Virtual Reality

Electromagnets promise precision finger-tracking for VR environments

Electromagnets promise precisi...
Finexus uses a set of electromagnets fitted to the user's fingertips and four magnetometers to detect their motion
Finexus uses a set of electromagnets fitted to the user's fingertips and four magnetometers to detect their motion
View 2 Images
In its current form, Finexus only works when the sensors and electromagnets are within 12 cm (4.7 in) of each other
1/2
In its current form, Finexus only works when the sensors and electromagnets are within 12 cm (4.7 in) of each other
Finexus uses a set of electromagnets fitted to the user's fingertips and four magnetometers to detect their motion
2/2
Finexus uses a set of electromagnets fitted to the user's fingertips and four magnetometers to detect their motion

Expected some time in the new year, the consumer launch of the Oculus Rift and Touch controllers soon after are set to open immersive new worlds for VR explorers. But how we interact with those worlds will continue to evolve long after the first headsets are strapped on. Offering a glimpse into the future of these input devices, one alumni of Oculus' research division has developed an electromagnetic sensor system that tracks fingertip movements, should you need to say, stop and play a piano or create an oil painting during your VR travels.

Oculus Research is a mostly autonomous group inside the company that pretty much picks and chooses which ideas to pursue, with a view to making a big impact on the Rift years down the track, CEO Brendan Iribe told us in June. For PhD candidate at the University of Washington, Keyu Chen, who interned there earlier this year, this free reign inspired the development of a new multipoint tracking system to monitor detailed movement of fingertips in real time.

Dubbed Finexus, the system uses a set of electromagnets fitted to the user's fingertips and four magnetometers to detect their motion. With a special algorithm and custom PCB, Finexus is able to use magnetic field sensing to track the position of the fingertips in a 3D environment. Each of the electromagnets operates at a different frequency, reportedly making it easier to identify them separately, while also making the system less vulnerable to surrounding electromagnetic noise.

One potential advantage of Chen's approach over other gesture-sensing technologies that use infrared cameras like Leap Motion or Microsoft's Kinect, is that Finexus won't need to maintain a line-of-sight to function. So you a quick dip into your pocket to grab a weapon or item, for example, won't stop the system from tracking your fingers.

In its current form, Finexus only works when the sensors and electromagnets are within 12 cm (4.7 in) of each other. Chen hopes to extend this working range to as much as 25 cm (10 in) and develop the system to one day comprise a glove embedded with electromagnets and sensors hidden inside the wristband or a smartwatch.

The video below shows Finexus in action.

Source: Keyu Chen via MIT

Finexus: Tracking Precise Motions of Multiple Fingertips Using Magnetic Sensing

1 comment
noteugene
I don't see why Caterpillar or some other heavy duty machinery manufacturing company could not incorporate a glove similar to this to install into a backhoe cab. Ever driven a backhoe? Sure...if you've driven one for a few months or so your speeds not so bad. With one of these and an improved bucket, one narrow bucket for each finger, you could use 1 finger to 4 depending on width of cut. You could pinch your fingers and thumb together to scoop up dirt. You could spread your fingers apart a bit...or not. You could over lap your fingers to add strength if you hit a hard spot. You could scratch like a dog.....with 1 or both hands. You could dig with 1 hand & redeposit material with the other..... simultaneously. If all you used was 1 or 2 fingers to install, say, a 6" pipe, think of all the dirt you don't have to dig. Time is money is not exactly right. Time is worth way more than money. You could build some backhoes about the size of a riding mower & it'd do the job a whole lot faster. Small jobs anyway. The design or engineering hasn't changed all that much for this piece of equipment in 100 years which is pretty pathetic. Getting the neighborhood lawn boy to install a pipe for $10 hr beats paying a backhoe operator $40 - $50 hr. Let's hope somebody with money to build a prototype gets the same idea.