Comments on Improving Transferability Between Different Engineering Stages in the Development of Automated Material Flow Modules
📝 Abstract
In the paper by D. Regulin et al. (IEEE Trans. On Automation Science and Engineering, vol. 13, no. 4, 1422-1432, October 2016) authors claim that they present a meta-model for the modeling of the Automated Material Flow Module (aMFM) and a model-driven design approach for aMFMs. In this letter we comment on the presented meta-model and the proposed model-driven approach regarding their potential for exploitation. We present specific arguments and make cases that call the authors design decision into question.
💡 Analysis
In the paper by D. Regulin et al. (IEEE Trans. On Automation Science and Engineering, vol. 13, no. 4, 1422-1432, October 2016) authors claim that they present a meta-model for the modeling of the Automated Material Flow Module (aMFM) and a model-driven design approach for aMFMs. In this letter we comment on the presented meta-model and the proposed model-driven approach regarding their potential for exploitation. We present specific arguments and make cases that call the authors design decision into question.
📄 Content
1 Comments on “Improving Transferability Between Different Engineering Stages in the Development of Automated Material Flow Modules”
Kleanthis Thramboulidis Electrical and Computer Engineering, University of Patras, Greece.
Abstract—In the paper by D. Regulin et al. (IEEE Trans. On Automation Science and Engineering, vol. 13, no. 4, 1422-1432, October 2016) authors claim that they present a meta-model for the modeling of the Automated Material Flow Module (aMFM) and a model-driven design approach for aMFMs. In this letter we comment on the presented meta-model and the proposed model- driven approach regarding their potential for exploitation. We present specific arguments and make cases that call the authors design decision into question.
Index Terms—Industrial automation systems, model-driven
development, meta-modeling.
I. INTRODUCTION
For the definition of an efficient model-driven development
process the source and the target models as well as the
automation of the transformation process of the source model
to the target one are required. Meta-models are usually used to
describe the source and the target models. Authors claim in [1]
that they present: a) a meta-model, namely AutoMFM, for the
modeling of an automated Material Flow Module (aMFM)
and b) a process to facilitate the development of Automated
Material Flow Systems (aMFSs) which are considered as
compositions of aMFMs.
In this letter, the proposed in [1] meta-model is discussed in
the context of the model-driven development paradigm.
Specific arguments that call into question the authors’
decisions and claims, regarding the maturity, robustness and
effectiveness of the AutoMFM, are presented. Taking into
account that a rigorous process assumes a robust, mature and
effective meta-model, we present only a few comments on the
process presented in [1]. The remainder of this letter is
organized as follows. Section II establishes the context of the
discussion which is presented in Section III.
II. THE CONTEXT OF THE DISCUSSION
Model-driven engineering (MDE) has been successfully
used in the manufacturing domain to alleviate the complexity
of platforms and express domain concepts effectively [2].
Models are used to represent in a formalized way the structure,
the behavior, and the requirements of the system under
development. Several approaches focused not only on the
software discipline of manufacturing system but also on the
system level where the system is considered as a composition
of mechatronics/cyber-physical components, e.g., [3].
The objective of constructing a model for a system, and thus
for an aMFM, is to address its complexity by describing it in
an abstract way. This model should be independent of the final
execution platform and thus it has to be transformed to an
executable one. The automation of this transformation leads to
a model-driven development process. The initial, platform
independent model is known as the source model and the final
executable one is known as the target model. A model for an
aMFM would be a description of the aMFM at the system
level expressed in a well-defined language. The meta-model is
used to model/describe this well-defined language. Thus,
system models constructed using a specific meta-model
conform to the meta-model and this check is automated in a
model-driven development process. Moreover, meta-models
are used to formalize the domain knowledge and facilitate the
job of the system engineer [2]. Obviously this should be the
intent of the AutoMFM but this is not as it is argued in this
letter.
For the development of a multi discipline system, such as
manufacturing systems, it is common to use several tools and
models for the various disciplines. AutomationML, as is also
the case for ISO10303-233 [4], defines a standardized neutral
data exchange format based on XML for the storage and
exchange of plant engineering information among the various
discipline tools involved in the engineering of the plant.
III. DISCUSSION
Authors claim in [1] that the proposed meta-model and the
approach have specifically developed for the domain of
aMFSs. However, it is not clear what makes the proposed
meta-model specific for aMFSs since the only domain specific
information captured in the meta-model is the types of logistic
tasks [1, Ref. 31]. This is also evident from [1, Sec. II] and [1,
Ref. 31] where none of the defined requirements is specific for
aMFSs. Arguments should be given to claim that material flow
modules cannot be effectively modeled by an industrial
automation domain modeling language. Moreover, the benefits
of such a very specific modeling language have to overwhelm
the cost of developing and maintaining another language. The
case of extending an industrial automation domain language to
address the specific requirements of material flow systems (if
these exist) is the obvious way to go.
2 Autho
This content is AI-processed based on ArXiv data.