TECHNOLOGY AREAS: Information Systems, Materials/Processes, Human Systems

ACQUISITION PROGRAM: Program Management Test, Measurement and Diagnostic Equipment (PMTMDE).

OBJECTIVE: Develop and leverage existing and emerging efforts to dramatically advance the state of the art in cognition based agent technologies for maintenance across DoD.

DESCRIPTION: Due to the large number of existent and emerging weapons systems platforms (to include all systems that move, shoot, communicate, detect), the challenge to the maintainer base continues to grow. The need to more accurately and rapidly diagnose and prognostically detect, current and incipient systems failure or degradation continues as expeditionary warfare tempo increases and costs rise.

The ability to accurately and rapidly troubleshoot complex systems requires sophisticated analytical ability and experience on the part of the maintainer.
Retaining and maintaining this capability across the Joint Services requires continual effort. With refocusing of manning requirements and a shift to considering methods of increasing efficiencies and operational responsiveness of the maintainer base, there is a need to investigate supplemental/advanced technologies. Multiple efforts have been undertaken over the years within the maintenance community to aid the maintainer community. These efforts have ranged from built in test (BIT), built in test equipment (BITE), to sophisticated test equipment and electronic interactive technical manuals. These also included troubleshooting trees and other early efforts at artificial expert systems to allow the maintainer more structured methods of diagnosis.

Government, industry, and academia have undertaken extensive work over the last three decades in these areas. We seek to leverage these efforts and advance them greatly to provide a superior, truly innovative and novel capability to the field maintainer during the troubleshooting process. An effort to put together an intelligent agent form of technology that accurately simulates and realistically augments human troubleshooting of weapons systems under significantly different scenarios does not yet exist.

The goal is to provide an extraordinary, intelligent troubleshooting agent/technology that appears to the user as a systems expert. The system must be capable of rapid learning and designed to closely interface with the maintainer unlike any previous systems. Enormous strides have been taken within the credit industry, insurance analysis, network security, homeland security, nuclear, and others to set up pseudo-intelligent systems to successfully aid in human oversight of vast amounts of data and activities. Many industries are looking at ways to reduce cost and improve readiness by having systems in place that anticipate and even predict incipient systems failure to their respective maintenance infrastructures. Leveraging this kind of technology to the DoD maintenance infrastructure and driving it to the next level could provide a revolutionary jump to the newly emerging test hardware and complex weapons systems that would thus benefit. Additionally, a system of this type of intelligence-based software maintenance technology should be adaptable to any system within the DoD.

PHASE I: Using a current weapons platform as a test bed, demonstrate the ability to rapidly and accurately troubleshoot operational failure and degradation scenarios under a variety of conditions with maintainers from a low experience skill-set. This innovative system must be capable of understanding it�s environment and rapidly changing conditions. The system must be able to closely, accurately, and directly interact with the user on a level approximately similar to a junior maintainer and a system expert under real world situations. Contrast and compare the differences and capability of this form of highly advanced intelligent agent troubleshooting with all previous methods and constraints with each.

PHASE II: Demonstrate the ability of the system to rapidly assess complex electronic, mechanical, and diverging electromechanical troubleshooting capabilities under random conditions with a variety of scenarios and widely varying maintainer skill sets. Demonstrate the ability to accurately resolve false diagnosis, unknowns, and to predict failure based on observed system behavior, data, and human verbal interaction. Demonstrate active learning within the system and the ability to assess novel (unpredicted) conditions.

PHASE III: Demonstrate the capabilities shown in Phases I and II with the ability to be rapidly incorporated across multiple platforms. Demonstrate additional enhancements that provide a seamless and fluid interface between the system and the maintainer.

Demonstrate how the intelligent system technologies can respond/react to chaotic scenarios and incomplete data while still providing benefit to the maintainer.

PRIVATE SECTOR COMMERCIAL POTENTIAL/DUAL-USE APPLICATIONS: The proposed intelligent system technology could have broad impact across the commercial maintenance community.

REFERENCES:
1. An Ecological Framework for Cancer Communication: Implications for Research
Kevin Patrick1, MD, MS; Stephen S Intille2, PhD; Marion F Zabinski1, PhD, MPH
1Department of Family and Preventive Medicine, University of California, San Diego, La Jolla, California, USA
2House_n, Massachusetts Institute of Technology, Cambridge, Massachusetts, USA
Corresponding Author:
Kevin Patrick, MD, MS
Department of Family and Preventive Medicine
University of California, San Diego
9500 Gilman Drive MC 0811
La Jolla, California 92093-0811
USA
Phone: +1 858 457 7296
Fax: +1 858 622 1463
Email: kpatrick [at] ucsd.edu

2. Technical Design of Condition Based Maintenance System
-A Case Study using Sound Analysis and Case-Based Reasoning
Marcus Bengtsson1, Erik Olsson2, Peter Funk2, Mats Jackson1
1Department of Innovation, Design and Product Development
2Department of Computer Science and Engineering
Mälardalen University
Box 325, SE-631 05 Eskilstuna, Sweden
marcus.bengtsson; erik.olsson; peter.funk; [email protected]
1phone +46 16 153 486, fax +46 16 153 610

KEYWORDS: Intelligent agents;artificial intelligence;expert systems; direct visualization

TPOC: Mike Shellem
Phone: (229)639-7692
Fax: (229)639-6172
Email: [email protected]
2nd TPOC: Mike Heilman
Phone: (703)432-3240
Fax:
Email: [email protected]

** TOPIC AUTHOR (TPOC) **

DoD Notice:   Between May 1, 2006 and June 13, 2006, you may talk directly with the Topic Author(s) to ask technical questions about the topics. Their contact information is listed above. For reasons of competitive fairness, direct communication between proposers and topic authors is not allowed starting June 14, 2006 , when DoD begins accepting proposals for this solicitation.

However, proposers may still submit written questions about solicitation topics through the SBIR/STTR Interactive Topic Information System (SITIS), in which the questioner and respondent remain anonymous and all questions and answers are posted electronically for general viewing until the solicitation closes. All proposers are advised to monitor SITIS (06.2 Q&A) during the solicitation period for questions and answers, and other significant information, relevant to the SBIR 06.2 topic under which they are proposing.

If you have general questions about DoD SBIR program, please contact the DoD SBIR Help Desk at (866) 724-7457 or email weblink.