As a 10-year machinist, and 8 years on the engineering side of the wall, let
me throw this out...

There is never a "finished" program.

Any experience on a screw machine will tell you that you can always find a
way to save five seconds here or there.  Proper setups can eliminate a lot
of .100" 'clearance' moves.
Rapid moves can always be 'a little closer' to the part before feeding, the
dividing line between tool life / surface finish / cycle time is always hard
to define with any kind of finality, lot size can always determine whether
or not tool life is critical, and rough cuts can always be .005" deeper to
save one extra pass, etc...  
As an operator, there is nothing better than loading the program, pushing
the green button, and going back to sleep.  As an engineer, there's nothing
better than an operator who pushes the green button and CAN go back to
sleep.
But to stamp a hardcopy "done", and prevent any further improvements is
unrealistic.  
I'm fortunate enough to work with operators who can find an extra 10 seconds
here and there to edit out of the program, and who can troubleshoot surface
finish/tool life/chatter on their own.
As an operator, I was expected to 'tweak' code on the night shift and on
weekends; the position of "engineer" was god-like and they would never
consider working any overtime. On the floor editing was almost a
requirement.
And, I have worked for a strong UAW shop, where the operators are nothing
more than button pushers, and the program COULD NOT be modified on the
floor. 

It comes down to this... if you want to hire a bunch of engineers, or work
the ones you have to death, then the code can be perfect first time every
time, and editing in the control can go away.  This allows you to hire any
Joe Blow off the street and pay him minimum wage to push buttons and say
'duh' when things go wrong.
Or, you can hire some very qualified operators, and let them fine tune the
programs on the floor, and let the engineer be more than just a programmer.
A competent operator is more valuable than the greatest cad-cam system on
earth.
The danger comes from NOT managing the edited files... all edits should be
approved (what are supervisors for?!) and all edited files should be
compared to the originals before blindly modifying the masters.  Everyone at
the machine, and the engineer as well, should all agree on what is and is
not a required change.  And if the system works well, then the ISO 'stuff'
can be modified to either allow editing, or not even mention it at all.  As
long as you have, and follow, a system, then all is right with the world.

I'm not saying that we should be writing junk code and making somebody else
fix it, but at least the guy running the machine 40 or 60 or 80 hours a week
have some say as to what makes things run best.  Cycle time and part quality
will immediately tell you if he's doing good things or not.

Greg Hyman
Mauch Labs, Inc.

======================================================================
To find out more about this mailing list including how to unsubscribe,
send the message "info mfg-smartcam" to [EMAIL PROTECTED]
======================================================================

Reply via email to