Process Modeling: A New Paradigm
Andrew Kusiak
ankusiak@icaen.uiowa.edu
Intelligent Systems Laboratory
Department Industrial Engineering
4130 Engineering Building
The University of Iowa
Iowa City, Iowa 52242 - 1527
Phone: 319-335-5934
Fax: 319-335-5424
Url: http://www.icaen.uiowa.edu/~ankusiak
Numerous methodologies have been developed for modeling processes
(including a design process) ranging from CIM-OSA to Integrated
Definition (IDEF) methodologies. The process modeling effort corresponds
to the product modeling effort. One would think that in the near
future process definition languages (e.g., IDEF like) will be
able to communicate with hardware description languages (e.g.,
VERILOG like). There is no reason for not merging the two areas
in the same way as engineering design and manufacturing have moved
closer to each other under the banner of integrated product and
process development. To date, reengineering might have been the
strongest driver behind process modeling tools. The full realization
of distributed design and manufacturing (virtual design and manufacturing)
will create a need for new product and process modeling methodologies
and tools. Neither the existing standards such as STEP or any
of the existing process modeling methodologies will be able to
fully meet the challenge of virtual enterprises because of three
main reasons:
- existing modeling methodologies are not comprehensive enough
- time is not well handled by the existing methodologies
- analysis and negotiation capabilities are limited
A technical mind tends to assume that when an information is available it is going to be used, or a good invention will always lead to a successful product. Unfortunately, there are many factors other than technical that make people use the right information or design a profitable product. Some call these factors ìsoft factorsî that may range from a personís cultural background to human trust. Any model that interacts with a human should include an appropriate human model. The incorporation of human modeling into existing process models must be addressed by future modeling tools.
The existing product and process modeling tools tend to be static. However, many aspects being modeled are dynamic. Temporal relationships have to be well reflected in the next generation modeling methodologies.
As product and process models have to analyzed from different perspectives, e.g., reliability, quality, time, new tools are needed to perform computational analysis of information, thus allowing one to improve the models and more often validate the product and process models. Negotiation schemes becomes necessary to handle optimization objectives that are often conflicting.
Decision making capabilities are needed to generate design process models and to realize designs according to the model selected. Process analysis and negotiation methodologies are examples of decision making tools that apply to the two areas.
Acknowledgment
Some of the ideas presented were discussed in five workshops organized by NIST (http://www.mel.nist.gov/workshop/iceimt97/) and IEEE System Level Design Language Workshop (http://www.cfi.org/sld/).
 |