Dr. Carol Smidts
POSTPONED - NEW DATE TBD
Abstract:
Digital systems have been increasingly used in nuclear power
plants. This improves the efficiency of nuclear energy but exposes nuclear
power plants to the threats of cyber-attacks, in addition to the problems
caused by both software and hardware component failures. To improve
the cybersecurity posture of nuclear energy, it is important to develop
effective methods for cyberattack detection, cyberattack response, and
cybersecurity risk assessment. In particular, it is important to distinguish a
cyber-attack from a safety event caused by component failure, so that correct
remedial actions can be taken. Dynamic Bayesian Networks (DBNs) provide a
framework to probabilistically infer the state of the system and differentiate
the nature of abnormal events using both physical sensor measurements and
network data. DBNs also provide interpretability in the inference
process. Compared with regular events, for example, those caused by pure
hardware failures, the presence of malicious attackers in cyberattacks
significantly increases the complexity of incident response. This is because the
evolution and consequence of a cybersecurity event depend not only on actions
taken by plant operators but also on actions taken by the attackers. Game
theory provides a suitable technique to study the interaction among multiple
players. In this seminar, the speaker will discuss a DBNs based abnormal event
classifier and how game theory can be applied to nuclear power plant
cyberattack response modeling and real-time cybersecurity risk update.
Biography:
Dr. Carol Smidts is a Full Professor at The Ohio
State University, Department of Mechanical and Aerospace Engineering, Nuclear
Engineering Program. Professor Smidts’ research lies risk and reliability
analysis and in human factors-instrumentation and control, including human
reliability analysis, probabilistic analysis of dynamics for complex systems,
reliability analysis of digital instrumentation and control systems, software
reliability modeling and software test automation, and distributed test
facility design. She is the author of more than 175 refereed journal and conference
publications, is the recipient of multiple awards and 3 patents. Dr. Smidts’
research has been sponsored by Government (DOE, AFSOR, AFRL, NRC, NASA, NSF,
FAA, DOD, NSA) as well as by industry (Texas Instruments, IBM). She is an IEEE
Fellow, and was the conference co-Chair of the IEEE International Symposium on
Software Reliability Engineering (2006 and 2013), IEEE High Assurance Systems
Engineering (2008), is an Associate Editor for Software Testing Verification
and Reliability, was Associate Editor for the IEEE Transactions in Reliability
(2008-2016), was one of Technical Program Committee Co-Chairs of
NPIC-HMIT 2019, is the Program Chair of 2021 International Topical Meeting on
Probabilistic Safety Assessment and Analysis (PSA 2021), is the Vice
Chair of ANS’ Human Factors Instrumentation and Control Division (HFICD), and
is a regular member on review panels (DOE, NSF, FDA, ISSRE, HASE, DSN).