During laparoscopic surgery, tissues frequently need to be retracted and mobilized for manipulation or visualisation. State-of-the-art robotic platforms for minimally invasive surgery (MIS) typically rely on rigid tools to interact with soft tissues. Such tools offer a very narrow contact surface thus applying relatively large forces that can lead to tissue damage, posing a risk for the success of the procedure and ultimately for the patient. In this paper, we show how the use of Pneumatically Attachable Flexible (PAF) rail, a vacuum-based soft attachment for laparoscopic applications, can reduce such risk by offering a larger contact surface between the tool and the tissue. Ex vivo experiments are presented investigating the short-and longterm effects of different levels of vacuum pressure on the tissues surface. These experiments aim at evaluating the best tradeoff between applied pressure, potential damage, task duration and connection stability. A hybrid control system has been developed to perform and investigate the organ repositioning task using the proposed system. The task is only partially automated allowing the surgeon to be part of the control loop. A gradient-based planning algorithm is integrated with learning from teleoperation algorithm which allows the robot to improve the learned trajectory. The use of Similar Smooth Path Repositioning (SSPR) algorithm is proposed to improve a demonstrated trajectory based on a known cost function. The results obtained show that a smoother trajectory allows to decrease the minimum level of pressure needed to guarantee active suction during PAF positioning and placement.

No abstract