One central feature of our real-time world is the flood of data it generates. It used to be only financial markets (stocks, commodities, futures) that generated this much data. Now things everywhere are generating continuous and growing streams of data. Cars, planes and trains generate data from their onboard computers; companies generate data from their internal application networks; satellites beam down data from orbit; and the Internet moves all this data from anywhere to anywhere.

It’s creating a huge need to turn data into information and make it easy for people to understand what’s happening. Systems can’t just show data dumps anymore. If it takes more than a moment to understand what the data means, then it’s useless, because situations change so quickly and there’s so much new data coming at us.

We need effective user interface designs if we’re going to rise above this swamp of data and make sense of it all. We need to quickly understand and act, not spend lots of time analyzing. That’s what it takes to be agile – just ask a stock trader – he who hesitates is lost.

There are lots of people who claim to be designers and they create lots of user interfaces for lots of systems. What makes some of them good and others not so good? Who’s to say if one design is better than another?

The reigning Grand Poobah of user interface design is Edward Tufte. In 1983 he published The Visual Display of Quantitative Information and he’s followed that book with several more and with many lectures and seminars since then (check out his recent critique of the iPhone user interface).

I’ve read his books and taken some of his seminars and I come away feeling like I studied at the feet of the master, feeling like I heard great truth, and feeling pretty overwhelmed about how to start using what I heard. So I supplement my learning by talking to working designers who apply his principles and some of their own insights to create real-time user interfaces.

One such designer is Roy Nakashima. In an email recently, he articulated five points he calls “inspirations for turning data into information”. I love short and well articulated lists of pointers, so I’m quoting Roy’s email here:

1.  Story Telling. I have to understand the task the user is trying to accomplish while they are looking at the GUI. If I can tell a story about someone using the GUI to accomplish a specific task, in the telling I discover how well the GUI met their needs. If you watch an old “Columbo” episode, you can see an example of how the detective uses story telling as a tool.
 
  2.  Visual Hierarchy. Not all information has the same importance. I try to reflect this in the visual presentation. More important information is emphasized by size, font-weight, position (top or left most), highlighting, contrast and color.
 
  3.  Composition. I compose a web page the same way an artist composes a painting, i.e. I pay attention to how my eye moves over the page. Then I try to optimize eye movement by putting certain things closer together and removing unnecessary visual boundaries that inhibit eye movement.
 
  4.  Rhythm & Structure. When I use repeating visual elements or grid structures, it creates a visual rhythm in the page that makes it easier for the user to anticipate where something of importance is going to appear.
 
  5.  Transitions. In architecture and

12Next ››

Keywords:

design; user interface; real-time data

You do not have flash or javascript support.

Average (2 votes)

Reply
Digg this
Slashdot
Reddit