User-customizable Shared Control for Robot Teleoperation via Virtual Reality

TL;DR

This paper addresses the difficulty novices face in understanding and benefiting from shared control arbitration during robot teleoperation. It proposes a user-customizable framework in which arbitration parameters are observable and directly editable via a VR interface, formalized with $\mathbf{u}_{sc} = (\mathbf{I}-\mathbf{A}_{\bm{\theta}})\mathbf{u}_r + \mathbf{A}_{\bm{\theta}}\mathbf{u}_h$ and a bidirectional feedback channel $ \bm{\psi}(\mathbf{v}_r,\mathbf{v}_h)$. The approach is instantiated for SE$(3)$ teleoperation using a 7-DoF Kinova arm in a buzz wire game, featuring a real-time assistive wrench computed from potential fields and a spider-chart UI to edit arbitration factors. In a longitudinal study with 12 participants over four sessions, users who could actively tune arbitration achieved higher precision, smoother control, and better transfer to a novel wire task compared with direct teleoperation or fixed heuristics.

Abstract

Shared control can ease and enhance a human operator's ability to teleoperate robots, particularly for intricate tasks demanding fine control over multiple degrees of freedom. However, the arbitration process dictating how much autonomous assistance to administer in shared control can confuse novice operators and impede their understanding of the robot's behavior. To overcome these adverse side-effects, we propose a novel formulation of shared control that enables operators to tailor the arbitration to their unique capabilities and preferences. Unlike prior approaches to customizable shared control where users could indirectly modify the latent parameters of the arbitration function by issuing a feedback command, we instead make these parameters observable and directly editable via a virtual reality (VR) interface. We present our user-customizable shared control method for a teleoperation task in SE(3), known as the buzz wire game. A user study is conducted with participants teleoperating a robotic arm in VR to complete the game. The experiment spanned two weeks per subject to investigate longitudinal trends. Our findings reveal that users allowed to interactively tune the arbitration parameters across trials generalize well to adaptations in the task, exhibiting improvements in precision and fluency over direct teleoperation and conventional shared control.

Figures (8)

• Figure 1: A user wearing a virtual reality (VR) headset teleoperates the Kinova Gen 3 robotic arm to play the buzz wire game. The VR controller acts as a handle in the virtual world, where its simulated effects also have a direct consequence on the physical counterpart played by the robot.
• Figure 2: The user-customizable shared control architecture.
• Figure 3: Illustration of how the attractive and repulsive forces are computed given a control point $\mathbf{p}_i$ (dark blue) and environment point $\mathbf{p}_k$ (green/red).
• Figure 4: The customizable spider chart interface in virtual reality. The operator can modify any of the five factors in steps of $5$ before confirming their choice. Each factor is scaled from $[0.1, 1]$ to $[10, 100]$.
• Figure 5: New wire testbed for the last session to assess skill generalization.