Percutaneous nephrolithotomy (PCNL) is a minimally invasive procedure for removing kidney stones. Gaining kidney access is the most challenging task in this procedure and the cause of many complications.

This thesis presents semi-autonomous solutions for improving PCNL procedure outcomes. First, a cyber-physical PCNL simulator incorporating haptic feedback and simplifying surgeon mental workload via teleoperation of a nephroscope, the surgeon controls the tooltip position while a robotic agent controls its orientation.

This thesis then explores how subtask automation can further improve the procedure. A multi-objective path planning algorithm is implemented to generate multiple suitable paths for kidney access, from which an expert surgeon selects one for execution. The robotic agent then steers the tool along the path autonomously. A further advancement adds a tool/tissue interaction model which determines tool bending;​ thereby providing accurate trajectory tracking. The concepts are validated experimentally in ex-vivo and phantom tissues.