising of contaminated
chaotic signals
2. Sensor-based reconstruction of n-SDE formal models
3. Recurrence based piecewise eigen
representation
|
|
Applications
to Quality Monitoring in Machining and Other Manufacturing Machine
Operations
|
Sponsors: NSF, Zumberge
Fellowship, SC2 Foundation, HP, National Instruments, ONR-MURI, Powell
Foundation, OSU Startup Grant
1. Experimental characterization of machining process
dynamics
2. Sensor-based modeling and control of machining
dynamics and chatter
3. On-line continuous tool wear estimation using
fractal analysis
4. Nonlinear analysis and microdynamic
modeling of acoustic emission
5. Model-based tampering for improved performance of
aerospace precision grinding processes
|
|
|
RFID
and RF Sensor network applications
|
Sponsors: ONR, NSF, OK-DoT
1. Damage monitoring of connected plate structures
2. Nonlinear Intrusion Detection in TCP/IP networks
3. Container Integrity monitoring using RFID Sensors
|
|
|
Applications
to Large Distributed Systems
|
Adaptive Interactive Modeling
Systems Technology (with S.C-Y. Lu))
Sponsors: USC, Ford
1. Backward mapping methodology for design synthesis
2. Ontological extension of axiomatic design
Control of Distributed Transportation
Systems (with M. Dessouky)
Sponsors: METRANS, NSF
1. Coordination of dynamic resource allocation in
trucking operations
2. Multiagent negotiations
for transit system control
|
|
Funded Research Activity
|
|
|
Current
Research Projects
|
1.
SST:
Collaborative Project: Self-sustainable sensor networks for
infrastructural integrity monitoring
PI
(with S.R.T. Kumara-Penn State, S-G. Kim-MIT
and X. Zhang-Berkeley),
NSF,
$774,000 (OSU allocation $198,000), 2004-2007
This collaborative project brings together
the complementary expertise of the investigators in piezoelectric power
generation (MIT), miniature high efficiency RF units (Berkeley), wireless
signal analysis (OSU) and survivability of large sensor networks (Penn State).
The two main objectives of the project are to (a) design and fabricate
self-supporting miniature wireless sensors (currently for vibration and
temperature measurement) capable of harvesting energy from the
host-environment and therefore do not need batteries, and (b) derive
principles to harness information from these sensors for quality and
integrity monitoring of large scale infrastructure systems. These systems
include underground utility pipeline networks. Eight journal papers and 3
conferences papers have been accepted or published, and 10 manuscripts
have been submitted for review based on the research conducted under this
grant. This project has been the basis for 2 MS and 1 PhD theses.
2.
Nonlinear
Continuous Flow Modeling for Real-Time Performance Prediction of
Automotive Assembly Operations
PI
GM,
$76,500, 2005-07
Manufacturing
enterprises are investing in a variety of sensors and IT infrastructure
to increase plant floor systems visibility. This offers an unprecedented
opportunity to track performance of a manufacturing system from a dynamic, as opposed
to a static sense. Conventional
static models are inadequate for predicting performance variables in
real-time from these large data sources. Dynamic models are necessary to
compactly capture information from vast data sources for real-time
performance prediction. Among the relevant
approaches, flow
modeling offers an effective balance between accuracy and speed. This
approach treats the part movement in a manufacturing system as a fluid
flow and models the dynamics in the form of differential equations.
However the presence of finite capacity buffers and machine breakdowns
make the models discontinuous, thereby rendering the simulations slow and
less accurate. We have investigated the use of hybrid continuous-discrete
modeling approach. The results show that the model speeds are comparable
to those of commonly used discrete event simulations, and the model
estimates are with 10% of actual line observations. We are currently in
the process of further enhancing the model speed through the use of sigmoidal function theory and degradation dynamics.
3.
RF
Sensor Application for Container Integrity Monitoring
Co-PI
(with V. Sarangan)
OK-DoT, $60,000, 2005-07 (OSU-IEM allocation $14,500)
Global
supply chain operations use hundreds of thousands of container trucks to
transport valuable goods within and across continents. The vibration
patterns of a container and its contents can reveal significant
information related to its integrity during transport, handling and
storage. The patterns of container vibrations are sensitive to the
following four major factors that define the operating condition of a
container: terrain type, speed of
the vehicle, weight of the container and dimensions of the container. The
primary purpose of this project is to establish quantitative
relationships between the vibration patterns and the operating conditions
of a moving container truck and the stability of contents therein through
a series of experiments involving the use of a model container truck and
a wireless (T-mote) sensor capable of discerning vibrations at 200Hz
sampling rates. The idea is to classify the operating conditions by
analyzing the complex dynamics underlying vibration signals. Using
nonlinear analysis, we find that the the
Lyapunov exponents are in the range of 0.01-0.02 for dynamics underlying
signals from Stage 1 and 0.005 for those from Stage 2, implying that the
dynamics may be nonlinear and chaotic. The statistical and nonlinear
dynamic features together are successfully mapped using a neural network
to classify between the different operating conditions. The neural
networks were able to accurately identify the operating conditions from
the vibration sensor features about 90% of the times. In real life, the
research results can be applied to accurately capture the environmental
condition in which the vehicle is moving and thus proactively address
possible serious integrity losses. One manuscript has been submitted for
review, and the student team that works on this project has won multiple
awards (in the OSU-ECE Design Day and OSU Research Week) for their
efforts. This research has been the basis for one ongoing PhD thesis and
one MS thesis.
4.
Technological
and Economic Analysis of RFID and RF Sensors for Tinker AFB operations
PI
CASI-Tinker
AFB, $50,000, 2006-07.
The
project aims to survey and evaluate alternative Automatic Identification
Technologies (AITs) that complement and/or
integrate with the current WiFi systems for
improving the real-time tracking of assets in Tinker AFB ramp operations.
We are investigating the technological and economic viabilities of
passive RFID and sensor-integrated wireless sensor networks (e.g., Motes)
along with other alternative AITs. This project will lead to the
development of design and operational guidelines of a modified AIT system that fuses information
from multiple AITs (including WiFi) for effective asset tracking.
5.
Heterogeneous
wireless sensor based modeling of chemical mechanical planarization
process
Co-PI
(with R. Komanduri and Z. Hou)
NSF,
$394,000, 2007-10.
The semiconductor industry is rapidly moving
towards ultra-large scale integration (ULSI) of
microelectronic circuits. Planar surfaces are necessary to enhance wafer
yield and reduce stack heights. Chemical mechanical planarization
(CMP) has become the process of choice for this application. Material
removal rate (MRR) and within wafer nonuniformity
(WIWNU), among other performance variables, should be optimized for
improved productivity and reduced costs. Interestingly, PIs’ recent
studies showed that the dominant process parameters for determining MRR are
not so significant for WIWNU and vice versa. Also, the process-machine
interactions play an important role with the CMP process dynamics likely
to be nonlinear and stochastic. Research to-date has not addressed the
physical bases for such behaviors. This is a critical scientific and
technological barrier for effective quality monitoring and control of
CMP. This recently approved project will
yield deeper insight into various chemo-mechanical interactions in CMP
and will integrate a heterogeneous sensor network into CMP operations to
improve productivity and IC quality.
6.
Experimentation
test-bed for evaluation of RFID and RFID sensing technologies
PI
CELDi (NSF/IUCRC), $30,000, 2005
This project was aimed at
creating a laboratory infrastructure to conduct research in RFID systems,
evaluate and benchmark performance of various RFID and high throughput RF
sensor systems. A laboratory to conduct research in Sensor Networks and
Complex Systems Monitoring (COMMSENS) has been established in Fall 2004.
It now houses various components of RFID systems, sensors, data
acquisition systems procured from multiple commercial vendors as well as
a new architecture for high throughput wireless sensing of large scale
infrastructure. This project has also led to an approach for statistical
design of RFID system, a statistical model for predicting performance of
RFID systems in various environments as well as a method to overcome
current technological barriers in reading RFID sensors in highly metallic
environments. This research has been the basis for one MS thesis and two
patent disclosures.
7.
RFID
business case study
PI
FAA,
$42,800, 2005-06
This study had focused on
evaluating the technological and economic feasibility of using RFID and
2-D barcodes for identification and tracking of assets in FAA logistics
center operations. A spreadsheet framework to perform the economic
analysis was developed. One of the research assistants involved in this
project was subsequently hired as a summer intern to follow-up on the
research results and to customize the framework to FAA operations. A
Phase 2 of this project is currently being considered.
|
|
|
Prior
Research Funding
|
1. Co-PI (with B. Khoshnevis) for "Free-form
fabrication of advanced ceramic components using contour
crafting," Office of Naval
Research (Sub-contracted from Rutgers), $105,000, 2000-02
2. PI for “Dynamic modeling of Contour Crafting,”
Powell Foundation Fellowship, USC, $50,000, 2001-
2002
3. Co-PI (with M. Dessouky and R. Leachman) for “Real-time dispatching and routing in
urban rail
networks,” National Science
Foundation (approved for funding), $149,800, 2002-2004
4. Co-PI (with S.C. Lu) for “Adaptive Interactive
Modeling System,” Ford Motor Company, $10,000, 1997-
98
5. PI for “Implementation of nonlinear real-time
chatter control in machining,” Zumberge Grant,
USC,
$24,977, 1998-1999
6. PI for “Academic Grants Program: Development
of instrumentation for the Manufacturing and Controls
Lab,” National Instruments,
received $12,000 worth of instrumentation and software, 2000
7. PI for “Virtual machine tool laboratory to
meet local manufacturing needs,” SC2 Grant, USC, $6,000,
1999-00
8. PI for “HP Foundation Program: Development of
Computer Network for Manufacturing and Controls
Lab,” Received $30,000 worth of
computing equipment through HP Foundation Grant, 1999
9. Co-PI (with S.C. Lu) for "USA-China
Workshop on Advanced Machine Tool Research," NSF, $60,500,
1998-00
10.PI for “Dynamic coordination of resource allocation
in trucking operations," METRANS Center, $20,000,
2000
11.PI (with M.M. Dessouky) for "Multiagent coordination of transit operations," METRANS Center,
$47,000, 2000-01
|
|
