Steering Versus Teleporting in Mobile Virtual Reality

1 Introduction

• MVR can leverage widely adopted smartphones and costs $10–60 USD, compared with $300–3500 USD for standalone VR headsets.The authors frame affordable hardware as a way to improve access to immersive experiences.Source: §1 · p. 1
• Primitive MVR input—including 3DOF head tracking and inconsistent discrete controls—limits comparable immersive experiences and 3D interaction.MVR devices rarely provide the tracked controllers commonly bundled with standalone VR headsets.Source: §1 · p. 1
• The study compares gaze- and controller-directed steering and teleportation using a low-cost tracked controller, acoustic button, and controller-pitch speed control.It evaluates whether travel techniques and velocity control can work on low-fidelity MVR hardware.Source: §1 · p. 2

2.1 Steering-Based Travel

• Steering supports movement unconstrained by real-world space, but MVR steering often lacks speed control and relies on automated motion or start/stop input.Controller tilt-based speed control is introduced as a way to provide continuous, omnidirectional control on MVR hardware.Source: §2.1 · p. 2

2.2 Selection-Based Travel

• Teleportation is an efficient selection-based travel technique, while prior MVR studies and input constraints leave controller-based teleportation open to further exploration.Prior work describes teleportation as avoiding cybersickness and environment-size constraints, though it may disorient users.Source: §2.2 · p. 2

2.3 Comparing Teleportation and Steering

• Field of view can shape travel strategies: prior work found gaze-directed travel better suited to small-FOV systems and controller pointing to CAVE environments.Earlier comparisons also report mixed results across teleportation and steering outcomes.Source: §2.3 · p. 2
• No prior work had examined how travel technique affects MVR performance, user strategy, and preference in navigating large virtual environments.The study addresses whether advantages reported for higher-fidelity VR hold under MVR constraints.Source: §2.3 · p. 2

2.4 Locomotion Performance Evaluation

• Trajectory sampling can reveal steering behaviour and path conformity over time beyond cumulative measures such as completion time or total distance.The paper notes that trajectory-based locomotion evaluation remains underexplored and that waypoint tasks may not suit such analyses.Source: §2.4 · p. 3

3 System Design

• The Cardboard Controller uses camera-based 6DOF tracking and an acoustic button; click detection filters audio and can still be falsely triggered by loud speech or door slams.Controller orientation ignores yaw changes above 12° and inversions above 90° to limit jitter and awkward movement.Source: §3 · p. 3, Figure 2 · p. 3

3.1 Steering

• Controller pitch controls continuous steering speed: ±10° stops movement, forward tilt reaches 3 m/s at 60°, and backward tilt reaches the same maximum at -40°.On tracking loss, speed and controller-directed steering direction remain constant for 1.5 seconds before slowing to a stop; backward movement begins slowing after 200 ms.Source: §3.1 · pp. 3–4, Figure 3 · p. 4

3.2 Teleportation

• Teleportation uses an aiming-direction projectile arc, accepting ground collisions as destinations and rejecting walls or decorative objects.Pressing the button at a valid position teleports instantly; the maximum range is approximately 20 m when aimed 45° above horizontal.Source: §3.2 · p. 4

3.3 Pilot Study

• In the eight-person pilot, pitch control had fewer mean wall collisions (2.5 vs. 3.1) but lower mean speed (1.65 vs. 2.98 m/s); coin collection was comparable (50% vs. 48%), and pitch was selected for the main study.Six of eight participants preferred pitch, with several valuing speed control for avoiding walls and stopping for coins.Source: §3.3 · p. 4

4.1 Participants

• The experiment used an open-source testbed and recruited 28 right-handed participants aged 18–35, all with prior experience using 3D-tracked pointing devices.Three participants had no prior VR experience, and one had used only cardboard VR.Source: §4 · p. 4, §4.1 · p. 4

4.2 Apparatus

• A Samsung Galaxy S23 Ultra in a VRG Pro headset provided 3DOF head orientation and an approximately 80° horizontal field of view, with the cardboard controller as the pointing device.The experiment had ethics-board approval.Source: §4.2 · p. 4, §4 · p. 4

4.2.1 Virtual Environment.

• Four condition-specific, single-directional paths were approximately 735–780 m long, with 64 alleyways and 67 coins each; participants collected coins by overlapping a gaze-controlled cursor for 50 ms.Temples and alley walls supported collisions, while turn arrows served as wayfinding cues.Source: §4.2.1 · p. 5, Figure 5 · p. 5, Figure 7 · p. 5

4.4 Procedure

• Participants practiced each travel technique on a separate path, then were instructed to finish quickly while avoiding collisions, collecting all coins, and staying near the path centre.The 60-minute procedure included sickness ratings during each condition and questionnaires during breaks.Source: §4.2.1 · p. 5, §4.4 · pp. 5–6

4.2.2 Task.

• The task recorded collisions with nature elements, temples, and exterior walls, counting continuous contact as one collision; steering positions were sampled every 5 seconds, while every teleport endpoint was logged.The distinct sampling intervals reflected the continuous structure of steering and discrete structure of teleportation, as well as smartphone logging load.Source: §4.2.2 · p. 5

4.3 Design

• A 2 × 2 within-subjects design crossed gaze- versus controller-directed control with steering versus teleportation, and measured travel, collision, coin, teleport, and subjective outcomes.Performance measures were recorded every 25 m and at the finish; the study also collected UEQ-S and Fast Motion Sickness Scale responses.Source: §4.3 · p. 5

4.5 Hypotheses

• The hypotheses predicted teleportation advantages in travel time, collisions, and path efficiency; gaze-directed teleportation preference; and controller-directed steering preference.They also predicted more coin collection while steering, higher presence for controller-directed steering, and greater physical demand for gaze-directed steering.Source: §4.5 · p. 6

5.1 Performance Metrics

• Gaze controls averaged 15.21 s versus 17.93 s for controller controls; CT had higher travel times than every other condition.Source: §5.1 · p. 6

5.1.2 Effective Speed.

• Teleport conditions averaged 0.86 m/s faster; GS and CS were significantly slower than both teleport conditions.Source: §5.1.2 · p. 6

5.1.6 Path Efficiency.

• Steering produced less efficient paths than teleportation; PEI compares travelled path length with ORP length, and values above 1 denote longer routes.Source: Figure 12 · p. 7, §5.1.6 · p. 7

5.1.5 Teleport Count.

• GT participants teleported more often than CT participants (M = 2.88, SD = .6 vs. M = 2.2, SD = .13).Source: Figure 11 · p. 7, §5.1.5 · p. 7

5.1.3 Wall Collisions.

• Steering produced 0.44 wall collisions per 25 m versus 0.08 with teleportation; overall, collisions stayed below one per 25 m.Source: §5.1.3 · pp. 6–7

5.1.4 Coins Collected.

• Participants collected 79–89% of 67 coins overall; steering yielded fewer coins, and gaze yielded 9.55 more than the controller.Source: §5.1.4 · p. 7

5.2.1 Fast Motion Sickness Scale.

• FMS scores and within-condition FMS change did not differ significantly across conditions.Source: §5.2.1 · p. 7

5.2.2 Questionnaires.

• GT received significantly higher preference rankings; CS exceeded GS descriptively but had polarized rankings, with 11 last-place rankings versus 7.Source: Figure 15 · p. 8, §5.2.2 · p. 8
• Button and controller ratings of 3.07 and 3.79, respectively, suggested both somewhat hindered performance.Source: §5.2.2 · pp. 7–8
• GT had the highest UEQ-S scores on all scales; CS usability was moderate, while participants noted tracking problems and arm fatigue.Source: §5.2.2 · p. 8
• GS and CT drew concerns about discomfort, restricted movement, and reacquiring the controller after looking away.Source: §5.2.2 · p. 8
• GT was less physically demanding than all other conditions; CT presence was lower than GT and CS, while CS presence exceeded GS.Source: §5.2.2 · p. 8, Figure 16 · p. 8

6.1 Teleport Control Scheme and User Strategies

• GT outperformed and was preferred to CT, but CT teleports were slower and more precise; short-range aiming remained unintuitive despite training.Source: §6.1 · pp. 8–9

6.2 Comparison of Teleportation and Steering

• CT's interval-level effective speed and cumulative travel time yielded different performance interpretations, warranting further investigation.Source: §6.2 · p. 9
• No significant cybersickness effects emerged, though 71% identified gaze as the main source of nausea; the authors suggest control type may matter more than travel technique.Source: §6.2 · p. 9
• A descriptive comparison with high-fidelity VR found several broadly aligned metrics, but hardware differences mean a direct controlled comparison is needed.Source: §6.2 · p. 9

6.3 Design Guidelines for Locomotion in MVR

• Guidelines favor gaze for teleportation and tracked handheld controllers for decoupled steering, although tracking instability and arm fatigue polarized CS preferences.Source: §6.3 · p. 9
• Tilt-based speed control was described as intuitive, and six of eight pilot participants preferred it to automated speed control.Source: §6.3 · p. 9

6.4 Potential of Lo-Fi Devices for Mobile VR

• The audio-based button supported teleportation, while controller pointing suited steering; camera-FOV tracking loss remained a common drawback.Source: §6.4 · pp. 9–10

6.5 Limitations

• Unequal trajectory sampling—steering every 5 s versus every teleport endpoint—may affect trajectory-level measures and limits direct comparisons.Source: §6.5 · p. 10

• Gaze-directed teleportation produced the best overall results, whereas controller-directed control was better suited to steering.Source: §7 · p. 10
• The study used a low-cost tracked controller and audio-based button to compare free-range teleportation with speed-controlled steering.Source: §7 · p. 10
• Controller-directed steering was broadly comparable to similar high-fidelity techniques, and participants valued its flexibility and intuitiveness.Source: §7 · p. 10