NASA STTR 2007 Solicitation

FORM B - PROPOSAL SUMMARY


PROPOSAL NUMBER: 07-1 T1.01-9880
RESEARCH SUBTOPIC TITLE: Information Technologies for System Health Management, Autonomy, and Scientific Exploration
PROPOSAL TITLE: Data Reduction Techniques for Real-time Fault Detection and Diagnosis, and Multiple Fault Inference with Imperfect Tests

SMALL BUSINESS CONCERN (SBC): RESEARCH INSTITUTION (RI):
NAME: Qualtech Systems, Inc. NAME: University of Connecticut
STREET: 100 Great Meadow Road, Suite 603 STREET: 438 Whitney Road Ext., Unit 1133
CITY: Wethersfield CITY: Storrs
STATE/ZIP: CT  06109 - 2355 STATE/ZIP: CT  06269 - 1133
PHONE: (860) 257-8014 PHONE: (860) 486-3994

PRINCIPAL INVESTIGATOR/PROJECT MANAGER (Name, E-mail, Mail Address, City/State/Zip, Phone)
Sudipto Ghoshal
sudipto@teamqsi.com
100 Great Meadow Road, Suite 603
Wethersfield, CT 06109 - 2355
(860) 257-8014

Expected Technology Readiness Level (TRL) upon completion of contract: 4 to 5

TECHNICAL ABSTRACT (Limit 2000 characters, approximately 200 words)
The recent advances in data collection and storage capabilities have led to information overload in many applications, including on-line monitoring of spacecraft operations with time series data. Such datasets present new challenges in data analysis, especially for implementation in memory-constrained DECUs. Also, the traditional statistical methods break down partly because of the increase in the number of observations (measurements), but mostly due to an increase in the number of variables associated with each observation ("dimension of the data"). One of the problems with high-dimensional datasets is that not all the measured variables are "important" for understanding the underlying phenomena of interest. In addition to the computational cost, irrelevant features may also cause a reduction in the accuracy of some algorithms. The first key issue we propose to address is that of data reduction techniques for onboard implementation of data-driven classification techniques in memory-constrained onboard processing units. Some of the classification techniques we intend to use with the above data-reduction techniques include, support vector machine (SVM), probabilistic neural network (PNN), k-nearest neighbor (KNN), principal component Bayesian analysis (PCA). To improve the diagnostic accuracy and efficiency of the above classifiers, we will apply classifier fusion techniques such as AdaBoost, Error correcting output codes, Voting to find which architecture will enhance the accuracy and under what conditions. Finally we will investigate Dynamic Multiple Fault Diagnosis that can work with imperfect fault/anomaly detection tests. As part of this task, we will develop novel factorial hidden Markov model-based inferencing techniques such as Lagrangian relaxation and Viterbi decoding algorithms to solve this difficult combinatorial optimization problem, for on-board vehicle health monitoring and fault diagnosis.

POTENTIAL NASA COMMERCIAL APPLICATIONS (Limit 1500 characters, approximately 150 words)
Any well executed on-board health monitoring program that can diagnose system health and can automatically reconfigure to respond to failures has tremendous use and importance in space aviation. The proposed work is in line with NASA's IVHM goals, as well as mission and contingency planning. NASA has a stated requirement for automated solutions for system health management, where prognostics and system recovery are required to support space initiatives. This proposed tasks addresses a small but important issue in achieveing that goal.

POTENTIAL NON-NASA COMMERCIAL APPLICATIONS (Limit 1500 characters, approximately 150 words)
The industries interested in the developed technology are expected to include the manufacturers (OEMs) and end users of complex systems and equipment that are used in environments where failure has serious consequences and where high availability and operational reliability are required. The aviation industry, primarily aircraft manufacturers and their customers are industry segments that are of interest and will be targeted as part of the commercialization effort.

NASA's technology taxonomy has been developed by the SBIR-STTR program to disseminate awareness of proposed and awarded R/R&D in the agency. It is a listing of over 100 technologies, sorted into broad categories, of interest to NASA.

TECHNOLOGY TAXONOMY MAPPING
Autonomous Reasoning/Artificial Intelligence
Expert Systems
Software Tools for Distributed Analysis and Simulation


Form Generated on 09-18-07 17:52