Locomotion is central to human navigation and spatial learning, with natural walking offering the most effective movement in immersive virtual environments (IVEs) due to its rich proprioceptive and vestibular feedback. However, space constraints often limit natural walking in IVEs, prompting the use of omnidirectional treadmills. The Cyberith Virtualizer Elite 2 enables users to walk in place by sliding their feet on a low-friction, tiltable platform while secured in a harness. It allows control over both tilt angle and walking speed, offering researchers flexibility in adjusting sensorimotor cues. In this study, we examined how tilt angle, speed, and training duration affect distance perception and cybersickness. In Experiment 1, we tested three tilt angles (0°, ~8.5°, ~17°) and three speeds (0.8×, 1.0×, 1.2× treadmill speed) during blind walking. A slower speed (0.8×) aligned best with natural walking and minimized cybersickness; users preferred the ~8.5° tilt. Additional analysis identified ~0.7× speed with ~8.5° tilt as preferred. In Experiment 2, we varied training durations (3, 6, or 9 minutes) and found that, with the preferred setup, treadmill distance judgments were comparable to natural walking. At least 6 minutes of training was sufficient for adaptation. These findings offer practical guidance for configuring treadmill locomotion in virtual reality research.

Multi-user virtual reality (VR) systems support multiple locomotion methods, but coordination may be challenging when partners use different methods. We studied dyads navigating a shared virtual maze using symmetric (same) or asymmetric (different) locomotion methods-specifically, steering (continuous movement with physical turning) and teleportation (point-and-teleport within 30 meters with physical turning)-and assessed survey knowledge using direction and straightline distance estimates between objects. We found that (1) dyads acquired better survey knowledge than individuals across locomotion conditions; (2) teleportation showed no significant difference from steering in distance error and yielded significantly lower angular error; (3) asymmetric locomotion methods did not significantly increase cybersickness compared to symmetric locomotion methods; and (4) symmetric and asymmetric dyads showed no significant differences in distance or angular errors. Overall, these results suggest that collaborative navigation improves spatial learning and that asymmetric locomotion can improve accessibility without significantly increasing cybersickness in multi-user VR applications.

Omnidirectional treadmills provide one solution for locomoting through large virtual environments in confined physical spaces. Through two experiments, this paper evaluated locomotion on an omnidirectional treadmill (Cyberith Virtualizer Elite 2) by comparing it to natural walking in an open physical space. In Experiment 1, participants judged distances and completed a path integration task using the treadmill and natural walking. Participants walked further on the treadmill but had larger angular errors during path integration, potentially due to increased cybersickness. Experiment 2 varied path lengths during path integration and found that longer paths led to higher cybersickness scores but did not affect performance. The paper offers interpretations and suggestions for using omnidirectional treadmills in virtual reality.

Consumer level virtual experiences almost always occur when physical space is limited, either by the constraints of an indoor space or of a tracked area. This observation coupled with the need for movement through large virtual spaces has resulted in a proliferation of research into locomotion interfaces that decouples movement through the virtual environment from movement in the real world. While many locomotion interfaces support movement of some kind in the real world, some do not. This paper examines the effect of the amount of physical space used in the real world on one popular locomotion interface, resetting, when compared to a locomotion interface that requires minimal physical space, walking in place. The metric used to compare the two locomotion interfaces was navigation performance, specifically, the acquisition of survey knowledge. We find that, while there are trade-offs between the two methods, walking in place is preferable in small spaces.

Navigation and wayfinding are often accomplished collectively, with groups of people. Yet most studies of navigation and wayfinding behavior, and of the acquisition of spatial knowledge more generally, focus on the individual. In this paper we extend the investigation of these topics to dyads. In particular, we focus on how well straight-line distances and directions between objects (survey knowledge) were learned by individuals and by the same individuals when comprising a dyad. Our experiment was carried out in a shared virtual environment and we report on the technical issues in conducting such a collaborative experiment in a shared virtual environment, such as the choice of locomotion mode and the provision of full-body self-avatars. Our findings indicate that dyads outperform individuals in their acquisition of survey knowledge.

Navigation, or the means by which people find their way in an environment, depends on the ability to combine information from multiple sources so that properties of an environment, such as the location of a goal, can be estimated. An important source of information for navigation are spatial cues generated by self-motion. Navigation based solely on body-based cues generated by self-motion is called path integration. In virtual reality and video games, many locomotion systems, that is, methods that move users through a virtual environment, can often distort or deprive users of important self-motion cues. There has been much study of this issue, and in this paper, we extend that study in novel directions by assessing the effect of four game-like locomotion interfaces on navigation performance using path integration. The salient features of our locomotion interfaces are that two are primarily continuous, i.e., more like a joystick, and two are primarily discrete, i.e., more like teleportation. Our main findings are that the perspective of path integration, people are able to use all methods, although continuous methods outperform discrete methods.

Locomotion in large virtual environments is currently unsupported in smartphone-powered virtual reality headsets, particularly within the confines of limited physical space. While motion controllers are a workaround for this issue, they exhibit known problems: they occupy the subject's hands, and they cause poor navigation performance. In this paper, we investigate three hands-free methods for navigating large virtual environments. The first method is resetting, a reorientation technique that allows for both translation and rotation body-based cues. The other two methods are walking in place techniques that use only rotation-based cues. In the first walking in place technique, we make use of the inertial measurement unit of the smartphone embedded in a Samsung Gear VR to detect when subjects are stepping. The second technique uses the Kinect's skeletal tracking for step detection. In this paper, we measure the survey component of spatial knowledge to assess three navigation conditions. Our metrics examine how well subjects gather and retain information from their environment, as well as how well they integrate it into a single model. We find that resetting leads to the strongest acquisition of survey knowledge, which we believe is due to the vestibular cues provided by this method.

Interesting virtual environments that permit free exploration are rarely small. A number of techniques have been developed to allow people to walk in larger virtual spaces than permitted by physical extent of the virtual reality hardware, and in this paper we compare three such methods in terms of how they affect presence and spatial awareness. In our first psychophysical study, we compared two methods of reorientation and one method of redirected walking on subjects' presence and spatial memory while navigating a pre-specified path. We further compared the two reorientation methods in a second psychophysical study involving free exploration and navigation in a large virtual environment. Our results provide criteria by which the choice of a locomotion method for navigating large virtual environments may be selected.

In this study, we examine subjects' spatial orientation in a head--mounted display (HMD) based virtual environment (VE) when objects are placed at different vertical viewing angles. More specifically, we examine spatial orientation when subjects memorize the location of objects on the ground, at eyelevel, and up in the air. Interestingly, we found that spatial orientation was the same across all conditions. This finding is important because it informs researchers who use similar spatial orientation tasks that variations in viewing neck angle to targets is not necessarily an issue.

This experiment investigates the spatial memory and attention when human acts as supervisor of one or two groups of distributed robot teams in a large virtual environment (VE). The problem is similar to learning a new environment and interpreting its spatial structure, e.g., [Mou and McNamara 2002], but less is known when attention is divided or locomotion is involved. Our motivation arises in the context of humans and robots acting cooperatively together as a team. Such teaming is becoming increasingly important in many scenarios, such as disaster relief, and wilderness search and rescue [Humphrey and Adams 2009].

When we locomote through real or virtual environments, self-to-object relationships constantly change. Nevertheless, in real environments we effortlessly maintain an ongoing awareness of roughly where we are with respect to our immediate surrounds, even in the absence of any direct perceptual support (e.g., in darkness or with eyes closed). In virtual environments, however, we tend to get lost far more easily. Why is that? Research suggests that physical motion cues are critical in facilitating this “automatic spatial updating” of the self-to-surround relationships during perspective changes. However, allowing for full physical motion in VR is costly and often unfeasible. Here, we demonstrated for the first time that the mere illusion of self-motion (“circular vection”) can provide a similar benefit as actual self-motion: While blindfolded, participants were asked to imagine facing new perspectives in a well-learned room, and point to previously-learned objects. As expected, this task was difficult when participants could not physically rotate to the instructed perspective. Performance was significantly improved, however, when they perceived illusory self-rotation to the novel perspective (even though they did not physically move). This circular vection was induced by a combination of rotating sound fields (“auditory vection”) and biomechanical vection from stepping along a carrousel-like rotating floor platter. In summary, illusory self-motion was shown to indeed facilitate perspective switches and thus spatial orientation. These findings have important implications for both our understanding of human spatial cognition and the design of more effective yet affordable VR simulators. In fact, it might ultimately enable us to relax the need for physical motion in VR by intelligently utilizing self-motion illusions.

The use of vection, the illusion of self-movement, has recently been explored as a novel way to immerse observers in mediated environments through illusory yet compelling self-motion without physically moving. This provides advantages over existing systems that employ costly, cumbersome, and potentially hazardous motion platforms, which are often surprisingly inadequate to provide life-like motion experiences. This study investigates whether spatialized sound rotating around the stationary, blindfolded listener can facilitate biomechanical vection, the illusion of self-rotation induced by stepping along a rotating floor plate. For the first time, integrating simple auditory and biomechanical cues for turning in place evoked convincing circular vection. In an auditory baseline condition, participants experienced only spatialized auditory cues. In a purely biomechanical condition, seated participants stepped along sideways on a rotating plate while listening to mono masking sounds. Scores of the bi-modal condition (binaural+biomechanical cues) exceeded the sum of both single cue conditions, which may imply super-additive or synergistic effects.