VoxBlog

I was at the local home improvement store the other day and saw a “talking smoke detector”. The idea behind it is that instead of a loud beeping alarm when it senses smoke, it actually says “Fire! Fire!”.

My In-Store Reaction

It reminded me of many user interface decisions I see: something that is done because you can and because it is fun to do, but in the end is a waste of time and effort. My guess was that someone at the smoke detector manufacturer said “hey, we can make this talk, why not?”. The exact same thing happens in software shops everywhere when developers see  gradients, animations, and flashing icons in their toolbox.

There are certainly times when voice prompts ate useful and there are certainly times when gradients and animations are useful, but when a simple alarm is expected and understood, stick with the simple alarm.  Adding complexity simply adds possible points of failure and problems.

But My Reaction Was Not Correct

Further research has pointed shown that the problem is not so simple.  Studies have pointed to the fact that for very heavy sleepers (such as children), a voice message works significantly better than an alarm, especially when paired with the sleeper’s name.

In the end, I see the talking smoke detector as a study in simple and unexpected UI improvements that can make a huge difference.  I doubt that few people would have said “I think there is something better than a blaring alarm for waking people and warning them of a fire”.

UI improvements are often seen as fluffy, personal-choice things, but they can (and can not) make a huge difference.  The trick is to determine which of those changes help, which make no difference, and which hurt.