Let's talk about haptic gloves

[rhinosix]

With the Oculus Rift on the horizon, the major component for functional VR will become accessible and continue to improve over time.

Hand and body tracking seem to be on their way. Audio is extremely important for immersion, and is only a matter of software support.

Locomotion devices are restricted mainly by cost, design, and size, rather than a huge engineering hurdle. And a locomotion device can be anything from an ODT, gamepad, modified exercise equipment, bicycle, head-tracker, steering wheel, harness, exoskeleton, etc.

We can build different types of locomotion devices that work well enough for various applications, but probably have some time before we have an all-in-one solution.

For immersion, I see haptics as the area most in need of attention. In particular, haptic data gloves.

To define what I’m talking about, I don’t mean gloves that give general force feedback from gunshots or explosions; binary touch feedback; or controls assigned to hand gestures. I mean gloves which can feel:

* texture (between smooth or bumpy surfaces)
* shape (like the edge of a virtual table, or a virtual pen)
* weight (holding a cup, or running hands through water)
* temperature (the cold of steel, or heat from a light bulb)

* and allow the user to manipulate virtual objects.

I’m largely ignorant, but interested. Just looking at the various available gloves, and hearing about Disney Research, flexible, wearable electronics, displays, electroactive polymers, HeatIt, muscle wires, touch-sensitive self-healing skin, and so on, I know there is a lot of interesting information out there. This thread is to post links, learn and ask questions such as:

What has been achieved so far in the engineering, military, and medical industries?
What’s the most advanced available glove? How is it limited?
What is an ideal design?
What kind of new and novel, or potentially interesting technologies might be transferable to a data glove?
What can a DIYer hope to achieve with currently available materials?
Is it a matter of using current technology in a new way?
Are we waiting on one specific new material or other technology?
How long until we have haptic data gloves on the level of Johnny Mnemonic?

Feel free to post your thoughts, insights, links, and comments. I’m going to try to compile links and information to post back later.

[EdZ]

All the haptic gloves I know of are research oriented, and exorbitantly expensive, way outside consumer price ranges. The general setup is very similar between all of them: a set of rigid linkages over each finger, with a Bowden cable (think a bicycle brake cable) linking between each of them to the actuators mounted either on the back of the 'glove' or the wrist/forearm (if small & light actuators) or to a backpack mount for older models. The more degrees of feedback you want, the more cables and the more actuators. Some will have a cable for every knuckle (i.e. 3 per finger), and some will only have one per finger and use a joint model assuming you're not contorting your finger unusually. These single-linkage-per-finger setups either use a mechanical linkage to provide a fixed amount of this force to each joint, or only effect the fingertip.
I've only seen one design that applies sideways force to the fingertip as well as grasp/release, and that used what is essentially a miniature robotic arm for each finger, linked at the tip and at the 3rd knuckle.

To my knowledge, haptic stimulation of fingertip touch sense is very crude at this stage, both physical pin-array stimulation or direct electrostimulation of sensory nerves.

What can a DIYer hope to achieve with currently available materials?

With the limitation of touch sensitivity, and the expense of multiple fine (and powerful!) actuators, and the difficulty of precisely and accurately measuring the current position and orientation of each segment of each of your digits, a DIY haptic glove is likely to be fairly crude. You might be able to use this to your advantage: build the glove around a suit-of-armour style glove. The inherent rigidity and lack of touch data will encourage users to work within the limitations of the system intuitively, and the fixed range of motion for most joints and ease of mounting makes it simpler to work with than a fabric glove. You can substitute thermoformed plastic/epoxy-fibre/poured resin casts for beaten metal for cost and ease of machining.

---

To provide 'weight' to objects (or 'kick' for guns, for example), you will also need a haptic arm setup, and that's another beast that's more rarely tackled.

There's a chance I might have my Manipulator Dymanics and Haptics lecture notes stored on an old disk, I'll try and dig them up.

[cybereality]

I have tried the Novint Falcon, and I think haptics added a lot to the experience. Of course you got nice kick-back from guns, but also more subtle things: like the weight of an oil drum when you pick it up. It really does take it to the next level. I think the haptics you could get on a glove would be limited unless it was something like the XIO, basically an exo-skeleton arm. I guess for fine manipulation it would be great, like picking up a glass or ball or something. But I think the force feedback (like gun kick) would be difficult, and having weight to an object nearly impossible. Even so, I would love for a good data glove with haptics. Heck, I'd even take just a nice glove without any feedback if possible. That would be nice.