ME/IE 8773-8774
Improving Automated Process Planning: Addressing
the
"Chicken-and-the-Egg Set-Up/Fixture" Dilemma
by
Caroline C. Hayes, Ph.D.
Associate Professor, I.E.
Department of Mechanical Engineering
University of Minnesota
Minneapolis, MN 55455
Wednesday, November 18, 1998
1:25 p.m. UNITE Channel A
Room 108 Mechanical Engineering
Coffee available in 152 ME following the seminar
Coordinator is an automated set-up planner that
addresses the "chicken-and-the-egg" fixture/set-up problem
in 3 and 5 axis CNC machining. This is a circular difficulty in
which one needs fixturing information in order to sequence set-ups.
However, one must already know the set-up sequence, and hence
the shape of the part in each set-up, in order to plan for fixtures.
Where is one to begin? Most automated process planners take a
cyclic approach. First they make an educated guess at a good set-up
sequence and then determine the fixturing implications. If feasible
fixturing cannot be found then the planner will go back and modify
the set-up sequence possibly several times. The result is a feasible
plan, but not necessarily one that makes particularly good global
choices. Coordinator addresses the fixture/set-up problem using
a least-commitment approach in which fixturing and set-up options
are simultaneously explored in increasing levels of detail. It
does so using a representation called "Medusa" of multiple
possible part shapes and faces that might arise during processing
are represented. This allows some of the higher level fixturing
analysis to be moved up-stream, prior to set-up sequencing, so
that it can be used to determine a better sequence on the first
try. This greatly reduces the need for iterative adjustment and
allows better overall setup/fixture decisions to be made, resulting
in higher quality plans.
Dr. Caroline Hayes received her B.S. in 1983
in Math, her M.S. in 1985 in Knowledge-Based Systems and her Ph.D.
in Robotics, all at Carnegie Mellon University in Pittsburgh Pa.
From 1991 to 1998 she was an assistant professor in the department
of Computer Science at the University of Illinois in Urbana, Illinois.
She joined the department of Mechanical and Industrial Engineering
at the University of Minnesota in February of this year. She is
an associate editor for IEEE Transactions of Data and Knowledge
Engineering. Her areas of research include decision support for
areas such as manufacturing planning, design, and military decision
making.
Informal Faculty Luncheon: Wednesday,
November 18, 1998, 11:45 a.m., Room 405, Campus Club. Professor
Hayes will be able to attend.