Carnegie Mellon Offers Wee QWERTY Texting Tech For Impossibly Tiny Devices

coondoggie writes "If smartwatches and other ultra-small devices are to become the text generators of the future, their diminutive keyboards are going to have to be way more useful for, um, big fingered typists. Carnegie Mellon researchers may have the answer to that problem. Called ZoomBoard, the text entry technique is based on the iconic QWERTY keyboard layout." The zoom board paper (PDF) has details. Entering a letter becomes a multi-step process; first you mash the general area of the keyboard containing the letter you want, and eventually it becomes large enough to hit. Test subjects managed to hit 9.3wpm after practice, versus 4.5 wpm for people trying to peck on a teeny-tiny virtual keyboard. They were inspired at least in part by the venerable Dasher input method.

Comments

Not as good as Morse

[Geoffrey.landis]

People should just learn Morse code, only one button. It's the original text message tech.
And good Morse code operators go vastly faster than a mere 9.3 world per minute.

Re:Not as good as Morse

[hairyfeet]

You joke (or half joke, can never tell on the net) but THIS, this right here, is why the PC and laptop will NEVER die, trying to input text on one of these little devices is an exercise in frustration and irritation. Hell I'm a two fingered typist (got some fingers broke in a bike wreck and while they play bass decently the curve that the joints are set in don't respond well to typing) and I can positively fly low on a regular keyboard compared to a 100 WPM touch typist on one of those little mini-key jobs, which needless to say is frustrating as hell for them. Dragging along a BT keyboard of course becomes more bulky and a bigger PITA than just carrying a little laptop so we are right back where we started.

I've seen everything from virtual keyboards to the "spin the thing until it lands on the right letter" and frankly nothing has even come close to a full size keyboard and until they do i just don't see anybody giving up their desktops and laptops, no matter how much the OEMs (and MSFT) want to force us onto "ecosystems" where they control everything and get a cut of every sale, it might be good for them but the interface just sucks.

Re:Not as good as Morse

[Anubis IV]

Speaking to a device is great an all, except that it's entirely impractical in many of the places where typing currently takes place. In an office environment, we'd all be drowning in a cacophony of noise that would shatter our concentration if we had to type up every TPS report verbally. And we already know from cellphones and Bluetooth headsets that people find it hard to ignore others (not to mention annoying) when only one side of the conversation can be heard. Plus, there's also the fact that privacy goes out the door when you have to vocalize everything you're writing.

At best, speaking works for people that are in a personal space (e.g. a car interior, a private office, a basement under mom and dad), but not really anywhere else.

No, what we need is something else.

Re:Seriously?

[afidel]

Or carry a smartphone with a full 5 row keyboard, with swype I can probably do 10-20 wpm with corrections, with the hardware keyboard probably 30-40. I'm really quite perplexed by the fact that hardware keyboards aren't more popular considering that kids spend most of their time on the phone texting/IMing.

brain mapping dystonia

[ankhank]

This is guaranteed to screw up people's ability to accurately place their fingers.

Same reason I turn off AutoCorrect -- because when the user can just wave and poke at the approximate area of the keyboard -- and get the right letter supplied -- the brain fuzzes over its map of the keyboard and the finger placement becomes imprecise. Or rather exactly precise enough to get the desired result -- which is pretty damn sloppy when the computer's taking care of the final accuracy.

'oogle brain mapping dystonia -- lots of academic work on this, it's a serious problem.

Sorry, CMU, this is going to cripple people if you implement it.
Not right away, it'll take some time before the damage is apparent.

Re:voice control

[wcrowe]

Because you might not want everyone in the room to hear the message you're sending.

Yay technology!

[femtobyte]

Isn't it amazing how great the advancement of technology is for device usability! Now people can type at an amazing *9.3* words per minute! It's hard to imagine how we ever got by back in the old days, when a casual typist could only achieve 30-60 WPM --- uphill through the snow both ways barefoot. Progress!

Re:Yay technology!

[drinkypoo]

All the advances in wristwatch IO are fixing problems created by insisting on doing stupid shit on a wristwatch that a wristwatch isn't any good for.

One day we'll all have natural voice DWIM interfaces and then this problem will be moot. Until then, some people will be trying to solve the problem of letting people make their devices do the things they want to do. It would be nice if we all had access to a big screen and all the input devices we could ever want no matter where we went, but until we can pull them out of a pocket dimension we're still going to want to use small devices to do jobs that could better be done with a large device if only we had someplace to store it that did not involve the rectum.

PS, so let the damn computer learn to read script

[ankhank]

If the computer's good enough to get the right letter out of a vague approximation of position on a mini keyboard, it ought to be able to read my handwriting.

Want to do input on a tiny little area or just by waving your hands in the air?

Penmanship. Just make the computer able to read handwriting.

This might be better than nothing...

[beaverdownunder]

but I'm sure hopeful there are better input methods than this!

This is the problem

[TheSkepticalOptimist]

On screen keyboards suck, period.

Having to tap several times to get to the key you want is a nuisance. This will be the one thing that ruins any new smart watch device, assuming you must type text on the screen using a keyboard to interact with it.

At some point you have to accept that just because it has a screen and processor in it along with wireless communications does not mean it has to become a social platform or even one that requires text input.

If app designers for new smartwatch devices are thinking about solutions for how to solve on-screen typing they have already failed. There is a reason why phones are moving to larger and larger screens because people find text input on smaller screens a horrendous experience. A small 2" watch screen is not going to have any adequate method for text input, so don't bother with the functionality.

I don't see why we need smart watches, but we need them even less as a social/texting platform.

Just write on the screen

[doconnor]

The old Palm watch came with a tiny stylus that let you write on the touch screen using their Graffiti system. A normal palm had a separate part of the screen for writing. The watch has some why of switch the screen from tapping mode to writing mode.

Re:Seriously?

[ArcadeMan]

Your wife knows you're at the grocery store, sends you a text to remind you of buying milk and you need to text her back with "1 or 2%?".

handwriting recognition

[highlander76]

For the size of the displays in the paper would a simple box for tracing letters and better handwriting recognition software be better?

Why qwerty?

[gmuslera]

Using the input method of a bulky device (whose letters were ordered that way to not let you write too fast to avoid jamming of mechanical parts) with fixed letter positions for very high tech, digital small devices, with no mechanical parts that could jam could not be the best approach.

Maybe entry could be arranged like in compressing algorithms, having the most common letters and words right at your reach (few bits/touches) and you could navigate to more uncommon ones that fits in your input. Or handwritting recognition, but with extended "alphabet" (where you can have different gestures for i.e. common words). Or hardware keyboards with soft keys.