Oblong's g-speak Brings "Minority Report" Interface To Life

tracheopterix writes "Oblong Industries, a startup based in LA has unveiled g-speak, an operational version of the notable interface from Minority Report. One of Oblong's founders served as science and technology adviser for the film; the interface was an extension of his doctoral work at the MIT Media Lab. Oblong calls g-speak a 'spatial operating environment' and adds that 'the SOE's combination of gestural i/o, recombinant networking, and real-world pixels brings the first major step in computer interface since 1984.'" The video shown on Oblong's front page is an impressive demo.

Comments

gorilla arm

[Anonymous Coward]

Gorilla arm.

That is all I've got to say.

Check the jargon file if you don't understand this.

Re:gorilla arm

[SanityInAnarchy]

Well said... I thought this comic illustrated it well, also.

Nice

[bb84]

...but until it shows me the future I won't be *too* impressed.

Not impressive at all

[avalys]

Actually, I call that an extremely unimpressive demo. It is a lot of technology with little purpose. In that entire video, what are they doing? Just spinning a bunch of pictures around.

Without a compelling application that requires that interface, it's a just a big, expensive toy.

Re:Not impressive at all

[plantman-the-womb-st]

Indeed, controlled with a Power Glove no less.

Re:Not impressive at all

[Iamthecheese]

Medicine, 3-D rescue mission/fire control mission planning, biology, CAD, art, anything with complex data sets, physics, movie editing, and 3-D movie creation come to mind. The intuitive 3-D control will allow whole new interfaces.

Re:Not impressive at all

[Louis Savain]

Are you kidding me? This is the future interface of parallel programming, among other things. Rotate'm, push'm, pull'm, drag'm and drop'm. This technology will allow us to walk inside or fly through our programs and quickly create and/or modify them through trial-and-error. Kinda like the way an interior decorator might rearrange the furniture and colors on the walls. This is the beginning of the end of keyboards and mice and typing. Add a voice recognition interface and this shit is going to kick ass. It will turn users, gamers and developpers into magicians.

Re:Not impressive at all

[TheModelEskimo]

Agreed. These people are demonstrating something almost completely useless while I use a very traditional method - text entry via keyboard - to learn programming in a console. And I'm a 3D illustrator.

People keep harping about 3D visualization being the next big thing, but while these awkward, hammer-seeks-nail inventions come and go, simple things like the classic terminal are *increasing* in popularity, if anything. New Linux users and experienced Mac users are saying things like, "actually, I just use the terminal to do such-and-such a task; it's faster that way."

Re:Not impressive at all

[Anonymous Coward]

I love the Power Glove. It's so bad!

Re:Not impressive at all

[Xiph]

I claim that this will be great for gaming, i already want to make games for things like this, seeing this video does nothing to remove that.

I also think this expensive toy will be great for things that requires complex data to be handled fast.
That's what gestures are good for, complex objects, needing complex handling, instead of going into a menu->submenu->item, click.
They're nice in the same way as keyboard shortcuts, they reduce strain, but can't be used for everything.

Gestures are great for somethings and really poor for other things.
This system is partly a system for gestures,
partly a system of semantics of the various gestures,
and partly a system for using these things over an arbitrary amount of screens(dig about a bit on the website).

I think that for some uses this will be awesome, for others it won't work. Don't do programming or other text-centric things on this system.
I have no illusion that talking will ever replace typing.
Just like I don't think the Wii will replace me going outside to play soccer with my friends, Or that an OMNIMAX will stop me going to beautiful places.

Re:Not impressive at all

[Smivs]

And while your 'free' hand is busy, what the hell is the image going to be doing?

Re:Not impressive at all

[rusl]

Well, I do think there are some interesting possibilities in that thing where they interact with the topography using the cut out shapes.

However, I too was thinking about my love of the command line. Basically, as they claim, 2D interface came along in 1984. It basically still has a lot to be worked out to make it useful. I do prefer point and click for many things there the command line options are just too complicating. It's easier to cut and paste 5 random files from one place to another than to find some common search thread between them all or to type it all out... Basically only simple things work well this way.

So the irony is that - it takes a long time to make a GUI useful, it will take longer to make a 3DGUI useful, and we've really only been able to work out the simpler applications (moving files, grouping non-text similar things) so it's like this technology will be only useful for very simple tasks for a long while yet... All the new and advanced stuff is likely to stay only really efficiently done via text commands. So the new fancy GUI will serve up old simple solutions. Probably the making of the complicated fancy GUI will be all accomplished using command line text =)

Re:Not impressive at all

[hairyfeet]

In most if not all of those you mentioned having a 3d view would barely get you halfway there. The problem is you need to be able to TOUCH, I mean really touch, to truly interact. And that is where we really suck right now. Because all of the sensory feedback devices I have seen so far including the really high dollar still in the testing phase ones, really only give you soft/hard. They can't give you warm or slippery or squishy or cold or kinda bumpy or.....you get the idea. We humans pick up so much by touch that we simply don't realize, and when you cut us off from those sensations we can still work but not nearly as well.

That is why IMHO this stuff will never be more than kind of an "ooh cool" kind of expensive toy or for really really tiny niche roles until we can interface the brain directly. Because trying to simulate all the things we can gleam from simple touch would be just to insanely expensive to ever be practical. But if we can figure out a way to send the data to the brain directly, either by some sort of implant or perhaps through sensors on the scalp, then we don't HAVE to come up with a physical way to fake all this data, we can send it to the brain directly. It would also get rid of the "gorilla arm" problem as you wouldn't have to wave your arms like a maniac trying to work since you could simply manipulate the data with your thoughts, or even basic eye tracking.

Call me crazy, but I think that an interface controlled by the mind could really give us a great leap forward. Even typing this post think of how much faster it would be if my thoughts simply appeared on the page? I guess it is because all these oversized 3d interfaces just seem like overkill, like a holodeck. I know the Star Trek fans will kill me, but let us be honest: holodecks are dumb. You are wasting all that space and energy to give ONE person a little fantasy land to play in. That is really really dumb. When I saw the Voyager episode "Equinox" I thought THAT was what a holodeck would really be like. Instead of wasting all this energy trying to create a physical simulation to interact with just send the signals directly to the brain where they can be experienced with minimal power required.

Maybe it is just me, but I think this thing too is going overboard with trying to give physical interaction, when it is mental interaction that we should be striving for. But it does look like it would be fun to play with for a half an hour or so, or until your arms feel like falling off, whichever came first.

Re:Not impressive at all

[- r]

not that i post that much here, but - this is the coolest thing iv/e seen in ages. *not that it applies to us as programmers*, but it does to our users. yes, i use terminal on my imac for programming, but not for seeing the result. i think someone out geeked the geeks here...

Re:Not impressive at all

[baggins2001]

Yeah but who would have thought that people would by teleconference rooms. I think it's a nice impressive toy, but someone with a lot of money (company money) is going to decide they need it to impress customers. I can already see someone swapping around Impress documents during a meeting. It'll happen, it'll make no sense, but it will happen.

Call me a luddite but I'll stick with 2D interface

[syousef]

I really don't want an interface where I have to gesticulate at a computer, while repeating words so the speech recognition engine picks them up correctly and moving cursors around with my eyeballs. Hell I don't even want 3D desktops and transparent windows - take all the damn effects away, and leave me with the folder metaphor, current UI for editing text and pictures, and a command line. These interfaces don't give me any new capabilities, and anything that requires more effort and doesn't empower the user is a waste of time. They aren't revolutionary - they're not even good sci-fi. They don't belong to the future, because the future will be built on interfaces that are MORE not less convenient and do actually give new capabilities. Good sci fi are things like the star trek communicator (not so different to today's mobile phone, or a walkie talkie of old, and were used to enable the characters).

cheapscates

[robi2106]

you mean to say, a startup centering around hi tech advances in visual interfaces.... can't afford to host their own demo? They have to go to the upscale HD version for YouTube to host the content?

Common. Get a real hosting account and a guy that knows how to embed JW to play your flash video.

Can you say.... gimmick?

[Ecuador]

Yawn... Another one of these. Why do I feel I read a /. article about "Minority Report interfaces" every week? And it would be interesting if we were talking about pre-cognitive interfaced etc. instead of the useless "do your best traffic officer impression" to move some videos around.
Yeah, IWTFV (didn't actually RTFA that came with it) and I guess it would be kind of cool for people who are not Real Geeks (TM). I especially enjoyed their "intuitive high bandwidth access to information" where they navigate this seemingly enormous 3D grid of what looks like boxes containing... the same japanese character! Yay, what a way to navigate through 2 bytes of info! Ok, maybe it is 1kb if the boxes were not identical, but there is no way to tell at a glance, as people who have tried to use lame 3D file managers would now. That scene also brought back fond cinematic memories... It's a Unix system! I know this!

The mouse is still better.

[wild_quinine]

Oblong calls g-speak a 'spatial operating environment' and adds that 'the SOE's combination of gestural i/o, recombinant networking, and real-world pixels brings the first major step in computer interface since 1984.

I'm tired of hearing about all these things that will replace the mouse. The mouse will be replaced one day, but not until something comes out which is better, not merely cooler.

This minority report interface will tire your arms out in less than five minutes. I'm embarrased to admit it, but I use a computer for upwards of eight hours a day. Sometimes upwards of twelve.

The mouse is ideal in that your fingers have precision, the feel of pointing is natural, and crucially your hand, wrist, arm, are all more or less at rest throughout the process. Sure, you move them. But you don't hold them anywhere. It's a fundamentally different type of task from minority reporting, or wii-ing, or other stupid-but-cool flailing systems.

So no, I don't know what will replace the mouse. Something, eventually. If I knew what it was, I'd make a bloody fortune. But improving on the mouse will take a damn shot more work than making me say 'Wow', let alone 'meh'.

Re:The mouse is still better.

[wild_quinine]

That's the old fart talking. We've been using a mouse most of our lives, so sure it's better for us. Take a look at the wii, it took me a while to adjust killing zombies by pointing at the screen instead of using my thumb. If kids growing up use this interface, not only will it be natural to them, they'll be in much better shape.

The wii's controller is unreservedly terrible. The only thing it's good for is amusement. If kids grow up using that interface, I'll be surprised if they can open 'My Documents' by the age of thirteen.

I'm sorry, but waving your arms around and working an office job are just mutually incompatible unless you're a manager.

I look forwards to something bettering the mouse, just as I look forwards to each new technological advance. I don't buy into stuff because it's cool, which is where the kids go wrong. I buy in when something is better.

Wow! I want one

[Prikolist]

I want one! I will disagree with everyone here saying that it's useless. I'd trade the mouse, and pen tablet, and the joystick, and all the rest of those for this. Looks way more convinient - not to mention instinctive - to use. It's like a touchscreen but you don't have to leave greasy fingerprints all over. With this I could even actually draw on computer, while so far any attempts with mouse just ended up with wrist pain and frustration. And just moving the cursor, moving windows, anything... Oh, and games, this will send Wii to an antique museum.

Re:Wow! I want one

[rusl]

As an artist myself I've often wanted to draw on the computer too. I've never suceeded. However, I've seen a very skilled person draw on the computer. The way to do it is this: use your right hand to carefully draw with the mouse - keep your left hand on Ctrl+Z. It's a computer so no matter how many times you erase you won't rub through the paper. he was really good with it, albeit his drawing style was somewhat limited - slightly gestural if you know what that means. He would make lots and lots of marks and CtrlZ most of them. But he had a good flow back and forth and the constant motion allowed him to stop being tentative and really express the line through the mouse motions. Also you have to have a good mouse that doesn't stick or act up... obviously. I've never seen anyone with those digipens draw right into the computer like this guy did using CtrlZ.

This is his website. But I' have no idea what medium he used for any of his drawings there... Suffice it to say I wouldn't be surprised if it was drawn via the mouse. He was that good with it.

Anyone remember that OCZ mind controller thing?

[w0mprat]

I see this having huge potential in CAD & design applications. Spatial controllers for CAD I've found to leave much to be desired. Gestures and natural motion are a huge improvement. This paradigm of interface will all hinge on a killer app, sure the engineering has been done and from what I can tell it works, effectively, but there are so many brilliantly engineered ideas that are simply nothing more than that.
Implementing a Good(tm) product, and getting a market for it is a whole different story. I would expect to see this kind of thing first coming to market as a expensive niche product for CAD/VR visualation set ups, or perhaps being bundled with a game that supports it. Many of these new things never get off the ground, not at least until the price/performance ratio reaches a point it becomes compelling.

Anyone remember that direct mind controller thing from OCZ? No? This'll be forgotten too...

Modern UI, I'm all for it

[deodiaus2]

Besides, I am still having hard time operating this mouse foot petal. It is so damn hard to get the selection of a word with my toes! Next thing you know, they'll design away my CD-RW coffee cup holder! I still miss my D parallel printer, what am I going to do with all the cheap cables I got at the discount bin at BestBuy!

Re:Call me a luddite but I'll stick with 2D interf

[syousef]

Not everyone thought the mouse was a good input at first. This type of UI may have speed advantages as well as visualization advantages we may not completely see yet. CAD comes to mind here. But I suppose ASCII art CAD is enough for some people :)

Show me speed advantages (without significant disadvantages in other areas) and I'll be pleased to accept change. In the meantime my office is enough of a nightmare without people gesticulating and yelling at their computers like Italian villagers.

I think speed advantages in CAD are more likely to come from more intuitive tools in the software. Right now doing anything in a CAD/CAM package or 3D Modeller requires specialized training. Whereas I could teach someone to edit (ie. touch up) a photo, or teach them to to use a word processor for something basic.in minutes.

I have used it, and it is impressive

[Anonymous Coward]

It is not just a gimmick - they have worked on the gestural language as well as translation software, and it works well. The glove is a bit of a bummer, but it is just a passive glove with spots the system can read. They already have clients, yes big data sets of SHARED computing environments, something that is being overlooked. But it will be quite some time before we have it on our laptops, probably on our TVs before that. And, yes, it will be a better UI than the mouse or accelerometers or voice for many things. But the future is a mixed environment not one single solutions.

Not a programmer, are you?

[Namarrgon]

I can't imagine a less efficient way to get any actual work done :-(

Apart from the arm strain, I think that saying, "if open-parenthesis p-underscore-temp-var-x-y-z-b-b-q close-parenthesis newline open-curly-brace newline temp-var-x-y-z-b-b-q equals asterisk p-underscore-temp-var-x-y-z-b-b-q semicolon newline close-curly-brace newline", more than, say, once, would engender homicidal rage.

Re:Not a programmer, are you?

[theheadlessrabbit]

I can't imagine a less efficient way to get any actual work done

tried Vista?

Re:Not a programmer, are you?

[L4t3r4lu5]

"Select Open parent this is p under score temp bar sex why z be be queue close apparent cheeses!"

Re:Not a programmer, are you?

[Bwian_of_Nazareth]

Oh, come on, how is this insightful? Since when is "actual work" equivalent to typing code? Just because something is not useful for one purpose does not make it useless. I don't think anyone even suggested that this is cool for coding or that all input devides should be replaced with this one. Sorry, but your comment is so narrow-minded that I would consider it "troll" rather than "insightful".

Uh huh

[bm_luethke]

OK, maybe this is the wave of the future. I will not say it isn't - but that promo didn't sell it. It looked like what they claim to be - based on a Hollywood custom script. I want to see how I would use this in the real world - I'm not going to be standing around and moving those text blocks around, nor did I really see why having that matrix of Asian language characters (I don't know which language - I can't read any of them) in that grid would help someone deal with the massive amount of letters anyway. It seems to me since most of them are based on pen strokes that that the arrangement is - hmm - only made to be visual appealing to westerners (which I am one of).

I had used an SGI CAVE a few years back for a few different things (well, others in the group I worked with wrote the stuff - I played with it simply because it was neat) and I see many similarities. Given that products history I do not see that as a Good Thing for them. In fact they seem to be a good 5-10 years behind the curve - the last time I used one was five years ago and they were already doing all this nice stuff from what I can see.

It was really good for things that were meant to be visual. For instance they had this really neat data set of a human (some convict that donated their body to science) and you could interact with a 3-dimensional representation of them. Their body "displayed" (or rather appeared too) in the center of the CAVE and then you could select (using a wand that the system kept tract of it's position in the room) a "window" and move/drag it around and see just that slice of the body in a high amount of detail. You could lock that and have as many 2-d slices going through the body as you want.

They also had a car wreck that you could do a similar thing - but you watched the "slice" as the wreck happened in real time. They actually crashed a car to get the data.

There were also quite a number of specialized tasks that benefited from it and I still run into some today.

But, other than that we pretty much played quake on it. Why? Well most data doesn't really need that type of visual representation. Our current screens work quite well and you are simply adding overhead for the heck of it. Even for those that the system worked well for they still did OK on a normal screen. A large monitor costs a few thousand, these systems cost a few hundred thousand. Well, you should get the picture there (and knowing that I worked in a govt research lab at the time should tell you why no one cared that it was a few hundred thousand more).

This system has the 3-d input but not the nice 3-d output that the SGI systems had so I can't see it working any better - it is just as specialized hardware intensive and I bet just as expensive. Even if it isn't - is the increased productivity for those specialized application going to be worth the cost? I also bet not.

You will note that even a group that has quite a bit of experience making true Hollywood scenes couldn't come up with better. Perfect for massive data - uh huh - and what did that wonderful things you show of arcs moving around *really* give you? You mean where you put a circle over one of the other circles and it turned yellow?

Is there *any* reason whatsoever that the majority of that could not be accomplished with a mouse and a large LCD? Nope - so why purchase this? At least the pretty much failed SGI stuff had the whole 3-d output to go with it - and trust me, there is no experience in the world like playing quake in a fully 3-d environment that you are freaking standing in the middle of and the virtual gun actually is being held by your hand. But then - how many are going to pay 250k for that?

This type of thing is so 1990's and dot com - ten years ago these guys would have been flush with cash from countless venture capitalist. Heck, their video even screams late 90's and early 00's. As is they better really be able to back up the claims they make to even have a shot at it, let alone be truly successful. I didn't particularly see anyt

Re:Uh huh

[zwei2stein]

The g-speak platform is in use today at Fortune 50 companies, government agencies and universities. Application areas include:

        * Financial services
        * Telepresence
        * Network operations centers
        * Logistics and supply chain management
        * Military and intelligence
        * Automotive
        * Natural resource exploration
        * Data mining and analytics
        * Medical imaging
        * High-touch retail
        * Trade shows and theatrical presentations
        * Consumer electronics interfaces

Oblong delivers room-sized and single-user g-speak environments as turnkey products.

A software development kit that runs on both Linux and Mac OS X is available. Applications are source-compatible across both operating systems and can run on ordinary desktop and laptop computers in addition to gesturally-equipped g-speak machines and clusters.

You were saying?

Re:Call me a luddite but I'll stick with 2D interf

[supernova_hq]

I don't forsee this technology being used on personal home computers in the near future.

Where I do anticipate (and look forward to) seeing it is for interactive public displays. It would be a very cool interface to have for a 3d map and directory in a mall or an informative display at a museum or aquarium.

As for home use, it could be used for family gatherings and birthday/wedding parties. Set it up with your DJ software and photos, then let your guests check out photos, pick out music to play, etc.

Most wedding parties, etc have a slideshow going on, why not let your guests upload photos as they take them, add them to the slideshow, maybe browse through them manually, change the background music, etc. It could make for a very entertaining and rewarding device to have for such occasions, even if you just rented them out.

Not from Minority Report

[Animats]

Actually, that idea first appeared in film in Johnny Mnemonic.

Autodesk put considerable effort into virtual reality in the late 1980s and early 1990s. The hope was that it would make it easier to design 3D objects. It didn't. The fundamental problem is that positioning your hands precisely in free space by eye, not touch, is slow and inaccurate. It looks really cool, but it's like trying to do precision work wearing mittens. Humans are much more precise when they have a surface to work against.

It's not a technology problem.

Comic is on topic

[TheLink]

How's the comic offtopic?

Back in my school days, one form of _punishment_ was being made to hold your hands up or out for many minutes. Imagine if you had to keep your arms extended for so long - talk about asking for a new set of RSI problems.

The full 3-D gesture stuff is overrated.

What would help me a lot more is the ability to quickly switch to a particular window in mind:

http://bugs.kde.org/show_bug.cgi?id=121349

Even if you don't have all your windows maximized, it would save a fair bit of time. Alt-Tab only works well if you are switching between two windows.

You can kind of do this on the Linux/BSD console but it's more limited. I'm looking for something like the text console but for the GUI and where you get to pick your "working set" of 9 or so windows from as many windows you have open.

Re:Comic is on topic

[mr_matticus]

You can kind of do this on the Linux/BSD console but it's more limited. I'm looking for something like the text console but for the GUI and where you get to pick your "working set" of 9 or so windows from as many windows you have open.

Sounds like a combination of Spaces and Exposé fits that bill exactly. KDE already has the multiple virtual desktops, and I'm sure there's some Exposé clone for Linux out there somewhere.

Re:Comic is on topic

[dreamchaser]

You would be surprised at how hard it is to keep your eyes steady. Unless there were an easy way to turn it on and off, you'd find your cursor jumping around quite a bit.

Re:Comic is on topic

[empaler]

Try nigh-impossible.

Re:Comic is on topic

[jebrew]

I seem to remember a camera or two having the feature of auto-focusing where your were looking. Seemed to work fine.

Does anyone remember this? I pretty sure my dad had a camera that did this. A Nikon if I remember correctly...which I usually don't.

Re:Comic is on topic

[jebrew]

Ah, found it in Canon EOS

Not again....

[1u3hr]

Search " minority"

"Minority Report"-Like Control For PC
On November 8th, 2008 with 138 comments
An anonymous reader writes "A startup named Mgestyk Technologies claims that they have an affordable solution for 'Minority Report'-like PC control. They have...

Obscura Digital Demos "Minority Report"-Like Display
On August 6th, 2008 with 124 comments
Barence and other readers sent along word of a demonstration by Obscura Digital of a new technology it's dubbed a multi-touch hologram reminiscent of...

Touch Screen Tech Comes of Age
On February 3rd, 2008 with 78 comments
pottercw writes "Good summary of today's touch-screen technologies on Computerworld the obvious Apple iPhone and Microsoft Surface, plus projected...

"Interface-Free" Touch Screen at TED
On October 30th, 2006 with 194 comments
Down8 writes, "Jeff Han, an NYU researcher, has recently shown off his 'interface free' touch screen technology at the TEDTalks in Monterey. Some sweet...

Correct me, but are all these breathless announcements still vapourware?

I'm getting a bit tired of this bullshit. It was just a stunt, it looked cool but completely impractical. And it's not like "Minority Report" (2002) actually invented the idea, even in the movies. Off the top of my head, same concept was used in "Johnny Mnemonic" (1995), Disclosure (1994), "Hitchhiker's Guide" (1978 (radio version)).

Not much progress since 1993...

[thbb]

Using datagloves, I did quite a bit of work in 1993 to see how the sort of UIs that we see in the Minority Report could work.

It turns out that there are 2 issues to overcome:
- Fatigue: the gesture vocabulary had to consist only of short sequences.
- "immersion syndrome": whatever I do can be interpreted against my will.

By designing the gesture vocabulary so that it would require alternating tense postures and relaxed aiming gestures, it was possible to overcome those issues in a pretty satisfactory way. Tension is particularly important, as it conveys intention: if you stress "Go There", people (and machines) can detect the fact that you want something to happen, as compared to using a monocord voice.

see Charade: Remote Control Of Objects Using Free-Hand Gestures published in Communications of the ACM in 1994 for more details.

--
The machine was rather difficult to operate. For years, radios had been operated by means of pressing buttons and turning dials; then, as the technology became more sophisticated, the controls were made touch sensitive ... now all you had to do was wave your hand in the general direction of the components and hope. It saved a lot of muscular expenditure of course, but meant you had to stay infuriatingly still if you wanted to keep listening to the same programme. D. Adams, The hitchhiker's guide to the Galaxy, Chap. 2. 1979.

What are the power requirements?

[oDDmON oUT]

Given that the new paradigm is "Reduce, reuse, recycle", how does a multiscreen, multi-projector, multi-everything system reduce my carbon footprint?

No, really, I'm curious.

It's about the 'real-world pixels'...

[TechnicolourSquirrel]

Ignore the dorky gestures and focus on the 'real-world pixels' -- pixels that are aware of not only their coordinates on a digital surface, but also their coordinates in the room at large. This is the big leap forward here, not all the arm-waving. Try to see the whole, bud.