SlimSoft's Blog

A while back, Renata posted about a company called 37 Signals. For those of you that are customers of ours, you use a product by 37 Signals almost every day... our Project Central is a hosted solution provided by 37 Signals called Basecamp. 37 Signals is an interesting company who specialize in web development, and most importantly simplicity in web development. 37 Signals also maintains a blog called Signal vs. Noise which has become one of my daily readings.

Today's posting by Jason "Apprehension is the enemy of interface design" really hit home to me:

Go to your local grocery story or Home Depot, or whoever else has self checkout and just watch. Watch for about 10 minutes. You'll see the physical manifestation of apprehension caused by people interacting with machine-based interfaces.

You'll see lots of squinting, lots of "almost pushed" buttons, lots of hand sliding and gliding (when someone moves their hand around the screen seeking out the button the want to press), lots of slow motion ("the slower I go the less mistakes I can make"), lots of corner-eye looks to see what everyone else is doing, and lots of quick pull aways. It's as if the faster you pull your hand away from the screen the less likely the machine will think you meant to do something wrong.

The posting goes on from there to recommend some guidelines to solve these problems...

One of the comments to Jason's posting really made me chuckle... a common problem with these machines is how they rely on the weight of your goods to determine if you are telling the truth about what you have purchased. Sometimes they get messed up. He talked about how when he uses the machines, he found that a quick solution to this problem was to just put your foot on the scale as you scanned the items. Maxing out the scale solved the problem. While I don't recommend this technique to anyone, it reminds me of all the clever solutions I have come across in factories to "work around" deficiencies in software design.

Think about the above from the perspective of an HMI panel, or any other Plant Floor application. The problem is exactly the same. This goes back to my posting last month about Usability vs. Maintainability. Which in the end costs more money? An operator who struggles through every interaction with their computer, or the extra burden on IT staff to maintain the solution. As I said in my post on the subject, I hope that we can eventually get to the best of both worlds, but until then, my vote is with the operator. I think if a cost benefit analysis was performed, which measured the cost of downtime caused by mistakes, as well as the cost of inaccurate data resulting from a less than obvious user interface, that this cost would far outweigh the cost of maintaining the solution.

As we embark on new projects and opportunities, let's all try to take a step back and think through the who, what, where, when, why, and most important of all the how.