Model Dependency Maps for transparent concurrent engineering processes
Sprache des Vortragstitels:
Englisch
Original Tagungtitel:
14th Mechatronics Forum International Conference, Mechatronics 2014, June 16-18, 2014, Karlstad University, Sweden
Sprache des Tagungstitel:
Englisch
Original Kurzfassung:
Mechatronic product development involves engineers
from different disciplines and from various workgroups, who
concurrently work on diverse aspects or sub-systems of that
product. They use several models which require input and
generate output information. Some of these outputs are inputs
for other models, in particular also for other workgroups, or are
essential inputs for decision making. For large mechatronic
systems these different models, decision making steps and the in- and output data establish a complex interdependency network.
Quite often this network is not explicitly known, which leads to
inefficient and error prone design processes. In fact, no
established methods and tools exist, to investigate and describe
existing model networks.
In this paper the problem of investigating existing model-, data- and decision-making-networks, their representation and
visualization are addressed. The proposed representation of the
network is called "Model Dependency Map" (MDM). It helps to
increase the mutual understanding of discipline-specific
engineering processes and their underlying interdependent
models. Also a short introduction to the information-acquisition
process is given, followed by a method how information can be
effectively visualised using various granularities of the MDM. A
small set of specified symbols is used to describe all the
accumulated information and knowledge to build up the MDM.
Basically, there are two essential types of symbols: nodes and
edges. Each node represents a model, which contains a
multiplicity of different parameters, but also information about
the model?s purpose, involved stakeholders, used tools,
knowledge and additional information. The dependencies
between the models are represented by a set of defined edges,
which, for instance, describe how information (e.g. work-results)
is transferred between different models.
Forschungs-