Senso is an experimental glove peripheral that tracks individual fingers for virtual reality hand controls, while offering haptic feedback that includes vibration and temperature fluctuation. If you want the technical details, you should check out Road to VR’s solid and appropriately skeptical rundown, because we’re not really here to talk about Senso. We’re here to talk about our eternal, unrequited, and misplaced love for the VR glove.
At first glance, gloves seem like a simple, obvious Controller Of The Future: they’re a semi-ordinary piece of clothing that can simultaneously track natural finger motion (unlike a handheld controller) and provide tactile feedback (unlike Kinect-style cameras.) That’s why there are a billion different versions of them. You’ve got the famous Nintendo Power Glove and its high-end inspiration, VR pioneer Tom Zimmerman’s VPL DataGlove. Road to VR names a half-dozen recent examples besides Senso, and that’s still not a comprehensive list. And then there are all the fictional iterations — perhaps most famously, Tom Cruise’s stylish three-finger gloves in Minority Report.
Yet despite all this effort, motion-tracking gloves aren’t part of any modern consumer virtual or augmented reality system. Because like an ostentatious mustache, cybergloves only work if you get them exactly right — and until that point, they’re just silly-looking and unpleasant. Why? Let me explain.
‘Fits like a glove’ is just an idiom
Gloves are only one-size-fits-all if they’re small but super-stretchy, and VR gear often needs to pack in a lotof electronics. This means that development kits — which is what most cybergloves are — tend to run large. If you’re not a big person, their much-vaunted freedom of movement gets hampered by the wads of fabric around your fingers, while air gaps dull haptic feedback.
Comfort isn’t the biggest issue for specialized industrial tools, and obviously a prototype will be rough. But most companies — including ones that intend to sell mass-market consumer products — seem barely interested in the difficulties of building intimate hardware for diverse bodies. If that’s the case, why make a controller that’s hugely dependent on meeting that challenge?
Hardware isn’t a silver bullet
Compared to cameras that track bare hands, gloves can take some of the guesswork out of finding your real-life finger position. Unfortunately, a computer still has to interpret whether those fingers are spread quite enough to mean you want to grab something in VR, or whether some offhand motion matches a gesture you recorded earlier. A physics system has to make sure objects don’t bounce off your fingers or stick to them like glue. Companies like Leap Motion have put a ton of work into software design, and it’s still imperfect. Not every hardware startup will get that far — or even have the time and expertise to try.
In fact, this is by far the biggest advantage Rift- or Vive-style controllers have over any kind of gesture interface: it’s really, really easy to tell when somebody has pulled a trigger.
Haptics are hard
VR doesn’t need to be lifelike to be fun. But if a device is touting full, immersive haptics, vibrations and heating elements aren’t enough. I want to grab virtual objects and actually feel my hand close around them, which is only possible if something physically stops my fingers.
The obvious solution is force-feedback hand exoskeletons, which are a real thing and look terrifying. But this is a perfect example of how VR gloves only seem like a viable product if a whole constellation of features — gesture recognition, haptics, comfort — comes together perfectly. It’s the only way to compensate for the fact that, well, gloves honestly aren’t much fun to wear.
Gloves are basically gross and uncomfortable
You know what your hands touch? Everything. And you know what they’re coated in? Oil and sweat. So any covering for them will be constantly picking up dirt both inside and out. Add the annoyance of an even slightly ill-fitting design, the inconvenience of having to either wear or carry a pair everywhere, and the misery they’ll cause in the heat of summer, and computing gloves had better offer something extra-special.
One of Senso’s big selling points is that unlike a Kinect or similar device, you don’t need an external camera. But cameras aren’t much of a problem if you can build really good ones straight into VR or AR headsets — which Microsoft, Leap Motion, and others are already trying. Gloves arguably still have some advantages, but again, it’s not enough for something to be slightly better if it’s significantly less convenient.
What else could we build?
I still love the idea of a natural-feeling VR control system with tactile feedback, however it’s done. Motion control jewelry offers some of the same advantages as gloves without covering your whole hand. Haptics don’t necessarily require skin contact at all. Gest was an ill-fated but clever example of what the future could look like, and although I had a lot of difficulties when I once tried out the Myo armband, I wore it a lot longer than I would have used a pair of gloves.
Maybe someone will find a sleek, self-cleaning, ultra-flexible electronic skin that makes me like gloves. In the short term, products like Senso are a fun proof of concept, but camera-based hand tracking and handheld controllers are still the two gold standards of VR. For gloves to have any place in the future of computing, they’ll have to beat both of these technologies while being legitimately wearable — and so far, that’s an elusive combination.