Abstract:

This seminar presents a comprehensive survey of agile quadrotor flight in constrained environments, with a focus on the systems-level integration of modeling, control, perception, planning, simulation, and hardware design. We outline key principles, representative methods, and recent developments across these domains, drawing connections between foundational techniques and emerging trends. Both classical and learning-based approaches to modeling quadrotor dynamics are reviewed, alongside advances in geometric and reinforcement learning control. Modern trajectory planning methods suited for tight and dynamic spaces are analyzed, and the role of onboard perception—including visual-inertial odometry, depth sensing, and state estimation—is discussed in relation to agile navigation. Additionally, simulation platforms are compared, and hardware configurations enabling real-world deployment are examined.

Speaker Bio:

Mr. Ziad Shoeib is a Master Student in the Control and Instrumentation Engineering Department at King Fahd University of Petroleum and Minerals, Saudi Arabia, specializing in Systems and Control Engineering. He received his Bachelor’s degree in Mechatronics Engineering from Misr University for Science and Technology, Egypt, in 2023. His research focuses on agile quadcopter flight, with emphasis on trajectory tracking, reinforcement learning, and geometric control in constrained environments. He actively explores the integration of physics-based modeling with learning-based approaches, combining theoretical insights with hands-on robotics development. His work spans control, planning, perception, and simulation, with a focus on enabling autonomous aerial systems to navigate tight and dynamic environments.