Techniques for environment parameter estimation during telemanipulation

Teleoperation allows surgeons to perform an operation that is remote in distance and/or scale. Extracting information about a patient, particularly the dynamic model of tissues, during a surgical procedure may be useful for improving telemanipulator control, developing simulations, and performing automated diagnosis. This study examines automated environment parameter identification methods for bilateral telemanipulation, with a focus on surgical applications. We first present a multi-estimator technique and demonstrate that, in practice, it finds the best estimator for a Kelvin-Voigt material. Using a one-degree-of-freedom teleoperation system, cubes with various material properties were palpated to acquire data under three control conditions: teleoperation without persistent excitation, teleoperation in which the operator mimics persistent excitation, and autonomous control with persistent excitation. The estimation performance of three online estimation techniques (recursive least-squares, adaptive identification, and multi-estimator) are compared. Neither the cube type nor the control condition affected the estimation performance. By considering practical aspects, recursive least-squares or multi-estimator would be suitable for online estimation of the tissue property.