While thinking about how to improve the Polder 989-90 Electronic Clock, especially in setting the time, some ideas hit me.
As I wrote previous article, it was extremely hard to find out how to set the time in this clock. That's because setting the time is very rare operation. For me, it was my third operation in several years, first one was just after buying the clock and second was after arriving New York from Seoul. Maybe I'll not do this task before changing a battery.
The most impressive thing I've learn in undergraduate was engineering methodology rules from computer architecture class.
Rule 1: Identify and optimize the common case
Rule 2: Make the rare case correct and reasonably fast
Although these rules may not a perfect fit for improving design of the clock, I got some ideas. Definitely, setting the time is a rare case. However, users should be able to do this task reasonably fast w/wo the manual since they don't have it in many cases.
Let's see the notation for setting the time of my previous article.
↓ press ↑ release ▽ press 2 seconds s- seconds to complete
Option 1:
[Cl▽↑]2- [S↓↑]1- [1↓↑][1↓↑][5↓↑][9↓↑][0↓↑][0↓↑]3- = 16 operations / 6 sec
I've missed very important thing in here. It can be called "Trial and Error" phase. Before noticing the right way of setting the time, I made so many trials and errors. In this case, proper labeling or button adding will recude the time and effort of "Trial and Error" phase dramatically.
How about put a "Time Setting" button near a battery cover in back side since it is rare operation and usually needed when changing a battery? In addition, put "AM/PM" label on existing button in front side for toggling AM and PM.
I think that what we discussed in the class skipped this important phase for ordinary user since we assume that every user is familiar with the instruction for specific task and calculate operations and time based on this assumption.
Anyway, this "Trial and Error" will be difficult to describe because it depends on users so that statistical approach is needed.
|
Commented by 
