Toggle search
Toggle menu
5.1K
5.9K
229
50.6K
VR & AR Wiki
Toggle preferences menu
Toggle personal menu
Not logged in
Your IP address will be publicly visible if you make any edits.

Haptic rendering

From VR & AR Wiki
More actions

Haptic rendering is the process of computing the forces, vibrations or other touch stimuli that a user should feel when interacting with a virtual or remote object, and of commanding a haptic device to display them. In a 2004 survey in IEEE Computer Graphics and Applications, Kenneth Salisbury, Francois Conti and Federico Barbagli described rendering as "the process by which desired sensory stimuli are imposed on the user to convey information about a virtual haptic object", and compared it with visual rendering: a sphere rendered haptically with a simple penalty function feels different from the same sphere rendered with techniques that also convey texture and surface friction.[1]

Haptic rendering is the touch counterpart of graphics and audio rendering in virtual reality (VR) and augmented reality (AR) systems. It differs from them in one basic way: the haptic channel exchanges energy and information in both directions, from the computer to the user and from the user back to the computer. Salisbury and colleagues wrote that this bidirectionality "is often referred to as the single most important feature of the haptic interaction modality."[1] Much of the foundational research used desktop force-feedback arms such as the PHANToM from SensAble Technologies, and produced algorithms such as the god-object and proxy methods, running in a servo loop that commonly updates at about 1 kHz.[1][2] In consumer VR, the OpenXR standard and platform tools such as Meta's Haptics SDK expose haptics mainly as controller vibration, while research and professional systems also render touch through haptic gloves, ultrasound arrays and grounded force-feedback devices.[3][4][5] Broader coverage of touch technology is in the Haptics article.

Reviewed 13 October 2026. Claims checked against the full texts of the 1995, 1997, 1999 and 2004 papers, Crossref metadata and abstracts of the other papers, and the cited OpenXR, Meta, CHAI3D, patent and TechCrunch pages. About review dates.

Definition and scope

Salisbury, Conti and Barbagli used "haptics" for touch interactions that occur for perception or manipulation of objects, whether between a human hand and a real object, a robot and a real object, or a human hand and a simulated object through a haptic interface device.[1] They separated two kinds of touch information. Tactile information comes from sensors in the skin and describes the distribution of pressure across the contact area; kinesthetic information comes from sensors in the joints. Real interaction forces are normally perceived through a combination of the two.[1]

The user's representation inside the virtual environment is called the avatar. Its shape depends on the application and on the device: a surgical tool can be simulated as a full volumetric object exchanging forces and positions in six dimensions, or as a single point at the tool's tip exchanging forces in three.[1] The number of degrees of freedom (DOF) used to describe the interaction force usually matches the number of actuated DOF of the device.[1]

Haptic interface devices are often described as small robots that exchange mechanical energy with the user. Most are impedance devices, which read position and output force; admittance devices read force and output position, and are generally used where high forces and a large workspace are needed.[1]

How it works

The haptic loop

The 2004 survey divides a haptic rendering algorithm into three blocks:[1]

Block Role
Collision detection Finds where, when and to what extent the avatar has contacted or penetrated virtual objects
Force response Computes the ideal interaction force (and torque) between the avatar and the objects it touches
Control algorithms Command the device so that the force actually applied to the user is as close as possible to the ideal force, within the device's limits

In a typical loop, low-level control code samples the position sensors at the device joints, converts the readings into the avatar's position in the virtual environment, runs collision detection, computes the interaction force, and sends it to the device while keeping the overall behavior stable. The simulation engine then uses the same forces to update the virtual objects.[1]

The survey states that there are "no firm rules" about how often these computations must repeat, but that a 1 kHz servo rate is common, as a compromise that allows reasonably complex objects with reasonable stiffness; higher rates give crisper contact and texture at the cost of scene complexity.[1] Diego Ruspini, Krasimir Kolarov and Oussama Khatib wrote in 1997 that haptic devices require high controller servo rates, typically over 1000 Hz, to achieve stability, and that missing these rates can make a system unstable, potentially causing device damage or user injury.[6] By comparison, they noted that desktop graphics systems of the time could render more than 20,000 shaded and textured polygons at interactive rates of 30 Hz, while most haptic systems could represent only a few dozen geometric primitives.[6]

Because graphics, physics and collision detection are much slower than the force loop, many systems split the work into threads running at different rates. The 2004 survey describes four: a visual rendering loop at up to 30 Hz, a simulation thread, a slower collision detection thread that builds a simple local model of the surface nearest the avatar (such as a plane or sphere), and a fast force loop that renders that local model at the full servo rate.[1] Ruspini's group ran the haptic server and the graphics client on separate computers; they credited Adachi and colleagues with first proposing the separation of haptic and graphics processes.[6]

Virtual walls and penalty methods

Rendering a virtual wall, the force from touching an infinitely stiff surface, is the standard test case for haptic rendering and a common benchmark in studies of stability.[1] Because the loop runs in discrete time, the avatar always penetrates the virtual object slightly. The simplest wall algorithm uses that penetration directly: for a wall at position xW, the force is zero in free space and F = K(xW - x) once the avatar has crossed the wall, where K is the wall's stiffness.[1]

Early three-DOF methods extended this idea with "vector field" approaches that mapped each position inside an object directly to a force. Zilles and Salisbury listed their drawbacks in 1995: it is often unclear which surface a given internal point belongs to, forces jump when the probe crosses a sub-volume boundary, and thin objects have too little internal volume to push back. On a thin object, once the probe passed the halfway point, the force reversed and pulled the user through to the other side.[2] Adding the force fields of overlapping objects also gave wrong results: as the angle between two surfaces became more obtuse, the apparent stiffness at their junction approached twice that of either surface alone.[2]

God-object and proxy methods

To solve these problems, C. B. Zilles and J. K. Salisbury at the Massachusetts Institute of Technology introduced the god-object method in 1995.[2] The god-object is an extra, virtual copy of the haptic interface point that is not allowed to penetrate objects. In free space it sits at the same place as the device's measured position; when the device moves into an object, the god-object stays on the surface, "where the haptic interface point would be if the haptic interface and object were infinitely stiff." Its new position is the point on the constraining surfaces that locally minimizes the distance to the device position, found with Lagrange multipliers, and impedance control with a stiffness and damping between the two points gives the force sent to the user.[2] Because the god-object keeps a compact history of contact, the force direction is never ambiguous, and the method works for thin objects and arbitrary polyhedra.[2] Their implementation, written in C++ for a 66 MHz Pentium connected to a PHANToM, ran at about 1 kHz for objects of about 600 triangles; only the collision detection part scaled with model size.[2] They also noted that the PHANToM-style device they used had a maximum force of 18 N (4 lbf), while most users exploring virtual objects used less than 5 N (1 lbf).[2]

Ruspini, Kolarov and Khatib at Stanford University and Interval Research presented the closely related virtual proxy at SIGGRAPH 97.[6] They modeled the proxy as a massless sphere that moves toward the device position like a robot greedily approaching a goal, sliding along constraint surfaces when blocked. Because the proxy has a finite size, it does not slip through the small numerical gaps found in most polygon meshes.[6] Their "HL" library had an interface almost identical to Silicon Graphics' GL graphics library, so graphics applications could add haptics with little change. In their tests a 200 MHz Pentium Pro haptic server driving a PHANToM rendered models of more than 24,000 polygons, with stable position gains over 1800 N/m and no artificial damping, and the proxy update time grew roughly with the logarithm of the polygon count.[6] The 2004 survey describes the two methods as built on the same principle and developed independently.[1]

Force shading, friction and texture

Polygon models make a curved surface feel faceted. Morgenbesser and Srinivasan introduced force shading, described in the 2004 survey as "the haptic equivalent of Phong shading": interpolating surface normals across a mesh makes the force change smoothly over the surface.[1] Ruspini's group implemented force shading, static and dynamic friction, surface stiffness and texture by changing the position of the virtual proxy.[6]

The 2004 survey lists several approaches to surface properties. Simple friction models based on the work of Leonardo da Vinci and Charles-Augustin de Coulomb were used in early three-DOF renderers, while models such as the Karnopp stick-slip model are more accurate but slower. Margaret Minsky first proposed haptic texture mapping in 2D, which Ruspini and colleagues later extended to 3D scenes; other groups generated textures from fractals or stochastic models.[1] Later work built texture models from recordings of real surfaces. At the 2014 IEEE Haptics Symposium, Heather Culbertson, Juan Jose Lopez Delgado and Katherine J. Kuchenbecker presented one hundred data-driven haptic texture models with open-source methods for rendering them on 3D objects. The University of Pennsylvania's Penn Haptic Texture Toolkit contains 100 haptic texture and friction models, the recorded data behind them, images of the textures, and code to render them on an impedance-type device such as a SensAble Phantom Omni.[7][8]

More than three degrees of freedom

A point probe cannot transmit torque, so a user can feel an object with a tool's tip while the rest of the virtual tool passes through it.[1] The 2004 survey lists methods with richer avatars: four-DOF "soft finger" contacts that support torsional friction (Barbagli and colleagues), five-DOF ray-based rendering of tools as line segments (Basdogan and colleagues), and six-DOF methods for contact between complex rigid objects (McNeely and colleagues; Otaduy and Lin).[1] Michael Ortega, Stephane Redon and Sabine Coquillart later extended the god-object approach itself to six DOF.[9]

William McNeely, Kevin Puterbaugh and James Troy of Boeing presented a voxel-based six-DOF method at SIGGRAPH 99. The static environment is stored as a voxel map and the moving object as a "point shell" of surface points with inward normals; every haptic frame samples the voxel map at each point.[10] They listed the extra requirements of six-DOF haptics compared with collision detection, including detecting all contacts rather than stopping at the first, computing force and torque at every contact, and reliably maintaining a 1000 Hz refresh rate regardless of the object's position and orientation. Their system drove a PHANToM Premium 6-DOF prototype at a constant 1000 Hz through SensAble's GHOST SDK, in a test scene of simulated aircraft geometry with 593,409 polygons voxelized at 5 mm, while graphics ran separately at about 20 Hz.[10] The authors judged the approach suited to maintenance and assembly simulations that can tolerate voxel-scale accuracy.[10]

Stability

A haptic device, the user and the rendering algorithm form a closed feedback loop, and that loop can become unstable. The 2004 survey gives four causes: devices have limited force; they have friction, inertia and backlash; position sensors have finite resolution; and the algorithm runs in discrete time while the user moves continuously. Because the sampled position lags behind the real one, a user pressing into and releasing a virtual object gets back more energy than a real object would return, which can make the device vibrate.[1] Researchers have relied largely on passivity theory, since the human operator cannot be described by a simple transfer function.[1]

For a virtual wall, J. E. Colgate and J. M. Brown showed in 1994 that the stiffness a device can render stably is linked to its mechanical damping, the digital damping commanded by the controller and the servo period. As summarized in the 2004 survey, stable interaction requires b > KT/2 + B, where b is the device's physical damping, B the virtual damping, K the rendered stiffness and T the servo period. Stiff walls therefore become unstable at long servo periods, which is a main reason high servo rates matter.[1][11] Colgate also introduced the virtual coupling, a spring and damper between the device and the avatar that caps the impedance the device must display: the virtual world may contain arbitrarily stiff objects, but the device only renders up to the coupling's limit.[1]

Blake Hannaford and Jee-Hwan Ryu at the University of Washington developed time-domain passivity control. A "Passivity Observer" measures the energy produced at the haptic interface, and a "Passivity Controller" computes damping that absorbs the excess energy which would otherwise cause vibration.[12][13] The University of Washington patent on the method, filed in September 2001 and granted in April 2006, covers both one-port virtual reality systems and two-port teleoperation systems.[13]

History

The 2004 survey traces the word "haptics" to early 20th century psychophysics and notes that a new usage, "computer haptics", emerged in the early 1990s, when force-controllable devices, geometric modeling, collision detection and affordable computing came together. It states that Knoll demonstrated haptic interaction with simple virtual objects at least as early as the 1960s.[1] Fred Brooks' Project GROPE, started in 1967, built force displays for molecular docking; its 1990 SIGGRAPH paper reported that haptic display added to visual display improved understanding of force fields, and that chemists using GROPE-III could reproduce known drug docking positions.[14]

Year Development
1990 Brooks and colleagues publish Project GROPE results on haptic displays for scientific visualization at SIGGRAPH[14]
Early 1990s The point-interaction paradigm, using a single probe point and three force components, is recognized as a useful simplification[1]
1994 Massie and Salisbury describe the PHANToM haptic interface; Colgate and Brown analyze the stable stiffness range of a haptic display[2][11]
1995 Zilles and Salisbury publish the constraint-based god-object method[2]
1997 Ruspini, Kolarov and Khatib present the virtual proxy and the HL haptic library at SIGGRAPH 97[6]
1999 McNeely, Puterbaugh and Troy of Boeing present six-DOF voxel sampling at SIGGRAPH 99[10]
2002 Hannaford and Ryu publish time-domain passivity control[12]
2003 The open-source CHAI3D framework is launched at Stanford University[15]
2004 IEEE Computer Graphics and Applications publishes a theme issue on haptic rendering[1]
2008 Ming C. Lin and Miguel Otaduy edit the book Haptic Rendering: Foundations, Algorithms, and Applications[16]
2014 Culbertson, Lopez Delgado and Kuchenbecker publish 100 data-driven haptic texture models; Long and colleagues render volumetric shapes in mid-air with ultrasound[8][5]
2016 Azmandian and colleagues present haptic retargeting for VR at CHI 2016[17]

Applications in VR and AR

Salisbury, Conti and Barbagli presented haptic rendering as one part of a VR application, alongside visual and auditory rendering, with a simulation engine computing the environment's behavior and transducers converting each signal into a form the user can perceive.[1] A 2019 review by Dangxiao Wang and colleagues in Virtual Reality & Intelligent Hardware argued that VR systems already had fairly realistic visual and auditory feedback but poor haptic feedback. It grouped 30 years of haptic display research into desktop haptics, surface haptics and wearable haptics, and named high-fidelity haptic rendering, together with handheld and multimodal devices, as open challenges for VR.[18]

Controller vibration

Handheld VR controllers render touch as vibrotactile feedback. The 2004 survey already noted that handheld devices in the gaming industry used low-cost vibrotactile transducers to produce synthesized vibration effects.[1] In the OpenXR standard, an application triggers a vibration output action with the XrHapticVibration structure, which specifies a duration in nanoseconds, a frequency in Hz and an amplitude between 0.0 and 1.0; the runtime may clamp duration and frequency to what the device supports.[3] The XR_FB_haptic_pcm extension instead lets an application stream PCM sample buffers, with values from -1.0 to 1.0 and a chosen sample rate, as a haptic waveform.[19]

Meta's developer documentation recommends its Haptics Studio desktop tool and Haptics SDK for Meta Quest development. Haptics Studio lets designers create and audition haptics without writing code, and the SDK plays the resulting clips through a "unified, high-level, media-based API" that detects the controller at runtime and adapts the pattern to its capabilities.[4] According to the Unity SDK documentation, a clip stores amplitude, frequency, emphasis and duration information; it plays as high-fidelity PCM haptics on Meta Quest devices and falls back to amplitude-only vibration on other PC VR devices or older Quest models, and Meta notes that Meta Quest 2 controllers have less advanced haptic actuators that support only a fixed vibration frequency.[20]

Gloves and wearables

The 2004 survey, which classifies haptic devices by where they are grounded, notes that force-feedback gloves can read finger-specific contact and output finger-specific resistive forces, but cannot reproduce an object's net weight or inertial forces.[1] In November 2021 Meta's Reality Labs described a research glove prototype from a team that had worked on haptics for seven years; the glove used small soft actuators moved by air, with a microfluidic chip controlling the air flow. Meta wrote that its haptic renderer "sends precise instructions to the actuators on the hand" based on the hand's location and the properties of the virtual objects it touches, such as texture, weight and stiffness. The team also combined auditory, visual and haptic cues to create the impression of an object's weight.[21]

Mid-air haptics

Mid-air haptics render touch without any worn or held device. Benjamin Long, Sue Ann Seah, Tom Carter and Sriram Subramanian of the University of Bristol described a method at SIGGRAPH Asia 2014 that uses acoustic radiation force from a two-dimensional phased array of ultrasound transducers; their algorithm controls the distribution of the force field in a volume so that users perceive a three-dimensional shape in the air.[5]

Passive haptics and retargeting

Some VR systems use real props instead of active force display. Azmandian, Hancock, Benko, Ofek and Wilson showed at CHI 2016 that a single physical prop can provide passive haptics for several virtual objects by exploiting the dominance of vision when the senses conflict. Their haptic retargeting techniques manipulated the virtual world, the user's virtual body, or both, to align the prop with the touched virtual object. All three improved presence compared with wand-based control of virtual objects, and the hybrid technique scored highest for satisfaction and presence.[17]

Simulation and training

The 2004 survey cited surgical simulators as an area where multirate rendering with simple local models had become popular, because accurate deformable-object dynamics were still too slow for the haptic loop.[1] McNeely's Boeing work targeted aircraft maintenance and assembly planning, where haptics lets a user feel collisions while moving a part or tool through a crowded virtual mockup.[10] In medical VR, FundamentalVR markets a system it calls HapticVR. A 2022 TechCrunch report described HapticVR as blending cutaneous (tactile vibration) haptics with kinesthetic force and positional feedback, and as compatible with handheld devices including haptic gloves and purpose-built controllers; in the same report chief executive Richard Vincent said that "not all haptics are created equal".[22]

Software

Several software layers hide the servo loop from application developers. McNeely's Boeing system used SensAble's GHOST SDK to manage the 1 kHz servo loop of a PHANToM.[10] Ruspini's HL library mirrored the GL graphics API so that graphics programmers could describe haptic scenes with familiar calls.[6] CHAI3D, first launched in 2003 at the Robotics and Artificial Intelligence Laboratory at Stanford University, is an open-source C++ framework for haptics, visualization and real-time simulation that supports commercial three-, six- and seven-DOF haptic devices.[15] It is released under the 3-clause BSD license, and its reference paper by Conti and colleagues appeared at Eurohaptics 2003.[23] For consumer VR, haptic output is exposed through OpenXR and through vendor tools such as Meta's Haptics SDK for Unity and Unreal Engine.[3][4]

Research

Heather Culbertson, Samuel Schorr and Allison Okamura's 2018 review in the Annual Review of Control, Robotics, and Autonomous Systems covers the design and control of haptic devices and best practices for generating distinct touch sensations, and addresses the need to consider the neuroscience and perception of human touch when designing and controlling haptic devices.[24] The 2019 review by Wang and colleagues likewise ends by emphasizing human haptic perception as the basis for effective haptic devices.[18]

Recent work borrows directly from graphics and audio. In a 2025 arXiv preprint, Tom Roy, Ferran Argelaguet and colleagues proposed "Haptic Tracing", a spatial haptic rendering method that propagates haptic information through a 3D scene using ideas from visual and audio rendering instead of physical simulation, and applied it to a multi-actuator vibrotactile system with a user study.[25]

See also

References

  1. ↑ 1.00 1.01 1.02 1.03 1.04 1.05 1.06 1.07 1.08 1.09 1.10 1.11 1.12 1.13 1.14 1.15 1.16 1.17 1.18 1.19 1.20 1.21 1.22 1.23 1.24 1.25 1.26 1.27 1.28 1.29 Kenneth Salisbury, Francois Conti, Federico Barbagli (2004-03). "Haptic rendering: introductory concepts". IEEE Computer Graphics and Applications, vol. 24, no. 2, pp. 24-32. IEEE Computer Society. doi:10.1109/MCG.2004.1274058. https://doi.org/10.1109/MCG.2004.1274058. Retrieved 2026-10-13.
  2. ↑ 2.00 2.01 2.02 2.03 2.04 2.05 2.06 2.07 2.08 2.09 C. B. Zilles, J. K. Salisbury (1995). "A constraint-based god-object method for haptic display". Proceedings 1995 IEEE/RSJ International Conference on Intelligent Robots and Systems, vol. 3, pp. 146-151. IEEE. doi:10.1109/IROS.1995.525876. https://doi.org/10.1109/IROS.1995.525876. Retrieved 2026-10-13.
  3. ↑ 3.0 3.1 3.2 "XrHapticVibration". OpenXR 1.1 Reference Pages. The Khronos Group. https://registry.khronos.org/OpenXR/specs/1.1/man/html/XrHapticVibration.html. Retrieved 2026-10-13.
  4. ↑ 4.0 4.1 4.2 "Haptic Feedback". Meta Horizon OS Developers. Meta. 2025-11-10. https://developers.meta.com/horizon/documentation/unity/unity-haptics-overview/. Retrieved 2026-10-13.
  5. ↑ 5.0 5.1 5.2 Benjamin Long, Sue Ann Seah, Tom Carter, Sriram Subramanian (2014-11). "Rendering volumetric haptic shapes in mid-air using ultrasound". ACM Transactions on Graphics, vol. 33, no. 6, article 181 (SIGGRAPH Asia 2014). University of Bristol. doi:10.1145/2661229.2661257. https://research-information.bris.ac.uk/en/publications/rendering-volumetric-haptic-shapes-in-mid-air-using-ultrasound-pr/. Retrieved 2026-10-13.
  6. ↑ 6.0 6.1 6.2 6.3 6.4 6.5 6.6 6.7 6.8 Diego C. Ruspini, Krasimir Kolarov, Oussama Khatib (1997-08). "The haptic display of complex graphical environments". Proceedings of SIGGRAPH 97, pp. 345-352. ACM. doi:10.1145/258734.258878. https://doi.org/10.1145/258734.258878. Retrieved 2026-10-13.
  7. ↑ "The Penn Haptic Texture Toolkit". GRASP Laboratory. University of Pennsylvania. https://grasp.upenn.edu/projects/the-penn-haptic-texture-toolkit. Retrieved 2026-10-13.
  8. ↑ 8.0 8.1 Heather Culbertson, Juan Jose Lopez Delgado, Katherine J. Kuchenbecker (2014-02). "One hundred data-driven haptic texture models and open-source methods for rendering on 3D objects". 2014 IEEE Haptics Symposium (HAPTICS), pp. 319-325. IEEE. doi:10.1109/HAPTICS.2014.6775475. https://doi.org/10.1109/HAPTICS.2014.6775475. Retrieved 2026-10-13.
  9. ↑ Michael Ortega, Stephane Redon, Sabine Coquillart (2007-05). "A Six Degree-of-Freedom God-Object Method for Haptic Display of Rigid Bodies with Surface Properties". IEEE Transactions on Visualization and Computer Graphics, vol. 13, no. 3, pp. 458-469. IEEE. doi:10.1109/TVCG.2007.1028. https://doi.org/10.1109/TVCG.2007.1028. Retrieved 2026-10-13.
  10. ↑ 10.0 10.1 10.2 10.3 10.4 10.5 William A. McNeely, Kevin D. Puterbaugh, James J. Troy (1999). "Six degree-of-freedom haptic rendering using voxel sampling". Proceedings of SIGGRAPH 99, pp. 401-408. ACM. doi:10.1145/311535.311600. https://doi.org/10.1145/311535.311600. Retrieved 2026-10-13.
  11. ↑ 11.0 11.1 J. E. Colgate, J. M. Brown (1994). "Factors affecting the Z-Width of a haptic display". Proceedings of the 1994 IEEE International Conference on Robotics and Automation, pp. 3205-3210. IEEE. doi:10.1109/ROBOT.1994.351077. https://doi.org/10.1109/ROBOT.1994.351077. Retrieved 2026-10-13.
  12. ↑ 12.0 12.1 B. Hannaford, Jee-Hwan Ryu (2002). "Time-domain passivity control of haptic interfaces". IEEE Transactions on Robotics and Automation, vol. 18, no. 1, pp. 1-10. IEEE. doi:10.1109/70.988969. https://doi.org/10.1109/70.988969. Retrieved 2026-10-13.
  13. ↑ 13.0 13.1 Blake Hannaford, Jee-Hwan Ryu (2006-04-11). "US7027965B2 - Time domain passivity control of haptic interfaces". Google Patents. University of Washington. https://patents.google.com/patent/US7027965. Retrieved 2026-10-13.
  14. ↑ 14.0 14.1 Frederick P. Brooks Jr., Ming Ouh-Young, James J. Batter, P. Jerome Kilpatrick (1990). "Project GROPE: Haptic displays for scientific visualization". ACM SIGGRAPH History Archive (SIGGRAPH 1990 technical papers). ACM SIGGRAPH. doi:10.1145/97879.97899. https://history.siggraph.org/?p=122061. Retrieved 2026-10-13.
  15. ↑ 15.0 15.1 "About". CHAI3D. https://www.chai3d.org/concept/about. Retrieved 2026-10-13.
  16. ↑ Ming C. Lin, Miguel Otaduy (eds.) (2008-07-25). "Haptic Rendering: Foundations, Algorithms, and Applications". A K Peters/CRC Press. doi:10.1201/b10636. https://doi.org/10.1201/b10636. Retrieved 2026-10-13.
  17. ↑ 17.0 17.1 Mahdi Azmandian, Mark Hancock, Hrvoje Benko, Eyal Ofek, Andrew D. Wilson (2016-05). "Haptic Retargeting: Dynamic Repurposing of Passive Haptics for Enhanced Virtual Reality Experiences". Proceedings of the 2016 CHI Conference on Human Factors in Computing Systems, pp. 1968-1979. Microsoft Research. doi:10.1145/2858036.2858226. https://www.microsoft.com/en-us/research/publication/haptic-retargeting-dynamic-repurposing-passive-haptics-enhanced-virtual-reality-experiences/. Retrieved 2026-10-13.
  18. ↑ 18.0 18.1 Dangxiao Wang, Yuan Guo, Shiyi Liu, Yuru Zhang, Weiliang Xu, Jing Xiao (2019-04). "Haptic display for virtual reality: progress and challenges". Virtual Reality & Intelligent Hardware, vol. 1, no. 2, pp. 136-162. doi:10.3724/SP.J.2096-5796.2019.0008. https://doi.org/10.3724/SP.J.2096-5796.2019.0008. Retrieved 2026-10-13.
  19. ↑ "XrHapticPcmVibrationFB". OpenXR 1.1 Reference Pages. The Khronos Group. https://registry.khronos.org/OpenXR/specs/1.1/man/html/XrHapticPcmVibrationFB.html. Retrieved 2026-10-13.
  20. ↑ "Haptics SDK for Unity". Meta Horizon OS Developers. Meta. https://developers.meta.com/horizon/documentation/unity/unity-haptics-sdk. Retrieved 2026-10-13.
  21. ↑ "Inside Reality Labs Research: Bringing Touch to the Virtual World". Meta Newsroom. Meta. 2021-11-16. https://about.fb.com/news/2021/11/reality-labs-haptic-gloves-research/. Retrieved 2026-10-13.
  22. ↑ Paul Sawers (2022-08-11). "Medical simulation platform FundamentalVR raises $20M to help surgeons learn through VR". TechCrunch. https://techcrunch.com/2022/08/11/medical-simulation-platform-fundamentalvr-raises-20m-to-help-surgeons-learn-through-vr/. Retrieved 2026-10-13.
  23. ↑ "License". CHAI3D. https://www.chai3d.org/download/license. Retrieved 2026-10-13.
  24. ↑ Heather Culbertson, Samuel B. Schorr, Allison M. Okamura (2018-05-28). "Haptics: The Present and Future of Artificial Touch Sensation". Annual Review of Control, Robotics, and Autonomous Systems, vol. 1, pp. 385-409. doi:10.1146/annurev-control-060117-105043. https://doi.org/10.1146/annurev-control-060117-105043. Retrieved 2026-10-13.
  25. ↑ Tom Roy, Yann Glemarec, Gurvan Lecuyer, Quentin Galvane, Philippe Guillotel, Ferran Argelaguet (2025-08-27). "Haptic Tracing: A new paradigm for spatialized Haptic rendering". arXiv. arXiv:2508.19703. https://arxiv.org/abs/2508.19703. Retrieved 2026-10-13.
Retrieved from "https://vrarwiki.com/index.php?title=Haptic_rendering&oldid=50129"