It's just eleven years since the infamous sci-fi film the 'Minority Report', set in the year 2054, impressed us with thought-provoking futuristic technology. This included what could be quantum levitation, computer guided cars, voice controlled basic appliances, electronic paper and sick sticks.
While many of these ideas are in the process of being brought to life - hopefully all but the sick sticks - the idea that has arguably received the most attention is the gesture controlled interface used by Tom Cruise's character. Since the Xbox Kinect arrived and brought gesture control to the mainstream audience almost three years ago, the race has been on between start ups to innovate the most accurate and convenient replication.
The Leap Motion is a freshly released attempt to conquer this market by packing three LEDs and two cameras into a surprisingly small desktop-worthy enclosure. Quite obviously, we had to get one.
After plugging it in and installing the software to get it going, the Leap was very easy to get working. Visualisations showed my precise hand movements without any lag, and all was going great until I installed a couple of programs designed to recreate some of the gesture support that my inner geek has been eagerly waiting for. Frustratingly, they were promptly uninstalled again after finding most tasks took 10x longer than they should. This was due to the fact the Leap was struggling to follow intricate gestures required for basic tasks such as clicking.
After more play, I found the Google Earth plugin took some time to get used to. Although, while being almost impossible to master, it is an intriguing new way to navigate the earth with a single hand. But my favourite app so far is 'Cut the Rope' - a popular game ported from mobile devices. Here, the Leap is used simply to cut rope by swiping with one finger. Currently then, it seems the best use for this gesture control is a simple one.
After a quick read through the JavaScript API I was surprised how straightforward they've managed to make developing with gestures, and I had a quick go at making my own very simple Leap controlled apps.
The first was a game comprising of a mediocre mismatch of the classics Block Breaker and Space Invaders, built in HTML5 Canvas (see it here). While the game was nothing to get excited about, building it did provide an good insight into the built in gestures. This game used 'pointables', which looks for any fingers or tools and provides data relating to its movements. These worked very well while one finger was correctly detected, although the Leap would often mistakingly find more objects apparently imperceivable to the human eye, making the experience very glitchy.
In my second experiment, I set up a tool to be able to write with a 'pointable' (see it here). This turned out to be slightly more successful. Though it still has a bit of a learning curve for whoever is using it, after a some practise it was possible to draw (basic) shapes. So to extend it, I tacked on a 'swipe to clear' feature, to give an easy method of starting from scratch without refreshing the website. This worked fantastically, and usually even when you needed it to, but all too often I found my beautifully crafted artwork destroyed by a trigger happy event. I fixed this by swapping the swipe event for one that would clear the canvas when you show the screen the palm of your hand. Not exactly intuitive, but it worked.
In all, the Leap Motion is a very impressive device and provides a great stride forward for updating the mundane but dependable mouse. It still has many kinks to iron out to become usable on a day to day basis, though in its current state it could still have its uses: think primitive public interaction with far less risk of breaking the inputs, or art installations where precise gestures aren't required.
Finally, Myo, Mycestro, Haptix and many more are all fighting for a slice of the market, while I'm sure improvements to the Leap software will also be released. The mouse isn't safe yet!
i love this approach ... freedom to code the front end of a site without worrying about the back end has got to be the way forward. There's so many ways to decouple these two now.
Unreal Engine ported to Javascript using asm.js and emscripten.
It's shame that the guys at Mozilla didn't share it in better quality but it's still very exciting to see such performance in the browser without plugins like Flash or Unity, using only Javascript and WebGL.
A new chip has been developed with low power usage that will allow gestures to control devices such as smartphones. The device uses an electrical field to detect gestures which uses much less power than previous camera-based solutions.
Yes, you can now play ping pong (wiff waff for the older generation) whilst waiting to cross the street - really smart little touchscreen addition to the boring old street crossing, if nothing else, just to tell you how long you have to wait!
A good check to make when building cross platform - how heavy is your page on a mobile device (you might be loading a whole load of crap you don't need, for example).
There's a million robots working in factories, but they're way too expensive for most companies ... until now: http://www.economist.com/node/21563705?fsrc=scn/tw/te/pe/baxtergoestowork
