Speaker:

Dr. Hassan ALMubarak

Assistant Professor
Control and Instrumentation Engineering Department

King Fahd University of Petroleum and Minerals

Dhahran, Saudi Arabia

Abstract:

Modern autonomous systems must make fast, high-stakes decisions under uncertainty while obeying hard safety constraints. This talk presents a unified framework for safety-critical decision and control via safety-embedded modeling. A new notion termed Barrier States (BaS) will be introduced, transforming constraints into barrier dynamics whose boundedness certifies forward invariance of the safe set. This embedding enables wide range of legacy and advanced controls including Differential Dynamic Programming, differentiable/tube-based MPC, learning-based methods, without ad-hoc penalties, achieving formal guarantees and optimal performance. Subsequently, swarm architectures for large-scale multi-agent systems that distribute safety-embedded optimization over sparse communication graphs are outlined, yielding provably safe formation/coverage/collision avoidance with near-linear scalability. Finally, this talk will preview fast adaptive MPC that couples ultra-lightweight online identification with model-bank/structure selection to track rapidly varying dynamics at control rates for agile platforms advancing data-efficient, certifiably safe autonomy.

Speaker Bio:

Dr. Hassan Almubarak is an Assistant Professor in the Department of Control and Instrumentation Engineering at King Fahd University of Petroleum and Minerals (KFUPM), Saudi Arabia, and an affiliate of the Interdisciplinary Research Centre for Smart Mobility and Logistics (SML). His research lies at the intersection of control theory, optimization, and machine learning, focusing on safety-critical, intelligent, and adaptive decision-making for autonomous systems. He is particularly working on bridging large-scale generative and multimodal AI models with control theory, developing frameworks for AI-driven control solutions, and control-inspired ​learning for trusted autonomy.

Before joining KFUPM, Dr. Almubarak served as a Lead Research Engineer at General Electric (GE) Vernova Advanced Research Center (ARC) in New York, where he led cross-functional teams developing Generative AI programs that delivered over $5 million in direct savings and resulted in multiple intellectual-property submissions. He also contributed to the development of advanced distributed model predictive control and estimation methods for large-scale multi-agent renewable energy systems.

He earned his Ph.D. in Electrical and Computer Engineering from the Georgia Institute of Technology, where his research at the Autonomous Control and Decision Systems Lab produced key contributions in safe model predictive control, distributed multi-agent optimization, and differentiable control algorithms. His publications appear in leading venues including IEEE TAC, RA-L, RSS, ICRA, IROS, CDC, and ACC, and has invited talks in various international centres including NASA JPL and the Center for Autonomous Vehicles in Air Transportation Engineering (AVIATE) in the USA.