Traceable teleportation: Improving spatial learning in virtual locomotion
Ye Jia, Zackary P. T. Sin, Chen Li, Peter H. F. Ng, Xiao Huang, George Baciu, Jiannong Cao, Qing Li
International Journal of Human-Computer Studies (2025)
Figure 1: Comparison between Classical Point-and-Teleport (top) and Traceable Teleportation (bottom), showing the Undo-Redo mechanism and Visualized Pathing features that enhance spatial orientation in virtual environments.
Abstract
In virtual reality, point-and-teleport (P&T) is a locomotion technique that is popular for its user-friendliness, lowering workload and mitigating cybersickness. However, most P&T schemes use instantaneous transitions, which has been known to hinder spatial learning. While replacing instantaneous transitions with animated interpolations can address this issue, they may inadvertently induce cybersickness. To counter these deficiencies, we propose Traceable Teleportation (TTP), an enhanced locomotion technique grounded in a theoretical framework that was designed to improve spatial learning. TTP incorporates two novel features: an Undo-Redo mechanism that facilitates rapid back-and-forth movements, and a Visualized Path that offers additional visual cues. We have conducted a user study via a set of spatial learning tests within a virtual labyrinth to assess the effect of these enhancements on the P&T technique. Our findings indicate that the TTP Undo-Redo design generally facilitates the learning of orientational spatial knowledge without incurring additional cybersickness or diminishing sense of presence.
The Teleportation Tradeoff, Revisited
Point-and-teleport (P&T) has become the default locomotion technique in consumer VR for good reason: it's easy to learn, minimizes cybersickness, and works within limited physical space. But it has a well-documented cost — instantaneous transitions destroy spatial learning. Users who teleport through a virtual environment build poorer mental maps than users who walk or steer continuously, because the abrupt viewpoint shifts disrupt path integration, the cognitive process by which we track our position relative to known landmarks. Animated interpolations (smooth transitions between teleport points) can help, but they reintroduce the optical flow that causes cybersickness, undermining the primary reason to use teleportation in the first place.
Traceable Teleportation (TTP) takes a different approach. Instead of trying to make teleportation feel more continuous, it provides cognitive tools that help users reconstruct spatial relationships despite discontinuous movement.
Two Features, One Goal
TTP adds two mechanisms to standard P&T:
Undo-Redo: a rapid back-and-forth movement capability. If a user teleports from point A to point B and realizes they're disoriented, they can instantly return to A, re-establish their bearings, and then re-teleport to B (or to a different target). This is not just a convenience feature — it directly supports spatial learning by enabling what the authors call "spatial hypothesis testing." The user can say "I think the exit is behind that wall" (teleport to check), "no, that's the wrong corridor" (undo back to previous position), and build a mental map through iterative verification. Each undo-redo cycle is a miniature spatial experiment.
Visualized Path: a persistent visual trace showing the user's teleportation history — where they've been, in what order, and how those locations relate spatially. The path is rendered as a line or trail in the virtual environment, visible at all times. This provides an external spatial reference that compensates for the disrupted internal path integration. Even though you didn't walk from A to B to C, you can see the line connecting them, and that visual trace supports the same spatial reasoning that continuous movement would have provided.
The Theoretical Framework
The paper grounds TTP in a spatial learning framework that distinguishes three types of spatial knowledge: landmark knowledge (what is where), route knowledge (how to get from A to B), and survey knowledge (how everything relates in a unified spatial model). Standard P&T impairs route and survey knowledge because it severs the temporal continuity between locations. TTP's Undo-Redo targets route knowledge (by enabling repeated traversal of the same path segment) while Visualized Path targets survey knowledge (by providing an overhead-like view of movement history). The features are complementary, not redundant, and the study was designed to test their independent and combined effects.
The Labyrinth Study
Participants navigated a virtual labyrinth — a controlled spatial environment where route complexity and landmark placement can be systematically varied. They performed spatial learning tasks after navigating with either standard P&T or TTP, including orientation judgments (pointing toward unseen landmarks) and route retracing. The labyrinth is a well-chosen testbed: it's spatially demanding enough to reveal differences between locomotion techniques, but structured enough to quantify spatial knowledge precisely.
The key finding is that TTP's Undo-Redo mechanism improved orientational spatial knowledge — participants who could undo and redo their teleports were better at pointing toward unseen locations, a standard measure of survey knowledge. This improvement came without any increase in cybersickness scores and without any decrease in presence. In other words, TTP added a cognitive benefit without paying the costs (sickness, presence loss) that typically accompany locomotion improvements.
The Visualized Path showed more modest effects, which the authors interpret as a ceiling issue: in a labyrinth with relatively simple geometry, the visual path may have been less necessary because the environment itself provided sufficient spatial cues. In more complex or visually uniform environments, the Visualized Path's contribution would likely be larger.
Where This Fits in the Locomotion Landscape
TTP is the pragmatic counterpart to the research group's own illumotion technique (IEEE VR 2024). illumotion proposes a fundamentally new locomotion paradigm based on optical manipulation; it's ambitious and novel, but requires users to learn a new interaction model. TTP, by contrast, enhances a technique that users already know and platforms already implement. The Undo-Redo mechanism could be added to any existing P&T implementation with minimal engineering — it's essentially a position stack with a UI trigger. The Visualized Path requires slightly more work (rendering a persistent trail in 3D space) but is still a standard graphics operation.
The two papers together represent a dual strategy for advancing VR locomotion: innovate new techniques (illumotion) while also fixing the ones people actually use (TTP). The latter may have more near-term impact simply because the adoption barrier is lower.
Boundaries
The labyrinth task tests spatial learning at a single scale (room-to-building sized environment) and a single session. Whether TTP's benefits persist in larger environments (city-scale), longer sessions (hours rather than minutes), or more complex navigation tasks (multi-story buildings, environments with moving elements) is untested. The Undo-Redo mechanism may also have a ceiling: excessive undoing and redoing could become disorienting in its own right, essentially creating a different kind of viewpoint discontinuity. The study doesn't identify the optimal undo-redo frequency for spatial learning.
The user study compared TTP against standard P&T but did not include a continuous steering condition. This means we can't quantify how much of the spatial learning deficit TTP recovers relative to the gold standard of smooth locomotion. TTP improves spatial learning relative to standard P&T — but we don't know whether it closes the gap with steering entirely, partially, or barely at all. That comparison is the natural follow-up.