Implementation of haptic feedback in minimally invasive surgical teleoperator systems may lead to improved performance in many common surgical procedures; however, most of the currently available surgical teleoperators do not provide force feedback, mainly because of the associated stability issues. In this paper, we study the effect of a special type of force reflection algorithms, called projection-based force reflection (PBFR) algorithms, on the stability and performance of a dual-arm haptics-enabled teleoperator system for minimally-invasive surgical applications. In particular, the teleoperator system's performance is experimentally evaluated in three common tasks, which are knot tightening, pegboard transfer, and object manipulation, in the presence of negligible as well as non-negligible communication delays. The results obtained indicate that, in almost all cases, the PBFR algorithms demonstrate statistically significant improvement of performance in comparison with conventional direct force feedback.

The projection-based force reflection (PBFR) algorithms were previously demonstrated to substantially improve stability characteristics of the force reflecting teleoperator systems and haptic interfaces without transparency deterioration in the steady state; however, the transient response of the PFBR algorithms suffers from relatively slow force convergence. In particular, the high frequency component of the contact force, which is very important for the haptic perception of stiff surfaces, is typically filtered out. In this paper, a solution to this problem is proposed which is based on the idea to separate different frequency bands in the force reflection signal and consequently apply the projection-based principle to the low-frequency component, while reflecting the high-frequency component directly. It is shown that, for bilateral teleoperators with irregular communication delays, stability can always be achieved by implementing the above described force reflection scheme, if the cut-off frequency of the complementary filters is sufficiently high and a certain weighting coefficient in the force reflection algorithm is sufficiently low. Experimental results demonstrate that substantial simultaneous improvement of stability and transparency is achieved using the proposed method.

Modeling errors are one of the elements responsible for deficiencies in sensorless collision detection of robotic systems. By providing a specific formulation of manipulator equations, this paper presents a time-variant threshold for observer residues in joint space of a manipulator in order to provide more accurate collision detection ability compared to constant thresholds. The time-variant threshold considers modeling errors in both robot's parameters and its friction forces. Simulation results using real-life collision forces and experiments on the Phantom Omni device are provided to validate the proposed scheme.

For cooperative force-reflecting teleoperation over networks, conventional passivity-based approaches are not always applicable due to possibility of nonpassive slave-slave interactions and irregular communication delays imposed by networks. In this work, we propose a design approach that is based on an advanced version of the small-gain theorem developed previously for complex network-based interconnections. It is shown that, using the proposed design approach, a networked cooperative force-reflecting teleoperator system can be made stable in the presence of irregular communications by adjusting the local control gains appropriately. Experimental results are presented that confirm the validity of the proposed approach.

Developments to the small-gain design approach to networked cooperative force-reflecting teleoperators are presented. Explicit assumptions on the human dynamics are incorporated into the stability analysis, and the conservatism of the small-gain design is eliminated using the projection-based force reflection principle. Results are established that describe the stability properties of the networked cooperative teleoperator system under different conditions on the human dynamics and for different types of the force reflection algorithms. Samples of experimental results are presented.