
Today, four main directions can be identified in the development of omnidirectional walking platforms for user locomotion in virtual reality (VR locomotion platforms). Most existing solutions originate from scientific research conducted by specialists at the VR Lab of the University of Tsukuba (Japan), which has had a significant influence on the development of virtual locomotion technologies.
Below, we will sequentially review each direction of omnidirectional VR platform development, their current state, and the reasons why most such devices still fail to provide comfortable, realistic, and physiologically correct user movement in virtual reality.
At the end of this page, we have prepared a curated selection of YouTube videos demonstrating the operation of all devices discussed here, presented in the same order as they appear in the text.
1. Omni-Directional Treadmills
The first category includes devices that structurally resemble conventional treadmills but feature additional degrees of freedom, allowing the user to move not only forward and backward but also sideways. The working surface of such platforms is typically divided into segments or modules capable of changing orientation, while the entire system dynamically returns the user to the central area of the working zone.
The foundations of this principle were established in the Torus Treadmill and Omni-Directional Treadmill (ODT), introduced in 1997.
The Torus Treadmill was developed at the University of Tsukuba (Japan), while the ODT was created by American engineers from Virtual Space Devices. A detailed description of this design is provided in U.S. Patent US5562572A, “Omni-Directional Treadmill.”


One of the largest commercial attempts to develop this direction was the European CyberWalk project. Its modern continuation can be considered the INFINADECK platform, which represents a reduced and technologically refined version of such systems, aimed at professional and industrial applications.
Limitations and Drawbacks
Key disadvantages of omnidirectional treadmills of this type include:
Attempts to reduce the size of such devices were implemented in the Cybercarpet and StriderVR projects; however, they did not eliminate the fundamental limitations of this principle. In both cases, a conventional treadmill mechanism is supplemented by a ball-bearing layer:
Despite implementation differences, the core operating principle and its associated limitations remain the same.
Disney Holotile
Within the category of omni-directional treadmills, the Disney Holotile deserves special attention. In this device, the walking surface is formed by an array of individually driven mechanized rollers that dynamically compensate for user movement in real time and return the user to the center of the platform.
This approach allows step-based movement in any direction while remaining within the working area. The concept is described in detail in U.S. Patent US20180217662A1 (2017), “Floor system providing omnidirectional movement of a person walking in a virtual reality environment.”
Despite its technological sophistication and originality, this approach retains the typical treadmill-related limitations, including high mechanical complexity, cost, and difficulty in implementing natural movement scenarios.
Realism Limitations
An additional significant drawback of all devices in this category is the inability to realistically implement stair climbing, descending, or complex movement over uneven terrain. This significantly reduces immersion and limits the applicability of such platforms in realistic VR scenarios.


2. Platforms with Sliding and Antifriction Effect
The second direction is based on the use of friction reduction effects.
Such systems employ:
• rollers mounted on the soles of the user’s footwear;
• rollers or ball-bearing elements on the platform;
• special antifriction coatings.
This approach originates from the experimental project "Virtual Perambulator," presented in 1995 at the SIGGRAPH conference by researchers from the University of Tsukuba. It was one of the earliest VR locomotion interfaces designed to simulate the user's walking. The device utilized a smooth plastic surface and slippery shoes, allowing the feet to slide freely during stepping movements.
The next stage in the development of this direction was the Omni-direction Ball-bearing Disc Platform (OBDP) (1999), which utilized a large number of ball-bearing rollers placed directly on the support surface.
Modern Analogues
Modern commercial devices of this type (Virtuix Omni, Wizdish ROVR, WalkMouse, Cyberith Virtualizer, Kat Walk, etc.) differ only marginally from their early prototypes. Improvements mainly concern safety and harness systems and do not increase walking realism.
The fundamental problem lies in the sliding principle itself: the user is forced to “ride” their feet across the surface, which subjectively resembles walking on ice and does not correspond to natural human biomechanics.



3. Powered Footwear (Powered Shoes)
The third direction is based on the use of electrically driven rollers mounted directly on the user’s footwear. The first such device was Powered Shoes, developed at the VR Lab, University of Tsukuba.
Modern analogues, including the Freeaim VR Shoes project, have not fundamentally advanced beyond the original concept and continue to suffer from several objective problems:
• impacts between footwear elements during walking;
• high noise levels;
• significant shock loads on the joints;
• inability to implement stair climbing and movement over uneven terrain.
Placing lifting mechanisms directly on the user’s legs makes such systems bulky and reduces their practical applicability.


4. Platforms with Independent Foot Platforms
The fourth direction includes devices with separate platforms for each foot. Most solutions of this type existed only as scientific publications and prototypes.
One of the earliest developments of this type was described in the work “A foot following locomotion device with force feedback capabilities.” The design consists of a platform with two movable arms and four degrees of freedom; due to significant operational loads, the device requires the use of high-power actuators, which constitutes its key drawback.
Devices belonging to this direction also include Gait Master systems:
• Gait Master 1 — two movable foot platforms mounted on telescopic mechanisms;
• Gait Master 2 and Gait Master 5 — described in scientific publications of the VR Lab, University of Tsukuba. The second and fifth versions were developed for use in medical rehabilitation.
Devices Outside the Classification
CirculaFloor and VirtuSphere should be mentioned separately, as they do not belong to any of the listed directions.
CirculaFloor is a system consisting of multiple robotic platforms that are sequentially positioned under the user’s feet.
VirtuSphere is a transparent rotating sphere with a diameter of approximately 2 meters, inside which the user is located.
The operating principle of VirtuSphere was first described in patent RU2109336C1 (1995)
“Method of immersing a user in virtual reality and a device for its implementation.”
Conclusion
After studying and testing existing omnidirectional walking devices for VR, we concluded that none of the available solutions simultaneously provides a high level of realism, comfort, and safety.
We are confident that the Radix Omnidirectional Treadmill, fundamentally different from existing analogues, is capable of advancing VR locomotion technologies to a new level. The concept embedded in this device allows significant simplification of the design, reduction of industrial production costs, and makes realistic virtual locomotion accessible to a broader audience.


