Bryan:
re: your comment "Bad gages or an eyeball glitch are bad enough, but a
logger can add
enormously to the garbage information load"--
I completely agree. I have encountered many cases in my "industrial" work
where (unrecognized) bad data led to bad decisions and costly results.
In LSR, I have had unexpected results from three channels that we decided to
ignore (assuming that they were related to transducer or logger problems)--
and later found that that data was probably trying to tell us how to save an
engine. Too bad we had not enough experience with the datalogging package
to have had confidence in it.
The bottom line is-- if you can't be REALLY SURE of your data, it's best not
to even have it. You can manage your program better by using instincts
developed through your racing experience. Bad data will only make you
ignore those instincts at the wrong times.
Having said that, a high priority for me is to get the datalogger going on
my bike so I can test all the sensors thoroughly and gain the needed
confidence in all the data.
Russ, #1226B
-----Original Message-----
From: Bryan Savage [mailto:basavage@earthlink.net]
Sent: Tuesday, February 25, 2003 12:19 PM
To: Russel Mack
Cc: James Tone; wmtsmith@landracing.com; Doug Odom; Land Speed List
Subject: Re: data recorder
Russ,
You have made the most important statement in this interesting thread:
BAD DATA IS MUCH WORSE THAN NO DATA
Bryan
/// unsubscribe/change address requests to majordomo@autox.team.net or try
/// http://www.team.net/mailman/listinfo
/// Archives at http://www.team.net/archive/land-speed
/// what is needed. It isn't that difficult, folks.
|