PNNL

Draguna Vrabie serves as interim Director of the Advanced Computing, Mathematics, and Data Division in the Physical and Computational Sciences Directorate. In this role, she leads the division’s strategic planning, capability development, and operational execution. Her leadership ensures alignment with institutional goals, fosters scientific and technical innovation, and strengthens internal and external partnerships across national security, energy, and scientific mission areas. As part of the PCSD Leadership Team, Vrabie plays a key role and contributes to shaping the directorate’s technical strategy, stewarding critical capabilities, and supporting cross-laboratory initiatives that advance data-driven science, scalable computing, and intelligent decision systems.

In January 2025, Vrabie was named Deputy Director for the Advanced Computing, Mathematics, and Data Division. From 2021 to 2024, Vrabie was the Team Leader for the Autonomous Intelligence Team within the Data Sciences and Machine Intelligence group, driving advancements in AI/ML-based control, optimization, reinforcement learning, and scientific machine learning. Since 2023, she has led the Autonomous Science strategy for the Physical and Computational Sciences Directorate, aligning foundational research in AI, control, and automation with application to scientific discovery. 

Vrabie joined PNNL in 2015 as a Senior Staff Scientist in the Energy and Environment Directorate, developing PNNL's capability for adaptive and predictive control for high-performance building energy systems. Between 2015 and 2017, she led the Control Theory portfolio part of PNNL's Control of Complex Systems Initiative, which focused on learning-based control, optimization-based control, and scalable computational methods for systems and control co-design. In 2019, Vrabie was named Chief Data Scientist. Between 2019 and 2023, she served as thrust leader for PNNL’s Data Model Convergence Initiative, where she developed and stewarded a Converged Applications project portfolio focused on algorithmic and computational methods for applications that integrate scientific modeling and simulation with data analytics and machine learning. Before joining PNNL, she was a Senior Scientist at United Technologies Research Center in the Control Systems Group, contributing to various commercial and aerospace applications, including energy-efficient buildings, power electronics, jet engine fleet maintenance, and mission planning for unmanned aircraft. She holds a Ph.D. in Electrical Engineering (2009) from the University of Texas at Arlington, and an M.E. and B.E. in Automatic Control and Computer Engineering from Gheorghe Asachi Technical University in Iași, Romania, where she specialized in predictive and learning-based control and robotics.

Vrabie's research bridges control theory, artificial intelligence, and scientific machine learning, advancing adaptive decision-making for high-performance cyber-physical systems where computational intelligence (software, algorithms, AI) operates in real time alongside physical processes (mechanical, electrical, thermal, etc.) to achieve precise, efficient, and reliable performance under dynamic conditions. She has made significant contributions to the theory of reinforcement learning, optimal control, and adaptive dynamic programming, with applications in robotics, energy systems, secure control, and autonomous systems.

Vrabie specializes in model predictive control, neural modeling, and scientific machine learning, advancing methodologies for physics-informed learning, differentiable predictive control, and constrained neural models with stability guarantees. Her team develops NeuroMANCER, PNNL’s most popular open-source differentiable programming library, which enables parametric constrained optimization, physics-informed system identification, and model-based optimal control. She is the co-author of highly cited works in reinforcement learning and control. She has co-authored three monographs, including Optimal Control and Reinforcement Learning and Adaptive Dynamic Programming for Feedback Control. 

Vrabie also holds several patents across a range of application areas, including cybersecurity for critical infrastructure, AI/ML-based optimal asset management and predictive maintenance for real-time decision support for aerospace and industrial systems, power and energy system optimization, with a focus on intelligent load response, and grid stability, intelligent automation for smart buildings and cold chain logistics. In 2021, she received an R&D 100 Award for her contributions to Shadow Figment, a cybersecurity technology designed to defend critical infrastructure, such as buildings and the electric grid, against cyberattacks.

Beyond her research and innovation, Vrabie has led many workshops on reinforcement learning and control, scientific machine learning, and AI-driven decision-making. She was a plenary speaker at the 2022 American Control Conference (ACC), where she presented her work on integrating AI and control for next-generation autonomous systems.

Vrabie is the recipient of numerous awards. In 2022, she received the Most-Cited Article Award from the Annual Reviews in Control Journal for articles published after 2019. In 2021, she received the Best Paper Award from the Journal of Building Performance Simulation. In 2020, she received the Best Paper Award from the Energy Systems Technical Committee of the American Society of Mechanical Engineers at the Dynamic Systems and Control Conference. In 2017, she was named “Engineer of the Year” by the Women in Engineering section of the Institute of Electrical and Electronics Engineers (IEEE) in Richland, WA. In 2013, she received the Operational Excellence Award, and in 2012, she received an Outstanding Achievement Award from the United Technologies Research Center. In 2010, she received the Best Paper Award at the International Joint Conference on Neural Networks. In 2004, she received the Best Paper Award in the Advanced Control, Modeling and Simulation Session, International Conference on Automation, Quality and Testing, Robotics.

• Control Systems Theory
• Reinforcement Learning
• Model Based Predictive Control
• Optimization; Game Theory
• Deep Learning
• Scientific Machine Learning
• Dynamic systems
• Autonomous systems
• Robotics

The University of Texas at Arlington
Doctor of Philosophy, Electrical Engineering

Gheorghe Asachi Technical University of Iasi
Master of Engineering, Advanced Tech Studies

Gheorghe Asachi Technical University of Iasi
Bachelor of Science, Engineering Systems

IEEE

• 2022 Most-Cited Article Award from the Annual Reviews in Control Journal, for articles published after 2019. 
• Best Paper Award from the Journal of Building Performance Simulation, 2021.
• R&D100 Award, 2021
• Best Paper Award, Energy Systems Technical Committee, ASME Dynamic Systems and Control Virtual Conference, 2020
• Engineer of the Year, IEEE Richland Section, Women in Engineering, 2017
• Operational Excellence Award, for setup of high-performance building testbed and demonstration of advanced applications, United Technologies Research Center Corporate Award, 2013
• Outstanding Achievement Award, for demonstration of advanced energy efficient buildings controls, United Technologies Research Center, 2012
• Best Paper Award, International Joint Conference on Neural Networks, Barcelona, Spain, 2010
• Outstanding Accomplishments, Department of Electrical Engineering, University of Texas at Arlington, 2009
• Automation and Robotics Research Institute Best Student Award, for distinguished scholarly accomplishments and outstanding service to pears and to the Automation and Robotics Research Institute community at large, 2009
• STEM Scholarship, University of Texas at Arlington, 2006 – 2009
• Best Paper Award in Advanced Control, Modeling and Simulation Session, International Conference on Automation, Quality and Testing, Robotics, 2004

A full list of Draguna Vrabie’s publications can be found on her Google Scholar profile.