Advanced Handsets Need Advanced UIs
By Tom Hume, Fri Apr 01 08:45:00 GMT 2005

As mobile phones become more capable, people are using them to store an increasingly wider variety and greater quantity of data. This raises a new problem for designers of handset user interfaces: how do you let owners find what they're looking for in a coherent and friendly manner?

Historically, mobile phone interfaces have been menu- or icon-driven: users select from a number of choices displayed on-screen and repeat until they find what they're after. While the phone was primarily a device for person-to-person communication this worked well. An address book, message inbox and call log are clearly linked strongly to one another and happen to map to everyday, real-world artefacts.

But analogies with the real world start to break down when your handset starts storing your photos, video clips, media you've purchased, payment details, account history, WAP sites, books and notes. Some of this information is well-structured and understood (such as call histories), while it's very difficult to perceive structure within others -- for example, handsets can't currently tell you which of your contacts is pictured in any given photo. The situation will become even worse when handsets deliberately start blurring the distinction between data held locally and that accessed across the network. It'll be all too easy to confuse end-users.

One strategy for dealing with this problem is to bring search capabilities to handsets. Qix, from Zi Corporation, is an excellent example of such an approach: start typing on your handset and it immediately looks through all your data for matches with what you've entered. So, if you've started typing a name which is in your phone book, this name is displayed alongside the number you're tapping in. Links and media are similarly indexed, leading to a remarkably thought-free experience -- the phone seems to know what you're after and bring it to you. Users of QuickSilver on the Mac will be familiar with how liberating this sort of interface can be. The really neat thing about it is the way it doesn't require the user to structure their data in any predetermined way, but rather allows them to quickly see across any structure they happen to be using: so if you don't use surnames because your address book is full of friends and family, for instance, that's fine.

MotionBridge offers a similar product targeting operator portals, which fills out keyword choices for users based on portal content, past usage, and so on. It's a quick way of short-cutting menus, which are fairly unwieldy on a mobile phone. Instead of users having to read a list of options, choose one, click repeatedly to navigate a highlight down to it and click again to select, they just type in what they want. It's a similar principle to that of Qix, but applied online rather than on-phone. In both cases it's interesting to note that the primary method of navigation is entering text on the keypad; perhaps we've been trained sufficiently by text messaging for this to finally be viable.

Another strategy is to make interactions with mobiles more expressive than current interfaces using keypads will allow. The most common way that this has been achieved to date is through the use of voice commands to trigger calls or other actions on the phone, but use of accelerometers and touch or tilt sensors will increase the range of ways in which phone owners can tell their handsets what they're after. Following the lead of the Sony EyeToy, even the camera embedded into most new mobiles can be considered an input device -- take a look at games like Moorhen X, which uses the camera in Series 60 handsets to let players pan around a large game world.

And there's an underlying trend expressed in these products: the user interface, once provided solely by the handset vendors themselves, is increasingly open to customisation and improvement by third parties, and this openness allows smaller design and technology shops to translate their innovative thinking into real products. As technology from companies like Savaje hits the market and allows operators, handset vendors and third parties to all customise or rework aspects of the handset user experience, we can expect to see this trend increase; and as the scale and nature of data that we access through our handsets grows, new ways of presenting this data will become vital.