RoboHaptics: Designing Haptic Interactions for Lower Body with Quadruped Robot Dogs

Paper Title

RoboHaptics: Designing Haptic Interactions for Lower Body with Quadruped Robot Dogs

Publication Info

• Topic area: Lower-body haptic feedback using quadruped robots in XR environments.
• Keywords: Haptics, quadruped robots, XR, lower-body feedback, tactile effects, immersion, precision, design toolkit, robot-mediated interaction, virtual reality.

Background and Problem

• Problem / challenge: Current haptic systems for XR primarily focus on the upper body and wearable devices, which are limited in scalability, modality, and encountered-type feedback for lower-body interactions. Quadruped robots have potential for mobile, off-body haptics but are underexplored for this purpose.
• Significance: Addressing lower-body haptics enhances immersion in XR by providing realistic tactile feedback to regions like legs and knees, which are critical for full-body interaction.
• Motivation and related work: Prior research has explored wearable haptics, grounded mechanical platforms, and robot-mediated haptics, but these approaches face challenges in portability, spatial flexibility, and scalability. Quadruped robots, such as the Unitree Go2, offer mobility and precise positioning, but their use for lower-body haptics remains largely unexplored.

Solution

• Proposed approach: RoboHaptics—a system leveraging commercial quadruped robots to deliver lower-body haptic feedback in XR environments.
• Novelty:
  1. Introduction of the first design space for quadruped-mediated haptics tailored to lower-body interactions.
  2. Development of a haptics design toolkit enabling programmable tactile effects on quadruped robots.
  3. Empirical experiments characterizing force profiles, precision, and reachability of quadruped robots for haptic feedback.
  4. Demonstration of application scenarios in XR, including sports simulations, object proxies, and encountered-type haptics.
• Procedure and key techniques:
  • Design space exploration covering user-robot scale, tracking configurations, body locations, robot parameters, and haptic interaction types.
  • Development of a software toolkit with motion primitives for haptic effects (e.g., Knock, Tap, Press).
  • Empirical evaluation of force output, precision, and accuracy using load cells and mannequin-based setups.
  • User studies assessing immersion, realism, and comfort in VR scenarios.

Results

• Concrete findings:
  • Force feedback ranged from 3–28 N, below nociceptive thresholds (~40 N).
  • Spatial accuracy: 3.7–5.5 mm in Laydown posture; up to 17 mm in Standing.
  • Precision: 2.1–5.1 mm across postures.
  • User ratings showed realism and immersion close to ground-truth objects in rigid, planar scenarios.
• Advantage over baselines:
  • Provides scalable, mobile, and reconfigurable haptic feedback without requiring worn devices.
  • Achieves high precision and accuracy suitable for lower-body tactile rendering.
  • Supports diverse tactile effects and application scenarios beyond traditional wearables.
• Experiments / evaluation:
  • Force characterization: Load cell measurements across six haptic effects and three robot postures.
  • Reachability and precision: Mannequin-based trials across postures and orientations.
  • User studies: 12 participants rated comfort, immersion, and realism in VR scenarios comparing quadruped-mediated haptics to real objects.
• Limitations and future work:
  • Hardware constraints: Limited sensor access, motor noise, and battery life.
  • Geometric mismatches for curved/compliant objects like balls.
  • Single-user focus; future work should explore multi-user scenarios and richer haptic vocabularies.

Summary

RoboHaptics introduces a novel system for delivering lower-body haptic feedback using commercial quadruped robots in XR environments. The study establishes a design space, develops a software toolkit, and demonstrates feasibility through technical experiments and user studies. Results show high precision, safe force delivery, and immersive tactile experiences, particularly for rigid and planar objects. While limitations exist in hardware and geometric fidelity, this work opens new avenues for scalable, mobile haptics in XR, positioning quadruped robots as versatile platforms for tactile feedback in immersive applications.