Session I — Operation and Maintenance

This session focuses on the applications of automation and optimization techniques for operating and maintaining current nuclear power plants, small modular reactors, microreactors, and fission batteries. Automation has the potential of significantly reducing operation and maintenance staff workload as well as human errors. For small modular reactors, microreactors, and fission batteries, there are significant opportunities for optimization of operation and maintenance activities to improve the efficiency and safety of nuclear systems. Optimization can also be performed for outage scheduling, inventory management, and fuel usage for current nuclear power plants to improve their efficiency. Topics to be considered include areas of applicability, performance targets, managing uncertainty, data sharing, interpretability and explainability, risk issues related to the use of Big Data & AI/ML, and other related topics.

Session II — High-Performance Computing, Massive Computation

High-Performance Computing (HPC) is the practice of aggregating computing to deliver much higher performance than one could get out of a typical desktop computer or workstation. The goal of HPC is to solve large problems in science and engineering. Although HPC has progressed remarkably over the past couple of decades, in recent years that progress has been achieved through greatly increased hardware complexity with the rise of multicore and manycore processors, and this is affecting the ability of application developers to achieve the full potential of these systems. In other words, HPC systems are becoming more and more complex and the hardware is exposing massive parallelism at all levels, making it a challenge to fully utilize these resources. In this track, we aim for the participants to address challenges and opportunities in several areas of HPC in support of Big Data at the university, company, and national laboratory scale for current nuclear power plants, small modular reactors, microreactors, and fission batteries. What hardware should be purchased (CPU vs. GPU)? What are the unique challenges that Big Data poses to hardware? How do we balance the needs of engineering level computation and data analytics? Can high-fidelity M&S support big data analytics? What are ways to use high-performance data analytics to improve the accuracy of predictions? How do smaller companies and universities with less access to computational resources utilize data analytics? Would high performance computing lead to better results/more utility of Big Data? These and other equally important questions will be identified and examined, and fruitful avenues to improve the handling of Big Data through HPC will be proposed.

Session III — Cybersecurity Session

With the digitalization and computerization of instrumentation and control, nuclear power plants have become more vulnerable to cyber-attacks than ever before. Such attacks may have severe implications for a plant’s operation and safety. This session aims at discovering and solving cybersecurity issues for Big Data & AI/ML applications in the nuclear field, as well as applying Big Data & AI/ML technologies to cybersecurity efforts incorporating robust cybersecurity policies, procedures, and practices to protect vital components of current nuclear power plants, small modular reactors, microreactors, and fission batteries. Topics of interest include but are not limited to confidentiality, integrity, and availability issues in Big Data & AI/ML applications, cyber threats in data sharing for AI/ML model training, adversarial uses of AI/ML models, as well as using Big Data & AI/ML for cybersecurity of Instrumentation and Control (I&C) system functions, identification and assessment of cyber threats, and analysis of the attractiveness of the I&C system to potential adversaries, vulnerabilities of the I&C system, operating environment, and potential consequences that could either directly or indirectly result from a compromise of nuclear power plants.

Session IV — Machine Learning in Nuclear Materials and Advanced Manufacturing

Big Data analytics tools are finding applications in advanced manufacturing and materials science for current nuclear power plants, small modular reactors, microreactors, and fission batteries. Examples include use of image processing tools to identify flaws for qualification of additively manufacturing process, characterization of microstructure obtained using electron microscopy and x-ray imaging tools, use of machine learning algorithms to accelerate prediction of materials properties and their behavior in extreme environments, and utilization of machine-learned interatomic potentials for atomistic simulations of microstructure evolution.

Session V — Big Data and Digital Twins

Digital twins (DTs) are now actively being developed and used in many stages of the nuclear power plant life cycle, including use for design, licensing, construction, operations, oversight, monitoring, and maintenance. A recent report published by the U.S. Nuclear Regulatory Commission identified several challenges and gaps associated with application of DT in nuclear, which included real time integration of sensor data with DT, and use of traditional modeling and simulation tools as data-informed models. Data quality, quantity, applicability, and uncertainty have been identified as some of the key technical considerations in regulatory decision making in the NRC’s AI Strategic Plan of 2023. This session focuses on considerations, opportunities, challenges, and gaps associated with the integration of big data with DTs in nuclear applications.

Session VI — Nuclear Non-Proliferation

As defined in the Nuclear Non-Proliferation Treaty (NPT), the International Atomic Energy Agency (IAEA) is tasked with promoting the peaceful use of nuclear energy technologies while deterring and detecting the spread of nuclear weapons, as mandated by NPT State signatories. Big Data technology can be used to facilitate new scientific discoveries in physics, chemistry, and biology to advance security solutions for current nuclear power plants, small modular reactors, microreactors, and fission batteries. This session will focus on Big Data technologies that can be reused in nuclear non-proliferation applications, such as the Big Data technologies supporting the detection of undeclared plutonium and uranium, the Big Data technologies for ultra-sensitive nuclear measurement systems, the Big Data technologies for nuclear explosion detection, and the Big Data technologies for legal and regulatory analysis, etc.