Entries in Augmented Reality New York (1)


Registration in Mobile AR

Last week, at the first meeting held by ARNY - Augmented Reality New York, since this month’s are2010, Augmented Reality Event, I gave a talk dedicated to a tech trend that is emerging in mobile AR registration. It is based on the talk I gave at are2010, but focused specifically on the technology portion of the presentation.

Steven Feiner (see episode 1 of AR & Emerging Tech), whom we are so fortunate to have as a member of our organization — based on hands-on access to the PTAM source code and many years of experience as a pioneer in the AR space — expressed his opinion that this kind of registration will not be available for many more years. Not only is our organization fortunate to have Steve as a participant, but I’ve personally been fortunate to have him available to me as I fleshed out the ideas in this very presentation, so I find myself in the rare position of disagreeing with Steve. There are several reasons I feel this specific form of AR technology is going to come to market within the next 6 to 8 months (or less). I’ll give three here:

1.) The release of Apple iOS 4, and more specifically, the iOS 4 SDK (Software Developer Kit) has made public what were previously private APIs, now available to developers. This means that there is: A.) a capable consumer hardware platform with significant user-base. B.) With an infrastructure in place to monetize it (App Store).

2.) Between Google and Microsoft, and several other smaller players, we now have an extensive portion of the world, especially urban areas, mapped. So the reference mapping is now in place.

3.) The incentive for 3rd party software developers to crack this quickly is immense. The competition to bring this to market is fierce. The reward, huge. Several players in the space including Earthmine (who has released an SDK for incorporating their tech into iPhone apps), Tanagram (who have filed recent patents and claim they’ve solved it), Wikitude (who recently released computer vision based Wikitude Drive for the Android, and are partnering on Mobile AR work with IBM… possible acquisition?), Metaio (recently released mobile feature tracking Junaio Glue) are all hustling to bring computer-vision into the mobile space, and Earthmine and Tanagram are both pursuing this very form of computer-vision-to-mirror-world implementation of mobile AR registration.

No doubt many other upstarts are working furiously in garages around the world — the first to pull this off for the iPhone will stand to make a lot of money. This is no longer a competition between universities at academic conferences. This is now being driven by market forces and financial incentive. As of two days ago (the release of iOS 4), the hardware to support it is widely available and the APIs are accessible.

I believe we are on the cusp, with implementations of this sort on the immediate horizon, in months not years. Even if Steve disagrees with my timeline, I’m sure he secretly hopes I’m right.