Mobile augmented reality

Mobile augmented reality is a big new thing, if not the only one. From the perspective of this blog (which only cares about things that help to make transport more sustainable/less damaging to the planet), it ought to be quite useful; one of the limitations to using ICT to make transport better is the user interface. Mobile augmented reality aims to use camera capability, coupled with other information available to a smartphone phone (both from the device itself, like the accelerometer or a compass, and from the internet), to present information about the world to the user as a visual signal. Since we are good at processing visual information if it's presented in the right way, this ought to make it easier to access and use this information. An example is the iPhone app already described on this blog, which presents info on how to find a nearby tube station as virtual street signs on a camera view - easier than scrolling round on a map.

Layar has some nice ideas for personal transport Augmented Reality apps - including bus stop information (see the next bus and where it's going) in Dublin, a metro station finder and even a ship tracker. In the interests of accuracy in blogging, I must report that I've downloaded the application to my G1 from the Android Market (it's free there, though it's been removed from the iPhone App store because it's not working right), and I'm buggered if I can make it do anything. Still, I'll keep trying - it might be me rather than it.

There is a nice article about Augmented Reality in Qualcomm's magazine (which annoyingly seems to be available as a pdf download but not as an online article - what's the point in that?). This contains the following rather frank admissions: "Our industry has a tendency to set high expectations about the potential for disruptive technologies like Augmented Reality. We need to focus on talking about applications for which Augmented Reality delivers real user value, and which apps the technology can realistically enable."

Does he mean...? Yes, he does. "The existing crop of mobile augmented reality applications have been developed around the device's compass and accelerometer. However, because a digital compass is subject to error of up to 20 degrees depending on magnetic interference, the user experienc for these applications is not optimal...compass-based augmented reality suffers from two problems with augmentation graphics: They aren't well aligned with the underlying imagery, and they tend to drift and be bouncy or jittery...the compass-based approach will not enable tight alignment between augmentation graphics and the underlying camera image. This level of alignment requires computer vision algorithms and more processing power than is available in today's mobile phones."

Well, Qualcomm would like to sell phone makers more, bigger and better chips - but it's nice to see someone admit that something just doesn't work.