Virtual Elasto-Plastic Compliance Control for Robotically Assisted Laparoscopy

Kurzfassung

Robotically assisted laparoscopic surgery (RALS) offers significant advantages over conventional laparoscopy but introduces new challenges regarding surgical workflow and requires integration of robotic platforms into complex operating room (OR) environments. Elasto-plastic robot compliance (EPRC), an extension of classical impedance control, enables robots to exhibit compliant behavior when encountering an active environment (AE) while maintaining the ability to apply forces in contact with passive environments. During the course of this thesis, limitations of the original EPRC formulation were identified. Inconsistencies in the sensitivity for activation depending on the direction of interaction and commanded velocity were observed, which reduce predictability in interaction. This thesis investigates the potential of EPRC for RALS and develops adaptations for consistent activation sensitivity. A systematic parameter study and characterization of coupling effects between movement direction and powers from system dynamics that could falsely trigger EPRC identified threshold settings that enable reliable activation while being robust against false activations. By rotating the EPRC frame, augmenting the control law by a factor to correct the position drift, and adapting the commanded velocity to improve differentiation between active and passive environments, the modified EPRC formulation achieved consistent activation sensitivity for all directions of interaction, independent of commanded velocities. The enhanced EPRC formulation was then evaluated in RALS-specific use cases. Control of the nullspace motion using EPRC resulted in compliant repositioning that matched manual button-activated motions. Preliminary evaluation of the EPRC-based adjustment of a tissue-retraction instrument in solo-surgery scenarios demonstrated that surgeons could reposition with the retraction instrument using other instruments with a reasonable effort from within the patient, potentially reducing workflow interruptions. Feasibility analysis of the role of EPRC in hierarchical control indicated that EPRC can be implemented on multiple priority levels when other priority levels are considered during threshold tuning. The verified general extensions of EPRC control improve its sensitivity for high precision and autonomous taks. Furthermore, the results of this thesis demonstrate its feasibility also in surgical robotics, enabling a consistent human-robot interaction and establishing a foundation for future applications. Future work includes the application of EPRC to enable compliant movement of the trocar constraint, e.g. allowing repositioning of the OR table without the need for an environmental model.