Some of the nicest AR interaction revolves around “tangible interaction” - augmenting physical objects that you can touch and hold and feel, where the augmentation responds to your handling of the object.
One of the great unappreciated advantages to “magic lens” style cell-phone AR is the screen acts as a “tangible proxy” for interaction with the virtual objects. If we want to interact with augmentations that don’t correspond to physical objects using a heads-up display we run into trouble.
There’s no such thing as “intangible interaction” - even in the folk record, when we’ve “interacted” with intangible things like “spirits”, we’ve resorted to tangible proxies like ouiji boards. The purely intangible stuff, like psychic visions, tended to be read-only too.
Most of the scenarios in this AR concept video try to solve this problem by projecting the AR elements on to a proximal plane (desktop, palm etc) that can act as a tangible proxy, but this leads to some trade-offs - the biggest of which is taking a 3D technology and limiting it to (sometimes awkward) 2D interaction.
Maybe this is the best we can do with HUDs without kitting ourselves out with haptic handware and giving ourselves gorilla arm or tennis elbow trying to interact with things that just aren’t there. Or maybe this is the killer flaw with HUDs - the “Holy Grail” of AR - you can see, but you can’t touch - HUDs are just a tease.
While the natural feature tracking on the new Ben&Jerry’s MooVision mobile AR app is impressive, the whole experience looks like it isn’t. Augmented Planet are fairly positive about the campaign for many good reasons, but I think we’re already at the stage where just augmenting a neat looking 3D model on top of something no longer provides any real “wow” moment (however technically hard it still may be). Expectations are so high from years of “faked” AR in movies, that simple models aren’t going to cut it, and a per-brand application landscape is not sustainable.
The application seems designed to a) giving you visual feedback that it has identified the tub so that b) you can then link to otherwise normal HTML web-content, where the real information lies. I have no problem using AR as visual-search with graphical feedback - I think that’s a great use-case for AR.
I’d just really like to see some actual information presented in the AR view, with some AR-dimension interaction. They could have made the words “New York” link to the city when pressed, instead of using buttons for interaction, or added a small map or globe indicating where in the world their organic cocoa comes from that I could touch to learn more. I’m just throwing these out off the top of my head, I’m not arguing that these are sensible AR interactions, but I think it’s time we worked on figuring out what sensible AR interactions are. We should really get around to solving the very difficult user interaction and user experience challenges before everyone decides our technology is a fad.
