Saturday, October 21, 2006

Elevator logic

The fidgety man in a business suit repeatedly jabs the call button. And when he pushes both the up and down buttons, you know more about him than you care to know.

However, I contemplate not the circumstances of his childhood and life, but a much simpler question. Why don't elevator designers punish the behaviour by resetting the call status?

It should have several advantages:

  1. Pressing the button a second time could delay the arrival of the car by canceling the previous call. (It would also require pressing the button a third time, but some users apparently derive pleasure from such effort, so we won't bring it up.)
  2. You could cancel a call for the wrong direction, if it was a genuine mistake or if you changed your mind.
  3. It would save wear and tear on the elevator buttons.

Why don't we design elevators like this? Have I missed something obvious?

 

Technorati tags: , ,

3 comments:

jbum said...

I've noticed that the common misconception that street light buttons and elevator call buttons *might* work this way often causes people who press the button twice to press the button an additional time, just in case their second press cancelled the first one.

Scott Fell said...

The real problem here is user TRUST. The user continues pressing the button because he does not trust his request is being served. The user interface does not provide the user sufficient vision into whether his request is being served. (I use the word "sufficient" because presumably the button remains lit after his first press)

A display showing the elevator's current floor number would be sufficient to relax the user that their action caused the desired effect.

However, such displays are quite expensive compared to simply replacing the "down" or "up" buttons a bit more frequently.

Users tend to treat well-designed interfaces the way they should be used.

Another reason not to punish the user is that machines/computers DO malfunction from time to time. It is logical that the user of any system maintains some level of suspicion that their requested activity may NOT come to fruition.

Sunil Bajpai said...

Thanks, Scott, for a thoughtful comment.

You are right that users tend to treat well-designed interfaces the way they should be used. Therefore, a better way to design the interface for elevators, or for any thing else, is to begin with this premise.