[conspire] Risks of automation

[Rick Moen]

A couple of details I forgot to include:

Q:  Why didn't you notice the automated watering not working at all?

A:  Because it's set to run once per day early in the morning before dawn,
    and once after dark.  (Watering when the sun's out is usually not 
    a good idea.)  I haven't been up at 6:30am to observe the front and
    then side/back yard watering runs getting water, and just didn't 
    happen to be out there at 7:30 pm to see that sequence, either.
    Nor was I up at 6am, three days a week, to see whether the lawn
    got watered.


Q:  How long was all watering shut down?

A:  About two weeks.  Ugh.


Q:  Were you serious when you said expertise is never required for 
    diagnosis, just sharp observation, keeping of records, and 
    consistent use of logic?

A:  Yes.  Even someone knowing nothing about watering systems could
    have solved this problem.  It's a matter of knowing what the 
    system's mission is (automated watering), and being willing to 
    tinker to test hypotheses and eliminate suspects.  And, my 
    point is, the same applies for any other sort of diagnosis.
    All that expertise, and good intuition, buys you is the possibility 
    of getting to the solution a little faster.