Telerobotic and Biorobotic Systems Group
University of Alberta
Department of Electrical and Computer Engineering
The Web TBS
 
   
 
 
 

"Bilateral" teleoperation

For haptic teleoperation, user's motion commands are fed forward from the master to the slave, and environment interaction forces are fed back from the slave to the master (thus, bilateral control). Using my master-slave system, I conducted haptic teleoperation experiments involving a tissue palpation task (probing tissue for estimating its characteristics). For different teleoperation controllers (position-error based; direct force reflection; four-channel), the teleoperator performance was assessed and compared in terms of objective criteria that evaluate accurate transmission to the user of certain task-related information (this is called "teleoperation transparency"), e.g., force versus deformation characteristics of a tissue in the case of a palpation task.

4-channel bilateral teleoperation system.

Typical torque–displacement relationships measured at the slave and as perceived by the user for a silicon-based
phantom (solid) and a foam object (dotted) as the environment.

This research also involved robot modeling and friction compensation, and hand force estimation using a state observer.

Time delay compensation in bilateral teleoperation

I then focused on time delay compensation in long-distance haptic teleoperation including telesurgery. The latency introduced by the communication channel reduces efficiency by requiring the user to slow down on each movement and can, in the worst case, cause instability in the teleoperation system. Moreover, time delays significantly change the “feel” or the perceived mechanical impedance of an object, which is a measure of teleoperation transparency. In this research, the scattering approach was extended and new models of a wave-based delay compensated communication channel were introduced. The research showed that using slave-side force measurements significantly improve transparency in comparison to the traditional delay compensation scheme. The research also bridged the passivity framework to the 4-channel bilateral control architecture, which has far better transparency characteristics compared to traditional architectures (position error based and direct force reflection). The proposed wave-based 4-channel teleoperation control architecture ensures ideal transparency under time delay, which is a significant advantage over the previous methods.  

(Please use Internet Explorer)

Discrete-time control in bilateral teleoperation

Discretization of a stabilizing continuous-time bilateral teleoperation controller for digital implementation may not necessarily lead to stable teleoperation. I studied the stability of master-slave teleoperation under discrete-time bilateral control. Stability regions were determined in the form of conditions involving the sampling period, control gains including the damping introduced by the controller, and environment stiffness. Due to the tradeoff between stability and transparency in bilateral teleoperation, such stability boundaries are of particular importance when the teleoperation system has good transparency.

Discrete-time controlled version of the 4-channel teleoperation system.

As the sampling period T is increased, the maximum environment stiffness k_e with which a slave robot can stably interact is reduced.

Effect of link and joint flexibility in a teleoperated robot

In applications such as space and surgical robotics, the use of thin, lightweight manipulators and cable-driven end effectors results in link and joint flexibility of the manipulator. In bilateral teleoperation, however, any flexibility in a link or joint of the robot reduces the effective stiffness of the slave and the transparency of teleoperation. 




(a) The rigid master; (b) the flexible-link slave.


Models of the operator, master, flexible slave, and environment.

I analyzed master-slave teleoperation transparency under slave robot joint and link flexibility, and evaluated the added benefits of using extra sensors at the end-effector of the flexible robot. Velocity (or position) feedback from the tip of the flexible robot improves free-space position tracking performance, which in the absence of such feedback is hampered by the system's anti-resonance. Also, the flexiblity in the joint or link will be transmitted to the user during a hard contact task unless end-effector velocity feedback is used.


Performance indices of different teleoperation architectures and sensor configurations.

Back to Research