Data glove

Data glove
Basic Info
VR/AR	Virtual Reality, Augmented Reality
Type	Input device
Subtype	Hand-worn finger and hand motion sensor
Platform	Varies by model
Creator	Electronic Visualization Laboratory (Sayre Glove, 1977); VPL Research (DataGlove, 1987)
Developer	Various
Manufacturer	Various
Announcement Date	1977 (Sayre Glove)
Release Date	1987 (VPL DataGlove, first commercial model)
Price	Varies (consumer to enterprise)
System
Storage
Display
Display	N/A
Resolution	N/A
Refresh Rate	N/A
Image
Field of View	N/A
Optics
Optics	N/A
Passthrough	N/A
Tracking
Tracking	Finger flexion sensing plus hand position and orientation tracking
Eye Tracking	N/A
Face Tracking	N/A
Hand Tracking	Yes (worn on the hand)
Body Tracking	N/A
Rotational Tracking	Yes (hand orientation, model dependent)
Positional Tracking	Yes (hand position, model dependent)
Audio
Audio	N/A
Microphone	N/A
Camera	N/A
Connectivity
Connectivity	Wired or wireless, depending on model
Device
Haptics	Yes on haptic models (vibrotactile and force feedback); none on tracking-only models
Sensors	Flex sensors (fiber-optic, resistive, capacitive, or magnetic), IMU, position tracker
Input	Finger bend, hand pose, often buttons or a directional pad

Property "Type" (as page type) with input value "Input Devices|Input device" contains invalid characters or is incomplete and therefore can cause unexpected results during a query or annotation process. <ul><li>Property "Creator" (as page type) with input value "Electronic Visualization Laboratory]] (Sayre Glove" contains invalid characters or is incomplete and therefore can cause unexpected results during a query or annotation process.</li> <!--br--><li>Property "Creator" (as page type) with input value "1977); VPL Research (DataGlove" contains invalid characters or is incomplete and therefore can cause unexpected results during a query or annotation process.</li></ul>

A data glove, also called a wired glove or dataglove, is an input device worn on the hand that senses the bending of the fingers and, on most models, the position and orientation of the hand. By measuring finger flexion and hand pose, a data glove lets a user reach into and manipulate a virtual or augmented environment with natural hand motion rather than a mouse, keyboard, or gamepad. Data gloves were among the first devices built specifically for Virtual Reality and remain in use for motion capture, enterprise training, and research, though optical Hand tracking and tracked controllers have largely replaced them for mainstream consumer VR.

The first data glove, the Sayre Glove, was built in 1977 at the Electronic Visualization Laboratory in Chicago. The device that brought the concept to commercial VR was the VPL DataGlove, developed by Thomas Zimmerman and Jaron Lanier and marketed from 1987. A simplified consumer adaptation, the Nintendo Power Glove, reached homes in 1989 but failed in the market. Modern descendants range from precision motion-capture gloves to force-feedback haptic gloves used in industrial and enterprise applications.

How it works

A data glove combines two kinds of sensing. The first measures how much each finger is bent, using flex sensors mounted along the back of the fingers. The second measures where the hand is in space and how it is rotated, using a separate position-and-orientation tracker. Software reads both streams and reconstructs a pose for a virtual hand or skeleton in real time.

Several flex-sensing technologies have been used over the history of the device:

• Optical (light-based and fiber-optic). The earliest gloves shone light down a flexible tube or fiber and measured how much reached a photocell at the far end; bending the finger reduced the transmitted light. The Sayre Glove used flexible light-piping tubes, and the VPL DataGlove used optical fibers scratched near each joint so that bending leaked light in proportion to the bend angle.
• Resistive (bend sensors and conductive ink). Many later and cheaper gloves use strips whose electrical resistance changes as they flex. The Power Glove used carbon-based resistive ink sensors to detect roughly four bend positions per finger.
• Capacitive and stretch sensors. Some gloves use stretchable capacitors whose capacitance changes with bending, shear, and pressure.
• Magnetic fingertip tracking. Recent professional gloves such as the Manus Quantum Metagloves place a magnetic source on the back of the hand and a small module on each fingertip, sensing the fingertips within the magnetic field for drift-free per-finger tracking.

Hand position and orientation are usually supplied by a magnetic tracker (historically a Polhemus or similar unit), by ultrasonic transmitters and receivers, by an onboard inertial measurement unit, or by optical tracking of the glove. Haptic gloves add a third element, actuators that push back on the skin or restrict the fingers to simulate touching and grasping virtual objects.

History

The Sayre Glove (1977)

The Sayre Glove is generally credited as the first data glove. It was designed in 1977 at the Electronic Visualization Laboratory (EVL) at the University of Illinois at Chicago by Daniel Sandin, Thomas DeFanti, and Richard Sayre, under a grant from the National Endowment for the Arts. The glove ran flexible tubes along the fingers with a light source at one end and a photocell at the other; bending a finger changed how much light reached the photocell, giving a continuous measure of flexion. It was lightweight and inexpensive and was used mainly to drive on-screen sliders.

The VPL DataGlove (1987)

Thomas Zimmerman built a prototype optical-flex-sensor glove and filed a U.S. patent on the optical flex sensor in the early 1980s. In 1984 Zimmerman met Jaron Lanier, who saw the device's potential for virtual environments; the two founded VPL Research (the initials stood for Virtual Programming Languages) and developed the device into the DataGlove. Engineer Young Harvill refined the optical sensor by scratching the fiber near each finger joint so it became locally sensitive to bending.

VPL brought the DataGlove to market in 1987, and that October it appeared on the cover of Scientific American. The glove used fiber-optic flex sensors to read finger bend and a magnetic tracker to read hand position and orientation, feeding the data to a host computer so a virtual hand could mirror the user's movements. A complete VPL DataGlove cost on the order of $10,000, putting it within reach of research labs and companies rather than consumers. VPL sold the DataGlove alongside its VPL EyePhone head-mounted display and a full-body DataSuit, assembling some of the first commercial Virtual Reality systems. VPL Research filed for bankruptcy in 1990, and its patents were later acquired by Sun Microsystems.

The Nintendo Power Glove (1989)

The Power Glove was a consumer controller for the Nintendo Entertainment System derived from VPL's DataGlove technology. It was designed by Grant Goddard and Samuel Cooper Davis at Abrams/Gentile Entertainment, with input from Zimmerman and Lanier, and manufactured by Mattel in North America (and by PAX in Japan); Nintendo licensed but did not design it. To hit a consumer price, the Power Glove replaced the DataGlove's costly fiber-optic sensors with carbon-based resistive flex sensors that resolved only about four bend positions for each of four fingers, and it tracked hand position with ultrasonic transmitters on the glove and three receivers mounted around the television. It launched in North America in October 1989 at about $75 and sold close to one million units, but it was a critical failure because of imprecise tracking, awkward calibration, and a near-total lack of dedicated software; only two games, Super Glove Ball and Bad Street Brawler, supported it directly.

The CyberGlove and research gloves (1990s onward)

In 1990 Virtual Technologies, Inc. introduced the CyberGlove, an instrumented glove aimed at research and professional use; it became a long-running standard for hand motion capture. Virtual Technologies was acquired by Immersion Corporation in September 2000, and in 2009 the CyberGlove product line was spun off to a new company, CyberGlove Systems LLC, which continued its development. Other research and prosumer gloves followed over the 1990s and 2000s, including the 5DT Data Glove line and the consumer-oriented P5 glove released by Essential Reality in 2002.

Modern data and haptic gloves

Contemporary gloves split into two broad groups: tracking-focused gloves that capture precise finger and hand motion, and haptic gloves that also feed sensation back to the wearer.

On the tracking side, the Dutch company Manus makes professional gloves for motion capture and Virtual Reality. Its Quantum Metagloves, shown at GDC 2022 and priced around $9,000 a pair, use magnetic fingertip sensors for drift-free per-finger tracking and stream data into Unity and Unreal Engine through the company's software; they capture finger motion but do not provide force feedback.

On the haptic side, several companies build gloves that physically resist or stimulate the fingers. HaptX announced its HaptX Gloves G1 in October 2022 and began shipping in 2023; each glove uses 135 microfluidic actuators that displace the skin to simulate touch, with up to about 9 psi of peak pressure and 1.5 mm of displacement, fed by a roughly 17-pound AirPack worn as a backpack. The gloves are sold to businesses for $5,495 a pair (or $4,500 a pair in a four-size bundle) plus a subscription starting near $495 a month. The Dutch firm SenseGlove sells the Nova 2, a wireless force-feedback glove that uses four magnetic brakes to resist the fingers with up to about 20 N of force each and adds vibrotactile cues, including a palm-contact element; it weighs about 350 g per glove, runs roughly three hours on a charge, and is marketed to enterprise customers at around $6,000 a pair. Other current or recent products in this space include the earlier SenseGlove Nova, the Manus Prime 3 Haptic XR, the Diver-X ContactGlove, and the Teslasuit Glove.

Meta has publicly described long-term research toward consumer force-feedback gloves, framing them as a way to make virtual training, handshakes, and object manipulation feel physical, but has not shipped a product; functional force-feedback gloves remained enterprise items costing several thousand dollars as of 2026.

Relationship to optical hand tracking and controllers

For mainstream consumer VR, data gloves were displaced by two newer approaches. Tracked controllers supply buttons, thumbsticks, and basic vibration with reliable tracking, while camera-based optical Hand tracking, as found on standalone headsets, reconstructs the hands from cameras with nothing worn at all.

Each approach trades off differently. Optical hand tracking requires no hardware on the hand and feels natural, but it fails when the hands leave the camera's view or occlude each other, and it cannot deliver haptic feedback because nothing touches the skin. Data gloves keep working regardless of camera view and can carry haptic actuators, but they must be put on and calibrated and, for force-feedback models, remain expensive and bulky. A 2024 study in Frontiers in Virtual Reality that compared a VR glove against motion controllers found that participants reported higher presence, embodiment, and cognitive absorption with the glove, which it attributed to the glove's more natural perceived movement; the authors also noted that without haptic feedback a glove is not a perfect substitute for real-world manipulation. Earlier work had reached more mixed conclusions, with one study finding that a VR glove did not necessarily yield a better user experience than standard controllers. These trade-offs help explain why data gloves today are concentrated in motion capture, simulation, and enterprise training rather than general consumer use.

See also

• Hand tracking
• Haptics
• Input Devices
• VPL Research
• Jaron Lanier
• Power Glove
• HaptX
• SenseGlove Nova
• Manus Prime 3 Haptic XR
• Teslasuit Glove