Elevators & Usability

David Foster recently wrote a post about a new elevator system:

One might assume that elevator technology is fairly static, but then one would be wrong. The New York Times (11/2) has an article about significant improvements in elevator control systems. The idea is that you select your floor before you get on the elevator, rather than after, thereby allowing the system to dispatch elevators more intelligently–a 30% reduction in average trip time is claimed. … All good stuff; shorter waiting times and presumably lower energy consumption as well.

(NYT article is here) Foster has some interesting comments on the management types who want to use this system to avoid being in an elevator with the normal folks, but the story caught my attention from a different angle.

I recently attended the World Usability Day event in Philadelphia, and the keynote speaker (Tom Tullis, of Fidelity Investments) started his presentation with a long anecdote concerning this new elevator technology. It seems that while this technology may have good intentions, it’s execution could use a little work.

Elevator KeypadPerhaps it was just the particular implementation at the building he went to, but the system installed there was extremely difficult to use for a first time user. First, the new system wasn’t called out very much, so Tullis had actually gotten into one of the elevators and was flummoxed at the lack of buttons inside. Eventually, after riding the elevator up and then back down to the lobby, he noticed a keypad next to the elevator he had gotten into. So he understandably assumed that he should simply enter the desired floor there, figuring that the elevator would then open and take him to that floor. He typed in his destination floor, and was greeted with a scren that had a large “E” on it (there’s an image of this on the right, but the presentation has lots of images and more information on the evolution of the Elevator). Obviously an error, right? Well, no. Tullis eventually found a little sign in the lobby that had a 6 page (!) manual explaining how the elevators work, and it turns out that each elevator cab has a letter assigned to it, and when you enter your floor, it assigns you to one of the elevators. So “E” was referring to the “E” cab, not an error. Now armed with the knowledge of how the system works, Tullis was able to make it to his meeting (10 minutes late).

Naturally, I think this is a bit of an extreme case (though there were a few other bad things about his experience that I didn’t even mention). The system was brand new and the building hadn’t yet converted all of their elevators to the new system, so it seems obvious that the system usability would improve over time. There are several things that could make that experience easier:

  • In the image above, note the total lack of any directions whatsoever. It’s especially bad because the placement of the keypad implies that it only applies to the elevator it’s next to.
  • Depending on the layout of the elevator area, I think the best way to do this would be to have a choke point with a little podium that has the keypad and a concise list of instructions. This would force the user to see the system before they actually get to the elevators.
  • Once you use the system once and figure out how it works, it’s probably much better, especially if all of the claimed efficiencies work out the way they sound.
  • As the NYT article notes, there are some other issues that need to be dealt with. For instance, most groups would naturally like to ride in the same elevator, but this presents a problem to this system, especially when only one person in the group actually uses the system. There’s also some frustration with not being able to get on the first available elevator, though that may be mitigated by an elevator ride with less stops. You also can’t change your mind once you get in the elevator…
  • It seems to me that this sort of system would be ideally suited to an extremely large skyscraper with a high volume of traffic (like a hotel). Most elevators probably wouldn’t need to be converted, which means that most people wouldn’t be exposed to this sort of thing until they make it to one of the larger buildings (which also means that the usability for first time users will still be quite important, even though it gets easier to use after your first time).