VR-Based Teleoperation Framework: Integration of Haptic Feedback and Singularity Management

This study presents a VR-based teleoperation framework enhancing collaborative robot stability and manipulability via hand-tracking, adaptive control, and dual-modality haptic feedback. It addresses critical synchronization challenges (singularity avoidance, tracking responsiveness, and workspace constraints), which are especially problematic in first-person VR where kinematic limits are not directly perceivable. The framework employs Adaptive Damped Least Squares (A-DLS) to maintain manipulability near singular configurations, workspace impedance control to enforce boundary constraints, and vibrotactile feedback delivered through a haptic glove to convey both workspace limits (fingertip vibration) and path deviation information (wrist vibration) to operators. Key features include real-time hand-tracking, workspace calibration, and adaptive controls to ensure seamless coordination between virtual and real robot workspaces. Experimental validation through two complementary studies demonstrates the system’s effectiveness. Experiment 1 evaluated singularity management and workspace stability, showing that the A-DLS algorithm maintained manipulability above critical thresholds for 92% of operational time versus 78% without adaptive damping. Experiment 2 assessed trajectory tracking accuracy through a path-following task with 10 participants. Results demonstrate that haptic-enabled control achieves a 24.7% reduction in mean path-following error (from 10.03 mm to 7.55 mm, p = 0.001) compared to haptic-disabled conditions, indicating improvements in both accuracy and consistency. Although haptic guidance modestly increases task time due to higher precision focus, the resulting gains in accuracy and stability make this framework ideal for precision-critical tasks. By ensuring stability near workspace boundaries, the system effectively facilitates VR-based teleoperation for applications like painting, polishing, and contour-following.