Software Panel

Home
Keynote Speakers
Tutorials
Future Directions Panel
Standards Workshop
Software Panel
Affordability/ROI Panel
NDE Panel
Research Panel
Design Techniques

 

PANEL:
Bridging Between the Software Engineering and the PHM Engineering Communities

 

Moderator:

 Jeffrey Voas, US National Institute of Standards and Technology (NIST)

 

Dr. Jeffrey Voas is currently a computer scientist at the US National Institute of Standards and Technology (NIST). Before joining NIST, Jeff was an entrepreneur and co-founded Cigital.  After 13 years at Cigital, Jeff accepted a director position at Science Applications International Corporation.  He has served as the IEEE Reliability Society President (2003-2005, 2009-2010), and served as the IEEE Computer Society’s Second VP (2010).  Dr. Voas is an IEEE Director (2011-2012) and IEEE Fellow.

 

Determining Software Trustworthiness in an Environmental Context

Environments are messy.  Nonetheless, if one can effectively define the software environments at play in federated systems of systems, families of systems or any other set of interactive systems, their combined impact may be evaluated.  In essence, their ultimate fusion to an intersection of requisite functions to satisfy a combined and well-understood operational environment would serve to increase resulting trustworthiness.  The only missing variable is how to define the combined operating environment as an intersection of the various dynamic environments comprising the constituent software components.  PHM can play a role here by applying its vast system sensing and monitoring algorithms, however they have yet to applied to external software environments due to the intractability of the size of most networked environments, particularly wireless ones.

**********

 Panelists:

 Prof. Dimitry Gorinevsky, Professor, Stanford University

 Dimitry Gorinevsky is a principal of Mitek Analytics LLC consultancy in Palo Alto, CA, and a Consulting Professor of Electrical Engineering in Information Systems Laboratory at Stanford University.  He worked in industry and academia.  His interests are in control, decision, and monitoring systems applications across many industries.  He has authored a book, 150+ papers, and a dozen patents.  He is a recipient of Control Systems Technology Award of the IEEE Control Systems Society, several other awards, and a Fellow of IEEE.

 

Monitoring Algorithms in PHM

 This talk will focus on off-line monitoring.  It will discuss algorithms for PHM monitoring in the statistical signal processing perspective.  Anomaly detection will be discussed as a Statistical Process Control (SPC) problem.  Training the nominal model on the historical data will be discussed as a regression problem.  The talk will introduce a fleet monitoring problem as a two-level fixed effects regression.  It will discuss an optimal approach to anomaly detection in the aircraft fleet that can be viewed as a generalization of Multivariate SPC.

 

**********

 

Prof. Keith Miller, Professor, University of Illinois at Springfield

 Keith W. Miller is the Schewe Professor of Liberal Arts in the Computer Science Department of the University of Illinois Springfield.  His research areas are software testing and computer ethics.  Keith is the editor-in-chief of IEEE Technology and Society, and writes an ethics column for IEEE IT Professional.  Dr. Miller received the 2011 Joseph Weizenbaum Award from the International Society for Ethics and Information Technology, and is the principal investigator on a recent National Science Foundation grant to study the effect of ethics education in programming classes.

 

PHM algorithms that interpret data are value laden

 As PHM algorithms grow in complexity and ambition, people are increasingly relying on software to interpret large volumes of data.  But transforming data into meaning, particularly when the data are not exclusively quantitative, requires subjective decisions on the part of designers, programmers, and operators of PHM systems.  Some of these decisions will be based on grounds that are considered purely technical.  Other decisions will require judgments about what system behaviors are most valuable, and what misbehaviors are most dangerous, for the users of a system, and for the people who aren’t users but are still affected by that system.  In this talk, we will discuss two short cases that involve PHM applications, and we will consider ethical considerations that are important when selecting appropriate algorithms and reasoning approaches to those applications.

 **********

 

Prof. Jia Zhang, Associate Professor, Northern Illinois University

 Dr. Jia Zhang is currently an associate professor of the Department of Computer Science at Northern Illinois University.  Her research interest centers on services computing, with a focus on collaborative workflows, service-oriented architecture, semantic services and cloud computing.  She has published over 100 refereed journal papers, book chapters, and conference papers, as well as a co-authored textbook titled Services Computing.  She is an Associate Editor of IEEE Transactions on Services Computing (TSC) and of International Journal of Web Services Research (JWSR).

 

The Cloud and PHM Data Volume

 Consumer computing may become more like running water and less like a battleship in our mental maps.  Cloud computing offers the opportunity to dissect and interpret the huge volumes of data that PHM employs from its many sensors.  This talk will pinpoint the relationship between the SaaS cloud services model and PHM.

 

**********

 

Dr. Chris Michael, Computer Scientist, SAIC

 Dr. Chris Michael has been engaged in security analysis and penetration testing of government and commercial software systems for more than 13 years and has initiated and led a variety of government-funded research projects.  He is currently the principal investigator of SAIC's IARPA-funded SPICE project, which focuses static vulnerability analysis and runtime protection for binary executable software. His research includes machine learning, software security, and algorithms for automated reverse engineering and software analysis.  He holds a Ph.D. in computer science (machine learning) from the College of William and Mary.

 

Health Monitoring for Software Security

 This talk discusses health monitoring of security properties, that is, software properties whose violation might indicate that a vulnerability in the software is being attacked. Because malicious attackers may try to hide their presence from a health monitor, such monitoring has to be fine-grained. We will argue that individual instructions in the binary executable code should be monitored. This can be prohibitively expensive, but it may be possible to use static verification technology to reduce the workload of the health monitor. The talk discusses the possibilities of such a synthesis between static verification and runtime supervision.

 

**********

 

Hart Rossman, VP and CTO for Cyber Security Services and Solutions, SAIC

 Hart Rossman is Vice President and Chief Technology Officer for Cyber Security Services & Solutions at SAIC.  He is a Senior Research Fellow with the Supply Chain Management Center within the RH Smith School of Business at the University of Maryland in the area of Cyber Supply Chain Assurance.  Mr. Rossman has been named to the state of Maryland Governor’s Workforce Investment Board’s Cyber Security Workforce Steering Committee as well as ISC2’s Application Security Advisory Board and is a named contributor to the CWE/SANS Top 25 Most Dangerous Software Errors <http://www.sans.org/top25-software-errors/>.  Mr. Rossman is a faculty member with the Institute for Applied Network Security, represents SAIC’s Incident Response Team in FIRST, and is a founding member of the Corporate Executive Programme.  Mr. Rossman co-authored NIST SP 800-64rev2, “Security Considerations in the System Development Life Cycle.”  He has earned a CISSP, CSSLP, received his B.A. in Communication from the University Of Maryland, College Park, and received his MBA from the University of Maryland, Robert H. Smith School of Business.

 

Defense in Breadth:  Software Assurance through Supply Chain Security

 It’s a national security imperative in a global economy that we have confidence in the supply chains of integrated systems and the integrity of the people, processes and technology that comprise them.  Traditional assurance methods focus on point in time quality and integrity assessments, generally on the finished product; post acquisition but prior to deployment.  This approach is no longer sufficient. A modern view of the cyber supply chain forces us to expand this view to consider the inter-relationships between system (product) development lifecycles across the supply chain; from the very first supplier to the true end user of the system.  This talk with introduce a cyber supply chain assurance reference model and discuss opportunities for instrumentation, analytics, and decision support in an effort to create more secure, reliable supply chains that can transfer these properties to the products they create and deliver.  

 

**********

 

Dr. Sam Keene, FIEEE, Consultant

 Dr. Sam Keene is a Recognized International Resource in the R &QA field.  He is a Past President of the IEEE Reliability Society and received the 1996 “Reliability Engineer of the Year” Award.  He received the IBM Outstanding Contribution award for his multimillion dollar savings through failure analysis activities.  The ASQ presented Sam the Allan Chop award in 1999.  He has also been recognized with the Outstanding Education Award from the Reliability Society.  He holds the office of Fellow of the IEEE, for his technical accomplishments, and is the recipient of the IEEE 2000 Medallion Award.  Dr Keene is also a Six Sigma Master Black belt, headed up the world wide MBB Council at Seagate Corp.  He also was one of 13 MBB’s that developed the Black Belt test Body of Knowledge for ASQ Black Belt Cerification.

 

PHM Augment’s Software Reliability

 Dr. Keene will introduce a new model that will set the expectation for how the software should normally perform.  This model, called the Development Process Prediction Model, was developed by Dr Keene, and can provide the software reliability prediction before software is developed.  Later, when the software goes through reviews and inspections, that data can be used to refine our estimate of the reliability of the delivered software.  Finally the beneficial PHM impact of software rejuvenation policy on software reliability will be shown.

 

Home | Keynote Speakers | Tutorials | Future Directions Panel | Standards Workshop | Software Panel | Affordability/ROI Panel | NDE Panel | Research Panel | Design Techniques

This site was last updated 06/06/11

Technical co-sponsors:
baseh
  logo