A Research Roadmap for Model-Driven Design of
Embedded Systems for Automation Components

T. Strasser1, M. Rooker1, I. Hegny2, M. Wenger2, A. Zoitl2, L. Ferrarini3, A. Dede3 and M. Colla4
1Robotics and Adaptive Systems, PROFACTOR GmbH, Austria, {thomas.strasser, martijn.rooker@profactor.at}
2Automation and Control Institute, Vienna University of Technology, Austria, {hegny, wenger, zoitl@acin.tuwien.ac.at}
3Dipartimento di Elettronica e Informazione of Politecnico di Milano, Italy, {ferrarini, dede@elet.polimi.it}
4Institute of Computer Integrated Manufacturing for Sustainable Innovation, University of Applied Sciences of
Southern Switzerland, Switzerland, {marco.colla@icimsi.ch}

Abstract-The aim of this paper is to describe a research roadmap
for a multi-domain model-driven embedded systems design ap-
proach and the corresponding meta-model which is applicable to
the domain of complex Industrial Automation and Control Sys-
tems (IACS). The special requirements of the industrial automa-
tion sector are taken into account by this novel approach, utilizing
existing model-driven techniques. This approach is currently be-
ing developed in the Framework Seven (FP7) Embedded Systems
Design project MEDEIA funded by the European Commission.
I. INTRODUCTION
Today’s production processes will be performed more and
more by automated machines as the level of automation in-
creases steadily. The main trend in industrial automation is the
increasing need for customized and individualized plants. The
production lines have to be constructed and adapted to new pro-
duction processes quickly [1]. Such highly automated systems
are mainly controlled by embedded hard- and software which
are heterogeneous in nature. Heterogeneous specification and
heterogeneous architectures lead to a tremendous increase in the
design complexity. The software engineering costs will increase
up to 80% of the overall systems costs in the next 15 years (cur-
rently this ratio is about 55%) [2],[3] as depicted in Fig. 1.
POWER
Allen- Bradley
QU ALITY
RUN
BAT
I/O
Rs2 32
OK
RU N PRO GREM
L ogix555 0 DC INPUT
O
K
ST
ST
ST
ST
0 1 2 3 4 5 6 7
1 1 1 1 2 2 2 2
6 7 8 9 0 1 2 3
2 2 2 2 2 2 3 3
4 5 6 7 8 9 0 1
1 1 1 1 1 1
0 1 2 3 4 58 9
DC INPUT
O
K
ST
ST
ST
ST
0 1 2 3 4 5 6 7
1 1 1 1 2 2 2 2
6 7 8 9 0 1 2 3
2 2 2 2 2 2 3 3
4 5 6 7 8 9 0 1
1 1 1 1 1 1
0 1 2 3 4 58 9
DC OUTPUT
O
K
ST
ST
0 1 2 3 4 5 6 7
8 9 10 11 12 13 14 15
QUAL ITY Alle n-Bradle y Co mpact Block I/O179 1D-16 8DI
16 INPUT S + DC POWER
QUALITY Allen-Bra dle y Compact Block I/ O17 91 D-8B8P
16 INPUTS / 8 OUTPUTS + DC POW ER
Mod/ Net
Status
Logi c
Status
DeviceNet
QUALITY Alle n-Br adley MicroLo gix
1 500
PO WER
RU N
FAULT
FOR CE
BAT.LO
CO NN 0
DCO MM
LSP
DC INPUT S
24 V SIN K / SOUR CE
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
DC / R ELAY O UT
24 V SO URCE
0
1
2
3
4
5
6
7
8
9
10
11
28BXB
DC PO WER
24V
0
1
2
3
4
5
6
7
8
9
10
11
A
C
INP
U
T
23 0 VAC
0
1
2
3
4
5
6
7
8
9
10
11
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
D
C
INP
U
T
S
2 4 VDC SINK / SOURCE
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
D
C
OU
TP
U
T
2 4 VDC SOURCE
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
0
1
2
3
4
5
6
7
A
C/
DC
O
U
T
RELAY
0
1
2
3
4
5
6
7
Q UALITY Allen-Bradley Pane lView 300
2
1
3 4
5
6
100
50
0
F1 F2 F3 F4
F5 F6 F7 F8
QUALITY Allen-Bradley PanelView 300 Micro
2
1
3 4
5
6
100
50
0
F1 F4F3F2
NO
TI
C
E
NO
TI
C
E
NO
TI
C
E
NO
TI
C
E
Al le n-Bra dl ey
PLC - 5/80 E
1
2
1
2
D
C
O
U
T
P
U
T
1
0
3
0
v
TO
0001
03
04
07
02
05
06
10
1415
11
12
13
16
17
10
1415
11
12
13
16
17
0001
03
04
07
02
05
06

Fig. 1: Investment Costs for Control Systems Production Systems [3]
Therefore, the automation and control engineering of a pro-
duction plant takes a huge amount of time. Especially in the in-
dustrial sector, each plant or system can be seen as a prototype.
The need is to evolve the automation solutions in response to:
rapidly changing demands due to new production processes; the
evolution of hard- and software into more and more complex
forms. The design methods and approaches used depend strongly
on the application field. This means that different automation
and control systems have to work together to perform automa-
tion and control tasks [1]. Design and engineering of such hete-
rogeneous systems is tricky and needs the knowledge from dif-
ferent domains during all phases of development. During the
design phase, the system complexity, the domain/platform-
dependence and the design time are the critical factors. In the
design phase, the following items are essential [4]:
- System specification should be done at an easily unders-
tandable level of abstraction. The designer at each level of
the plant architecture should be able to describe what he
wants, not how to achieve it.
- Combination of different domain-specific techniques
should be manageable to overcome the limited applicability
of each, in order to allow modeling, analysis, and imple-
mentation of complex, heterogeneous embedded systems.
- The specification of systems architecture and system com-
ponents should not depend on languages used for embed-
ded system design.
One of the main current design trends in automation and con-
trol systems is to put components, blocks made of hardware,
software and intellectual property together (like algorithms and
data structures) [5]. This in turn calls for common, language
independent models for representing, saving and reusing such
components. Neither the state-of-the art or current trends in de-
sign and engineering in industrial automation [1],[5] are capable
of providing an applicable solution to the above mentioned is-
sues. Solutions from other sectors like automotive or aerospace
are developed for their special needs. These requirements are
mostly different from the needs in industrial automation. Moreo-
ver, most of these solutions require a deep knowledge in soft-
ware engineering, while in the industrial automation sector main-
ly electrical and control experts design the control systems.
The paper is organized as follows: section II provides an over-
view of the MEDEIA approach whereas section III gives argu-
ments for the rational of it. The MEDEIA Automation Compo-
nent meta-model is described in section IV. Section V provides
an outlook and talks about the future work in MEDEIA.
II. MODEL-DRIVEN DESIGN FOR
INDUSTRIAL AUTOMATION AND CONTROL SYSTEMS
One main obstacle for efficient engineering of Industrial Au-
tomation and Control Systems (IACS) is that the design metho-
dologies as well as the implementation technologies within
564978-1-4244-3760-3/09/$25.00 c© 2009 IEEE