RedirectedStepper: Exploring Walking-In-Place Locomotion in VR Using a Mini Stepper for Ascents

Research Background and Problem

• Identified Problems or Challenges:

  • Simulating the realistic sensation of walking on slopes and stairs in virtual reality (VR) remains a challenge. Particularly, due to physical space constraints, existing technologies cannot provide a realistic inclined walking experience in limited spaces.
  • Current methods, such as Walking-in-Place (WIP) techniques or simple motion capture methods, fail to naturally simulate real walking experiences, resulting in insufficient immersion and realism during walking.
  • Navigation experiences across scenarios with varying surface heights (e.g., straight stairs, spiral stairs, slopes) still have significant room for improvement.

• Significance:

  • Height navigation in scenarios involving elevation changes (e.g., climbing stairs, hiking) is a critical requirement in VR applications such as architectural tours, training simulations, and recreational games.
  • Providing immersive experiences and realistic walking sensations enhances users' spatial immersion and satisfaction in VR environments.

• Research Motivation and Related Work:

  • Early methods relied on bulky or non-adjustable devices (e.g., mechanical floors) to simulate tactile feedback for height changes, making them difficult to apply widely.
  • Some portable devices (e.g., treadmills) have demonstrated effectiveness in other applications but have not been fully utilized for walking in VR scenarios, and research on their impact on user mapping and experience remains limited.
  • Simplistic walking techniques (e.g., WIP and depth cameras) are convenient but lack realistic height motion feedback.

Solution

• Proposed Method or Solution:

  • The authors designed a novel in-place walking solution called RedirectedStepper, which combines a portable structured mini fitness stepper with an innovative visual mapping method to simulate stair-walking experiences.
  • A new exponential motion mapping function was proposed to remap the real foot movement trajectory into a curved path consistent with the height trajectory in the virtual environment.

• Innovations:

  1. By utilizing a mini fitness stepper, the solution replaces complex mechanical floors, addressing the trade-off between hardware simplicity and interaction naturalness.
  2. Through the exponential mapping function, a more consistent simulation effect is achieved between users' physical actions and virtual objects.
  3. The system also provides tactile feedback (via spring damping), allowing users to experience a more realistic "stair-climbing" sensation.

• Implementation Steps and Technical Highlights:

  1. Use the stepper's compression springs and tilt sensors to capture the user's physical foot pedal movements.
  2. Apply the Kalman filter algorithm to filter sensor noise and optimize the collected angle measurement data.
  3. Remap the real and virtual foot movement paths using a custom exponential function and adjust the VR perspective to ensure user presence.
  4. Render various walking scenarios (e.g., straight stairs, spiral stairs, and slopes) in Unity 3D and integrate user testing based on Oculus Quest VR devices.

Research Outcomes

• Specific Outcomes:

  1. Compared to traditional Kinect-based WIP methods, RedirectedStepper significantly improved users' walking realism in stair scenarios.
  2. Users were better able to perceive height differences, and the force feedback for walking felt more realistic.
  3. Demonstrated potential for practical application in VR game design (prototype game "SnowRun").

• Advantages Over Existing Solutions:

  • Compared to WIP and other non-mechanical systems, RedirectedStepper utilizes tactile feedback and motion mapping for a more realistic simulation, better meeting the needs of height navigation scenarios.
  • Compared to high-end mechanical floors, the design is more affordable, lightweight, and operationally feasible.

• Experiment and Evaluation Results:

  1. Task Performance:
     • Test participants completed custom tasks involving stairs, spiral stairs, and ramps, showing improved realism.
     • RedirectedStepper significantly outperformed traditional Kinect methods in the "realism of foot sensation" metric.
  2. User Feedback:
     • Users reported positive feedback on the perceived height differences and the realism of walking feedback in the system.
  3. Limitations:
     • RedirectedStepper performed less effectively in continuous slope scenarios compared to stair scenarios, with some users finding the stepper's pedal direction and angle unnatural.

• Limitations and Future Directions:

  • The current design still exhibits simulation errors in continuous slope scenarios, and future iterations could expand to multi-terrain adaptive solutions.
  • The current test sample consists mainly of younger individuals with relatively uniform physical parameters; future studies should include more diverse groups.
  • The hardware (e.g., spring stiffness) is currently statically designed; future designs could incorporate adjustable stiffness for better adaptability.
  • Additional research and development are needed for derivative applications (e.g., downhill scenarios, free navigation across multiple environments).

Through this research, RedirectedStepper not only addresses some technical gaps in realistic walking simulation but also opens up new possibilities for future applications in VR training, entertainment, and health.