Dr. Galina Rogova
The State University of New York at Buffalo, USA
Dr. Rogova is a research professor at the State University of New York at Buffalo. She is recognized internationally as an expert in information fusion, machine learning, decision making under uncertainty, and information quality. She has worked on a wide range of defense and non-defense problems such as situation and threat assessment, understanding of volcanic eruption patterns, computer-aided diagnosis, and intelligent transportation systems, among others. Her research was funded by multiple government agencies as well as commercial companies. She published numerous papers and co-edited 7 books with one of them devoted to Higher Level Fusion. She served as a program committee member, special session organizer, and tutorial lecturer for numerous International Conferences on Information Fusion. Dr. Rogova was also a member of the organizing committee for multiple NATO Advanced Study Institutes and NATO Advanced Research Workshops on information fusion and decision support.
Higher level fusion and situation management: challenges and computational approaches
Situation management is a collection of methods and tools aimed at helping decision makers to monitor, understand, and control dynamic situations, and act effectively to mitigate their impact. Situation management problems exist in many domains such as traditional and asymmetric warfare, man-made and natural disasters, network management, border and harbor protection, and transportation. One of the critical enabling technologies of situation management is higher level fusion transforming multiple data and information streams into actionable knowledge to provide for user situation awareness that is essential for effective decision making and actions. Higher level fusion processes are designed by analyzing spatial and temporal relations of the situational entities considered at different levels of granularity and their dynamics within the overall situational context. The problem of building such processes is complicated by the dynamic environment, erroneous or poor quality context characteristics, variable quality of heterogeneous information obtained from both human and automatic processes as well as incomplete and uncertain domain knowledge. The talk will provide a discussion of major challenges of designing higher level processes and suggest some approaches to confront these challenges.
Mr. Dale Reding
Director General Science and Technology Air Force and Navy (DGSTAN)
Department of National Defence, Canada
Mr. Reding holds the position of Director General Science and Technology Air Force and Navy (DGSTAN) in the Department of National Defence (CANADA). Reporting directly to the Assistant Deputy Minister, Science & Technology (S&T), he is responsible for the strategic direction of S&T for both the Air Force and Navy. In this role Mr. Reding is the Scientific Advisor to the Commanders of the Royal Canadian Air Force (RCAF) and the Royal Canadian Navy (RCN). Mr. Reding has enjoyed a 27 year career as a Defence Scientist and manager within the Canadian Department of National Defence, carrying out operational research and analysis across a variety of operational environments and commands, as well as conducting research in the areas of machine learning, optimization, logistics analysis and defence planning. His career includes assignments as a senior scientist at the Center for Aerospace Analysis at NORAD/USSPACECOM (North American Defence Command/United States Space Command) in Colorado Springs (USA), and as a principal scientist at the NATO C3 Agency in The Hague (NL). Mr. Reding was the Chief Scientist at the Centre for Operational Research and Analysis (2008-2009), and Director General Defence R&D Canada – Toronto, Canada’s defence ‘Center of Excellence’ for Human Performance research (2010-2013). Mr. Reding was appointed Director General Science and Technology Air Force and Navy in April 2013.
Post-SSE Fusion and Sensing Challenges (Air)
Dr. Chee-Yee Chong
Chee Chong received his S.B. S.M. and Ph.D. degrees, all in Electrical Engineering from the Massachusetts Institute of Technology. He was on the electrical engineering faculty of Georgia Institute of Technology until 1980, when he gave up the security of a tenured position to join a Silicon Valley startup that performed research on advanced information and decision systems. He served as fusion technology research lead at this startup and three other companies until he retired in 2013 to perform independent research in areas that interest him. He has been involved in research and development in tracking, fusion, and resource management for undersea, surface, ground, air, and space targets for over 25 years. In particular, he is known for his work in distributed tracking and fusion. He developed the first fusion rule for optimally combining probabilities and state estimates, and led the development of the first distributed multiple hypothesis tracking algorithms under DARPA’s Distributed Sensor Networks (DSN) program in early 1980’s. He has been involved in all levels of information fusion, including Bayesian multiple hypothesis tracking for general target and sensor models, optimal track fusion and association, close loop tracking with sensor resource management, analytic model to predict association performance, Bayesian networks for situation assessment, hard and soft data fusion, and graph approaches for data association. He is the co-author of over one hundred conference papers, journal papers and book chapters, and the co-inventor of three U.S. patents, including one on information fusion for cyber security. He is co-editor of the book “Distributed Data Fusion for Network-Centric Operations”, which has been translated into Chinese. He co-founded the International Society of Information Fusion (ISIF), and served as its President in 2004. He was general co-chair for the 12th International Conference on Information Fusion held in Seattle, USA in 2009. He was associate editor for IEEE Transactions on Automatic Control, associate editor for Information Fusion, and is associate editor for Journal of Advances in Information Fusion published by ISIF. He received the ISIF Yaakov Bar-Shalom Award for a Lifetime of Excellence in Information Fusion in 2016.
Forty years of distributed filtering
Filtering or estimating the state of a dynamic system is a core function for target tracking, providing the necessary inputs for higher level fusion. In multi-sensor tracking and fusion systems, distributed filtering requires less communication and provides more robustness than centralized filtering. Decentralized filtering algorithms first appeared about forty years ago. Since then, many distributed filters have been developed with different assumptions. This paper reviews popular distributed filtering approaches including information decorrelation to reconstruct the global estimate, covariance based algorithms to find the best linear estimate, and consensus/diffusion filters to obtain stable solutions. Suggestions for future research will be discussed.