KEYNOTE LECTURE

Shervin Shirmohammadi

University of Ottawa, Canada

Artificial intelligence is increasingly embedded in defense and security technologies, including imaging and sensing, radar, sonar and acoustic signal processing, multisensor data fusion, marine and underwater systems, surveillance, transportation, and logistics. In these applications, measurements are used to train AI models, while AI models are used for indirect measurement, detection, classification, tracking, diagnosis, and prediction. For safety- and mission-critical systems, however, an AI output must be accompanied by an indication of its reliability. Measurements may be noisy, incomplete, corrupted, mutually inconsistent, or acquired under conditions that differ substantially from the training data. Uncertainty quantification is therefore essential for risk-aware decision making, robust sensor fusion, human–machine collaboration, and trustworthy deployment.

This talk examines uncertainty-aware AI for defense and security applications. It introduces the use of AI for indirect measurement and decision support in imaging, radar, sonar, acoustic, marine, underwater, surveillance, transportation, and logistics systems. We discuss uncertainty in AI regression, classification, and large language models. The talk presents a unified taxonomy from the perspective of both AI and measurement science. Finally, examples from the literature are used to show how uncertainty can support more reliable, resilient, interpretable, and trustworthy AI-enabled technologies for defense and security.

Shervin Shirmohammadi received his Ph.D. in Electrical Engineering in 2000 from the University of Ottawa, Canada, and after spending 3 years in the industry as a senior architect and project manager, joined as Assistant Professor the same University, where since 2012 he has been a Full Professor with the School of Electrical Engineering and Computer Science. He is Director of the Discover Laboratory, doing research in AI-assisted measurements and sensing systems, specifically fundamentals of AI in measurement, human behavior and activity sensing, multimedia systems and network measurements, IoT measurements, and biomedical measurements. The results of his research, funded by more than $28 million from public and private sectors, have led to over 450 publications, over 80 researchers trained at the postdoctoral, PhD, and Master’s levels, 40 patents and technology transfers to the private sector, and five Best Paper awards. He was the Founding Editor-in-Chief of the IEEE Open Journal of Instrumentation and Measurement in 2022 and 2023, the Editor-in-Chief of the IEEE Transactions on Instrumentation and Measurement from 2017 to 2021, and the Associate Editor-in-Chief of IEEE Instrumentation and Measurement Magazine in 2014 and 2015, and is currently on the latter’s editorial board.
He has been on the Administrative Committee (AdCom) of the IEEE Instrumentation and Measurement Society (IMS) since 2014, currently serves as IMS’s President, and was a member of the IEEE I2MTC Board of Directors from 2014 to 2016.
Dr. Shirmohammadi is an IEEE Fellow “for contributions to multimedia systems and network measurements”, and recipient of the 2019 George S. Glinski Award for Excellence in Research, the 2021 IEEE IMS Distinguished Service Award, and the 2023 IEEE IMS Technical Award “for contributions to the advancement of machine learning-assisted measurements”.