Last month I lambasted Nokia's mixed reality concept video for showing a future-woman using a pair of ludicrous augmented reality glasses. As it turns out, Nokia's researchers aren't sold on them either. "If the assumption is that everyone will need data glasses for this to work, then it's not going to work," says Kari Pulli, a mixed reality research lead at Nokia's lab in Palo Alto.
The glasses in the video below (and bracelet, and earphones) are interesting to Nokia researchers, he says, because they are one way to give users a layer of information over their world without destroying their social skills. "Using the camera as a 'magic lens' is awkward—you're holding your hand up all the time," he says. "If you want to be more discrete about it, you'd need... something like an outward-looking lens on glasses." But the glasses of this decade will be too big, too heavy and too annoying to charge, he says—to say nothing of the dork factor (speaking of destroying social skills).
The important part of the concept, Pulli explains, is the whole "magic lens" idea. The company has developed a few proof-of-concept apps to demonstrate the technology, two of which are Image Space (seen above), a track-your-trip breadcrumb app, and Point and Find, which allows users to get more info on real-life things like cars or movies by using their phone's camera.
But the video isn't being fatuous when it shows off those glasses; Nokia is indeed investigating gaze-tracking, or "near-to-eye" displays that they say will augment the mixed reality experience. Pulli says this outward-facing camera technology might ultimately be more practical built into a neck pendant, or embedded in a user's clothing.
But glasses or no glasses, Pulli believes there are bigger obstacles to achieving "mixed reality," which Nokia says is a superset of "augmented reality." Specifically, the servers and infrastructure that make it all possible. "There is a lot of additional work needed for recognition and tracking," Pulli says, "because we need more robust algorithms. We need to create the backend that can do all this matching. Only then can we make the killer application." The features that the video above demonstrates (which I called "useless" in my post last month) are what he calls "nice to have" features that would only arrive after the system can do the "truly useful" stuff like recognize people, landmarks and objects.
To read more about Nokia's MR/AR strategy, see a paper Nokia drafted in June 2009 on just this topic, embedded below.