Public defence, Automation, Systems and Control Engineering, MSc Ogulcan Isitman

Title of the thesis: Control and Planning for Magnetic Robotic Manipulation Across Scales

Thesis defender: Ogulcan Isitman
Opponent: Prof. Sarthak Misra, University of Twente, The Netherlands
Custos: Prof. Quan Zhou, Aalto University School of Electrical Engineering

Magnetic robotic manipulation studies how magnetic fields can be used to move and control small objects without direct physical contact. This thesis focuses on developing control methods for magnetic manipulation across micro and meso scale biomedical systems, where the target may be fragile, very small, or located inside the human body. At the micro scale, magnetic fields can guide particles, cells, and microrobots for tasks such as targeted therapy, diagnostics, and micro assembly. At the meso scale, they can steer medical devices such as capsule endoscopes through anatomically constrained environments. The purpose of the study was to make magnetic manipulation more reliable, precise, and constraint aware despite nonlinear magnetic forces, strong coupling effects, and scale dependent physical challenges.

The thesis shows that coordinated magnetic actuation can enable safe and precise manipulation of a single microparticle, that multiple nearby microparticles can be controlled simultaneously and independently using a data driven magnetic field model, and that magnetic capsules can be navigated safely using trajectory planning and feedback control under dynamic, actuator, and anatomical constraints. The main result is a coherent optimization based framework for constrained magnetic manipulation across scales. The research contributes new methods for coordinating multiple magnetic actuators, learning complex magnetic field interactions, and planning safe capsule motion. These results can support future biomedical robotic systems, including micromanipulation platforms, targeted therapy tools, and magnetic capsule endoscopy. Overall, the thesis demonstrates that combining modeling, optimization, learning, and feedback control can make magnetic robotic manipulation more scalable and suitable for complex practical environments.

Key words: magnetic manipulation, robotic micromanipulation, robotic control, trajectory planning

Thesis available for public display 7 days prior to the defence at Aalto University's public display page.