What's neat about this is that it essentially takes a skill, and packages it up, and allows anyone who can turn a wrench to do the same job with extremely minimal training.

Not enough Marines in the motor pool to fix all the broken trucks? No problem, augment them with some cooks.

I wonder if with this and some other stuff( for example, diagnostics using car computer) , people could easily fix their car on their own?

So how do these systems work? Do the goggles have some sort of position measuring devices or do they work through image recognition of what's in the field?

It must be positional; image recognition would seem a little too good to be true.

Would be delighted to be contradicted, though!

I've been looking into augmented reality technologies lately; vision systems are getting really good. There are some good ones here (click twice to load Flash videos, or download AVIs):

http://www.robots.ox.ac.uk/~bob/research/research_ptamm.html

Thanks for the link. Fascinating stuff, makes me want to quit and go jump into AR.

Can anyone explain the importance of reducing overall head movement?

Maybe it's a rough indicator of how much looking is required to locate a part. Maybe it reduces the odds of cracking One's head on large metal objects.

DIY, internet guided surgery here we come.

Seriously though, there seems to be some very interesting potential for this type of augmented reality.

This seems like is might be awkward for mechanics who were all ready good at the job.

It seems to be aimed at military in-field repairs, though, in which case it probably makes sense.

even an experienced mechanic will tell you that you can never have too much reference material

knowing how a vehicle mechanically works is one thing, figuring out how a certain manufacturer implemented those mechanical principles is totally another

neato. I did some work in his lab 10 years ago. Back then the AR getup looked something like the Ghostbusters backpack.

With these maintenance tasks seemingly pretty well-defined in terms of movements and actions with 3D objects in a 3D space, it doesn't take much imagination to see a suitably equipped robot could be doing this kind of thing pretty soon.