CAD-Portal.com:
Product Reviews:Unigraphics V17
Unigraphics
V17's Knowledge Driven Automation, the next killer app?
A few weeks ago we had the opportunity
to spend an afternoon with Ron Close, Marketing Director for the Unigraphics
line of business and Ken Hosch, Product Manager for Knowledge Driven Automation.
In the few hours we have together we plan to concentrate on the incorporation
of Knowledge Driven Automation (KDA) into the latest release 17 of Unigraphics.
The core of their implementation,
called UG Knowledge Fusion, uses the Heide Intent programming language
as an underlying architecture. The kernel executes the interpretive language
defined using Knowledge Fusion. The idea is that the engineers who are
comfortable with programming languages would develop a knowledge feature
which consists of a programmatic instructions, using the Heide language.
Unigraphics has extended the original Heide software with an entire group
of object classes for CAD/CAM objects.
We had the chance to look at this
implementation by exploring a few pre-developed knowledge features developed
using UG Knowledge Fusion.
We looked at developing an I-beam.
Again, to assist you with understanding how the system works and some of
the tests we reviewed, we have saved images for you to review. You
can review the images and the questions by clicking here. View
a sample KF program by clicking here. The key point is that the
Knowledge Fusion Language and the resultant geometric part are intimately
related. That is, both the Knowledge Fusion feature and the part are fully
two way associative.
One intriguing example was the development
of this I beam using an Knowledge Fusion I beam feature. We were able to,
when we had the I beam created, drill a hole in it and add fillets. In
most other competitive systems, rerunning the knowledge application would
create a brand new part. However, in UG V17 changing the KF parameter and
rerunning the knowledge feature also picks up the hole and the fillets,
so that downstream features are added to the eventual part. Thus, we now
begin to understand the real power if a tightly imbedded KF within a modeling
system -- the ability to marry both the CAD part and the knowledge fusion
feature. We have not seen and do not know of any other CAD system able
to do this.
We also looked at a more complicated
design, of which we have the images on the above link. In this case we
want to design a series of conduits under an aircraft structural floor.
Changing the location of the supporting beam for the conduits, changed
the conduits that ran through changed the geometry of the beam dramatically.
In
my estimation, these changes could not have been made (and certainly the
UG folks agree) using parts and assemblies built using only parametric
features.
To ease the user interface, UG V17
allows the creation of dialog boxes that present to the user, the data
input allowed by the knowledge feature. The dialog boxes are customized
to display only those specific parameters that the Knowledge Fusion feature
allows the user to change. After reviewing the programming skills required
to develop these KF's (knowledge features), we don't expect the average
user to write the necessary code. However, once the program is written,
users are only exposed the parameter changes, and like other features,
should easily be able to execute the KF's. Ron Close pointed out that V17
was UGS's first entry. They will be working hard to improve the ability
to more easily develop KF programs.
In summary, we believe that UGS'
KF offering represents a new wave of technology for the CAD design marketplace.
Customers with repetitive problems that are beyond the scope of variable
driven design systems should definitely plan on investing the resources
and skills to learn more about knowledge based engineering systems. Before
this product, these systems were admittedly useful, but UGS' implementation
takes them out of the realm of "interesting software" and tightly integrates
them into the mainstream of designers. In the long term, while we anticipate
that KF programs may become easier to write, that some degree of programming
skills will be necessary. UGS's main competitors are not standing still.
Last year we explored some of the differences of the knowledge thrust also
by Pro/ENGINEER's behavioral modeling and CATIA's KnowledgeWare. We expect
that UG V17's KF will provide added impetus to these vendors, as well as
mid range solid modeling vendors to offer comparable capabilities. In the
right design situations, knowledge based engineering presents the opportunity
to save substantial time and effort. Someday we might even see our dream
come true. Just input the overall specifications for the product -- and
out pops the design!
In the meantime, our advice is to
seriously begin thinking about how this capability might make a difference
in your company.
www.ugsolutions.com
|
