During my Master's at the Human & Robot Interaction Lab (TaarLab), I led R&D initiatives spanning telerobotics, control theory, and mechatronic design.
This project addressed the challenge of operating mobile robots in dynamic environments where time delays and obstacles compromise safety. I developed a shared-control architecture where a human operator controls the robot via a haptic device, but the robot's autonomy (Motion Planning) intervenes to avoid collisions.
Figure: Schematic of deployed teleoperation.
Experimental validation of the shared-control framework avoiding dynamic obstacles.
Experimental validation of the motion planning for a single mobile robot.
We introduced a novel mathematical concept—the Arc Length Function—to estimate the Domain of Attraction (DA) for non-linear control systems. By minimizing the arc length of the system's phase trajectory, we could design controllers that significantly reduced oscillation.
Figure: Estimated DoA for different nonlinear systems.
Experimental oscillation damping controller for cable parallel robots.
Experimental oscillation damping controller for delta parallel robots.
I served as a lead control and CAD engineer for a national initiative to build a haptic-enabled VR dentistry trainer. The goal was to allow dental students to "feel" the difference between tooth enamel, decay, and gum tissue during virtual drilling procedures.
Figure: Modified end effectors of the haptic device.
Virtual reality environment for dentistry training.
We designed and operationalized a novel parallel robot. Parallel robots offer high precision and speed but are complex to control due to their closed-loop kinematics. This project involved the full lifecycle from optimal mechanical design to fabrication and control.
Designed and manufactured robot in action.