Linearity, Time Invariance, and Passivity of a Novice Person in Human Teleoperation

TL;DR

This work addresses how a human follower in MR-guided teleoperation behaves dynamically and whether it can be modeled as a passive linear time-invariant system with minimal inter-axis coupling. Using a mass–spring–damper state-space model, linear grey-box identification, and experiments with 12 participants across Fourier and low-pass white-noise trajectories, the study demonstrates strong evidence for LTI behavior, approximate time-invariance, and robust passivity up to about $4.6$ Hz, while revealing stochastic variability that can be captured with an additive Gaussian term. The key contributions include a quantified axis-decoupling result, high-fidelity linear fits, a practical stochastic follower model, and an argument for using passivity-based control in low-cost teleoperation. The findings enable simpler controller design for stable, transparent bilateral teleoperation in MR-guided medical tasks, with implications for adaptive control and robustness to human factors.

Abstract

Low-cost teleguidance of medical procedures is becoming essential to provide healthcare to remote and underserved communities. Human teleoperation is a promising new method for guiding a novice person with relatively high precision and efficiency through a mixed reality (MR) interface. Prior work has shown that the novice, or "follower", can reliably track the MR input with performance not unlike a telerobotic system. As a consequence, it is of interest to understand and control the follower's dynamics to optimize the system performance and permit stable and transparent bilateral teleoperation. To this end, linearity, time-invariance, inter-axis coupling, and passivity are important in teleoperation and controller design. This paper therefore explores these effects with regard to the follower person in human teleoperation. It is demonstrated through modeling and experiments that the follower can indeed be treated as approximately linear and time invariant, with little coupling and a large excess of passivity at practical frequencies. Furthermore, a stochastic model of the follower dynamics is derived. These results will permit controller design and analysis to improve the performance of human teleoperation.

Figures (9)

• Figure 1: Overview diagram and corresponding two-port network representing the human teleoperation system, showing the follower block in blue. (A) the haptic device, (B) MR headset, (C) follower's instrumented tool and the virtual tool. The red arrow shows how the follower will move next to align the real tool to the virtual one. In this example, the system is used for teleultrasound.
• Figure 2: Diagram of the experimental system. (A) Follower haptic device. (B) Ultrasound probe-shaped handle with IR markers. (C) Virtual tool in the same shape. (D) Microsoft HoloLens 2 worn by the follower to render the virtual tool. (E) Follower-side computer to control the haptic device. (F) WebRTC communication between expert and follower PCs and the HoloLens. (G) Expert-side computer to generate trajectories and control their haptic device. (H) Expert haptic device (not used in these experiments).
• Figure 3: Fitted model versus measured output for a typical (A) white noise trajectory, (B) Fourier trajectory, and (C) Fourier trajectory with a virtual environment, leading to contact forces. This shows the accuracy of the model even for different types of inputs.
• Figure 4: Fourier transforms of the Fourier series (A) and white noise (B) trajectories, showing very close agreement in frequency content between expert and follower (input and output).
• Figure 5: Coherence function of a typical follower, showing high coherence between the input and output signals of the follower up to the cutoff frequency of 0.6 Hz.