Ready for a CyberWalk?

Roland Piquepaille writes "Even with recent improvements in virtual reality technology, it's still almost impossible to physically walk through virtual environments. Now, European researchers have started a project named CyberWalk and they'll demonstrate next week their omni-directional treadmill, named CyberCarpet. According to ICT Results, the researchers 'had to address five key issues: providing a surface to walk on, controlling the surface in a way that minimized forces on the user, developing a non-intrusive tracking system, displaying a high-quality visualization, and ensuring a natural human perception of the virtual environment.' The researchers think that their new virtual environments would be used by architects and the gaming industry." Additional details are also available via the project website.

Comments

CyberCarpet

[vrmlguy]

I'm interested in their omni-directional treadmill, CyberCarpet. I've tried to design something like this, but I inevitably wind up with a ten-foot (three meter, for our foreign friends) sphere that the user walks upon/within.

Re:CyberCarpet

[Ungrounded Lightning]

There was a science fiction short story involving just such a reverse trackball - though larger than ten feet - being used to trap a victim. Once on the ball, feedback rotated the ball to bring him to the center and keep him there, no matter how he ran, jumped, or dodged.

Not new

[nodrogluap]

We have a CAVE at our work, and the company that does the maintenance told me about and showed pictures of a 6 meter omnidirectional rolling carpet with VR display they built for the U.S. military. They were even trying to get wind and dust in the environment, but it would gunk up the works. It was probably a one-off though, whereas this new stuff seems to be going towards "commodity" vr.

Can fool accelleration sensors.

[Ungrounded Lightning]

Actually I'm pretty sure there's no way for this to convince someone it's real. You know that the ground is moving under you when you take a step so you'll always know you're not really going anywhere. Plus, people can sense acceleration.

Which is mostly done in the inner ear: Three "rate gyros" per ear (the semicircular canals) plus three linear accelleration sensors ditto (nerve cells with calcified masses on the ends).

But it turns out these can be fooled by elecrtostimulus from varying magnetic fields generated by coils mounted on a headpiece near them.

There has been at least one slashdot article on these - including adding them to a headpiece to mimic the head acceleration that would match a moving screen image to reduce "barfogenisis" and improve simulation game experience. Adding them to a 3-D VR simulation would be the next logical step.

With wall screens and projection onto the moving floor you could create the necessary visual illusion.

Gettin' there ... gettin' there ...

Re:Can fool accelleration sensors.

[tlhIngan]

Actually I'm pretty sure there's no way for this to convince someone it's real. You know that the ground is moving under you when you take a step so you'll always know you're not really going anywhere. Plus, people can sense acceleration.

Which is mostly done in the inner ear: Three "rate gyros" per ear (the semicircular canals) plus three linear accelleration sensors ditto (nerve cells with calcified masses on the ends).

But it turns out these can be fooled by elecrtostimulus from varying magnetic fields generated by coils mounted on a headpiece near them.

Actually, you can fool them quite easily, as any pilot knows. "Spatial disorientation" has downed many aircraft, usually because the pilot thinks the aircraft is doing one thing (flying straight and level), when it's really doing something else (descending/turning, or ascending/turning leading to stall).

One trick done during training is having the instructor do a maneuver like a turn continually for a minute or two, then ask the pilot to level the aircraft (by feel). The end result is quite... interesting. Let it continue for a minute, and the aircraft will be in a very strange attitude. Another thing is to simply let the pilot fly blind, by feel, and see what trouble it leads into.

It's why IFR pilots must trust their instruments and not their gut feel.

I don't see why you can't apply similar tricks into this to make it realistic. After all, those simulator rides feel pretty damn real even though the cabin only moves a few feet each way.

Wii 2?

[billy901]

The minds at Nintendo will quickly pick this up and expand on their Wii. People thought that the Wii was far away and would cost thousands of dollars. I'm sure we're on a pathway to playing ruining our fancy new plasma televisions by running into them! How will we create a safety strap for this one?