TUGraz OPEN Library

Asadzadeh, Mohammad Zhian

Fabrizio, Michele

Arrigoni, Enrico

Nonequilibrium variational cluster perturbation theory

Wolfsgruber, Christoph

Lichtenegger, Gerald

Originalarbeit
BHM (2016) Vol. 161 (5): 221–224
DOI 10.1007/s00501-016-0474-6
© The Author(s) 2016. This article is available
at SpringerLink with Open Access
Optimal Configuration in Production Planning and Control
Findings of a Simulation-based Evaluation Study
Christoph Wolfsgruber and Gerald Lichtenegger
Institut für Maschinenbau- und Betriebsinformatik, Technische Universität Graz, Graz, Austria
Received March 1, 2016; accepted March 30, 2016; published online June 2, 2016
Abstract: The operational practice is undergoing funda-
mental changes due to the increasing complexity and new
technological possibilities often summarized with the buzz-
word ‘Industrie 4.0’. Besides others, these trends will have
major impacts on the way production is planned. This
paper shows findings of a simulation-based evaluation of
production planning and control methods regarding these
challenges. Based on these fundamental insights, a frame-
work for an optimal configuration in production planning
and control is presented.
Keywords: Production planning, Production control, Push
systems, Pull systems, Discrete event simulation, Industry
4.0
Optimale Konfiguration in der Produktionsplanung und
-steuerung—Erkenntnisse aus einer simulationsbasierten
Evaluierungsstudie
Zusammenfassung: Die betriebliche Praxis erfährt zurzeit
grundlegende Veränderungen aufgrund der zunehmenden
Komplexität und neuen technischen Möglichkeiten – oft un-
ter dem Schlagwort „Industrie 4.0“ zusammengefasst. Die-
se Trends werden unter Anderem erhebliche Auswirkun-
gen auf die Art undWeise, wie die Produktion geplant wird,
haben. Dieser Artikel zeigt Erkenntnisse aus einer simula-
tionsbasierten Evaluierungsstudie von Methoden der Pro-
duktionsplanung und -steuerung in Bezug auf diese Her-
ausforderungen. Basierend auf den gefundenen Erkennt-
nissen,wird ein Framework für eine optimaleKonfiguration
in der Produktionsplanung und -steuerung vorgestellt.
Ass.-Prof. Dipl.-Ing. Dr.techn. G. Lichtenegger ()
Institut für Maschinenbau- und Betriebsinformatik,
Technische Universität Graz,
Kopernikusgasse 24/III,
8010 Graz, Austria
E-Mail: gerald.lichtenegger@tugraz.at
Schlüsselwörter: Produktionsplanung,
Produktionssteuerung, Push-Systeme, Pull-Systeme,
Ereignisorientierte Simulation, Industrie 4.0
1. Introduction
Today industry is dealing with challenges that have con-
stantly arisen and intensified over the last twenty years –
amassive increase in product variety, a shortening of prod-
uct life cycles, highly volatile demands [1]. Furthermore,
Industry 4.0 with its enabling technologies often referred
to as the Internet of Things (IoT) and Cyber-Physical Sys-
tems (CPS) is going to change the domain of production
in a sustained manner. Needless to say that in this sit-
uation strongly linked functions like production planning
and production control (PPC) have to be adapted and en-
hanced for their part. But the above mentioned newly of-
feredpossibilities especially in information technology also
offer new promising opportunities – dramatically increased
data quality, constant status updates from the shop floor,
and the tracking of activities and objects in real time [2].
With that said, the need for a review and an evaluation of
existing planning and control strategies as well as of con-
figurations in production is evident. Therefore a detailed
understanding of the production relationships and the ex-
isting concepts and methods of production planning and
control is needed.
In this paper a simulation model-based approach is pre-
sented that enables a performance evaluation of a variety
of PPC concepts appliedunder different production settings
(see Fig. 1). One of the advantages of using a simulation-
based approach is the possibility to even address produc-
tionsettingsandstrategiesperhapsnot usedat themoment
for different reasons and investigate their potentials for fu-
ture application. The great challenge in production plan-
ning and control which can be seen as the process of deter-
mining the best economic use of the available production
factors to fulfill the customer orders is to set improvements
BHM, 161. Jg. (2016), Heft 5 © The Author(s) Wolfsgruber & Lichtenegger 221
Originalarbeit
Fig. 1: Motivations for theevaluation
in many different target dimensions at the same time. This
means one has to deal with contradictory objectives all the
time. As an example Gutenberg describes the conflict be-
tween the utilization of production resources, the inven-
tory and delivery reliability as the dilemma of scheduling
[3]. This phenomenon lead to the development of hun-
dreds of different planning strategies and methods such as
lean concepts (e.g. JIT), material requirements planning
(MRP), or even complex linear programming optimization
approaches.
2. Simulation-based Evaluation Study
The goal of the presented evaluation study is to analyze
various PPC methods with respect to their flexibility (abil-
ity) to respond to external changes and the impacts of the
informatization in manufacturing. Based on different hier-
archical types of flexibility, the product flexibility and the
volume flexibility of Sethi & Sethi [4] and respectively the
mix flexibilityand volume flexibilityof Koste&Malhotra [5]
were identifiedas themajor sourcesofflexibility that areas-
sociated with production planning activities. The impacts
of informatization are modeled with different availabilities
and qualities (deviations) of the transactional data needed
for planning. Furthermore, the investigation is taking vari-
ability and lead time effects into consideration. Tab. 1 gives
an overview of the different evaluation dimensions ana-
lyzed in the study. The data sets used in the evaluation
study were captured during several industrial projects and
represent different practical scenarios.
The general structure of the evaluation model consists
of three main parts (see Fig. 2):
Manufacturing model
Customer model
Methods of production planning & control
TABLE 1:
Evaluation dimensions of the simulation study
Characteristic
Type Internal (Supply) External (De-
mand)
Flexibility
& Variability
Setup time
Supply variability
(MTTR)
Part variety
Demand mix
Data Quality
& Availability
Inventory devia-
tion
Planning cycle
Forecast changes
Forecast update
frequency
Others Lead time
Fig. 2: General structureof theevaluationmodel
Fig. 3: Selected resultsof thesimulation study (basedon [6])
In the show case, the manufacturing model represents
a flow-shop production with disconnected flow lines and
the customer model is based on the logistic principles of
automotive supply industry. These two model compo-
nents are implemented in a discrete event simulation-type
model (DES). In addition, the evaluation model contains
various PPC methods which are either linked via program-
ming interfaces to the DES or directly implemented in the
simulation program.
With this model various standard parameter scenarios
are analyzedover a simulation timeof several daysdepend-
ing on the problem. The scenarios include evaluations of
the impact of forecast changes, data availability, data qual-
ity, demand variability, and supply variability1. Represent-
ing the different analyzed scenarios, Fig. 3 shows selected
results of the simulation study in required setups over work
in progress (WIP) plots.
This chart shows the high influence of the product vari-
ety on the required number of setups. While, with a low ex-
ternal variability, the difference between the pull-type PPC
system (e.g. Kanban) and the push-type system (e.g. MRP)
is only minor (solid lines) at a high external variability, the
push type system can demonstrate its superiority (dotted
lines).
1 A more detailed explanation of the simulation model, the scenarios,
andtheresultscanbefoundin[6,8].
222 Wolfsgruber & Lichtenegger © The Author(s) BHM, 161. Jg. (2016), Heft 5
Originalarbeit
3. Findings
The research using the simulation-based evaluation study
showed to be a valuable approach to better understand the
influencing factors in PPC. A brief description of the find-
ings with respect to the following domains will be given
[6]:
PPC model accuracy
Data availability
Data quality
Product flexibility
Models are always only a limited representation of reality.
In the process of model building, delimitation, reduction,
decomposition, aggregation, and abstraction are used to
capture the major elements and influencing factors of the
complex reality. Several different PPC models with differ-
ent accuracies regarding their representation of time, the
used mechanisms for lot sizing, the capacity consideration
etc. exist. The simulations showed that data availability
and PPC modeling accuracy are related to each other. Fur-
thermore, the ongoing informatization will dramatically in-
crease data availability on the shop floor and in the sup-
ply chains. In PPC, mainly the inventory and demand data
availability will be influenced. The prior insight regarding
the PPCmodel accuracy already pointed out that this avail-
ability increase must go hand in hand with more detailed
planning models. Otherwise no significant efficiency ben-
efit can be achieved from this trend. The evaluation study,
moreover, showed that real time availability of data is only
partly necessary, especially when only a part of the data
(e.g. inventory data) is available in real time, while other
data which is relevant for planning (e.g. demand data) is
only updated every day. In such a case, no real perfor-
mance gain was measured in the simulation results. Apart
from the availability of data, the quality of the data, mea-
sured in the evaluation by the inventory inaccuracy, will
also increase through the informatization. Deviations be-
tween the data in information systems and the real invento-
ries on the shop floor have to be buffered using additional
safety inventory. The simulations using different inventory
inaccuracies showed that in-detail PPC models are more
sensitive to deviations in the stock levels than simple plan-
Fig. 4: SuggestedPPCconfigurationbestpracticeapproach (basedon [6]
ning models. This effect is mainly attributable to the al-
ready higher inventory levels of the simple PPC models in
comparison with themore sophisticated models which can
run with the same settings at a lower stock level. Conse-
quentially, in-detail PPCmodelsneedmoresafety inventory
to protect against the inventory inaccuracies than simple
models. Apart from the aspects of informatization, also the
product flexibility demanded by the customer is a major
concern in PPC. The evaluations showed that the range of
products has a high influence on the performance of the
planning system. Especially the Kanban system showed
only limited applicability when high product flexibility is
needed. Only with a range of two different products, the
Kanban system delivered a performance which was close
to the performance of more advanced MRP systems.
4. Configuration Framework for Production
Planning and Control
The selection of an optimal PPC configuration is a diffi-
cult task which depends on several factors and needs to be
taken more and more frequently, especially in the rapidly
changing economic environment. An interesting approach
for theselectionof the right PPCstrategycanbe found in the
information/control/buffer (I/C/B) portfolio by Schwarz [7].
This framework analyzes different PPC strategies in the di-
mensions information system, control system, and buffer
system. According to Schwarz [7], an operation must se-
lect the PPC strategy based on the total costs of these three
components. His framework provides an elegant solution
for the PPC configuration problem by combining the dif-
ferent dimensions on the common measure costs. How-
ever, the evaluation study showed that the I/C/B portfolio
needs to be updated in some of the dimensions to fulfill
present needs. The main drawback of the I/C/B framework
is the missing consideration of the requisite flexibility of
the manufacturing system based on the customer require-
ments. The simulations showed that the product flexibility
has a major impact on the performance of a PPC system.
Furthermore, the evaluation study showed a relationship
between data availability and PPC model accuracy. Based
on that, the I/C/B decision framework was adapted and ex-
tended (see Fig. 4).
BHM, 161. Jg. (2016), Heft 5 © The Author(s) Wolfsgruber & Lichtenegger 223
Originalarbeit
The new framework presented in Fig. 4 is grounded on
three basic systems: the production system, the informa-
tion system, and the PPC system. Furthermore, indirect
factors that cover e.g. knowledge impacts are introduced
additionally. The system performance results out of the
design and the interactions between these factors.
The production system is mainly characterized by its
supply variability and the requisite flexibility. The supply
variability includes effects originating from set-up activi-
ties, breakdowns, and quality defects. The requisite flexi-
bility originates from the customer requirements in terms
of product variations and delivery time expectations. The
evaluation study shows that the aforementioned factors
have a major impact on the overall performance of all PPC
approaches. Despite the best practices in PPC configura-
tion, standard measures using lean tools and complexity
management approaches have to be taken to gain control
over these highly influencing factors. The joint configura-
tion of the information system and the PPC system is the
major challenge in the selection process. Due to the inter-
action between the information systemand thePPC system
that was shown in the evaluation study, an approach differ-
ent to the I/C/B portfolio was followed (joint consideration
of the two systems). The quality and availability of exter-
nal (customer) and internal (supply) data have remarkable
influences on the selection of a fitting PPC model. Next to
the direct factors of the production, information and PPC
system, indirect factors also have an influence on the sys-
tems performance. These indirect factors can be mainly
attributed to the required skill and knowledge level for the
different PPC methods. Especially pull-type PPC systems
require a certain discipline and responsibility on the shop
floor to operate properly. Overall it can be said that these
factors result in a performance of the system with a cer-
tain use of the different buffering systems (inventory, ca-
pacity, and time) to match the demand. As the evaluation
study showed, simulation is a key technique to reveal the
performance. The decision on the optimal PPC configura-
tion should be based on the overall system costs for the
information system, the PPC system, the resulting system
performance and the costs for the indirect factors [6].
As a consequence of the previously described facts and
circumstances, simple planning methods could surpris-
ingly return a solution at the overall cost optimum when
inventory keeping is cheap. That does not necessarily
mean that an investment in PPC systems using in-detail
models and high quality data are not worth the effort –
amongst others it strongly depends on the cost factors. To
conclude, this generally means that an enterprise has to
first calculate the potential benefits and the return of an
investment in the informatization of the production and
then set up an aligned and cost optimal configuration for
its operations.
OpenAccessThisarticle isdistributedunder the termsoftheCreativeCommonsAt-
tribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/),
whichpermitsunrestricteduse, distribution, andreproduction inanymedium,pro-
vided yougive appropriate credit to the original author(s) and the source, provide
a link to the Creative Commons license, and indicate if changesweremade.
References
1. Nyhuis, P.; Reinhart, G.; Abele, E.: Wandlungsfähige Produktions-
systeme, Hannover: TEWISS, 2008
2. Bauernhansl, T.: Die Vierte Industrielle Revolution: Der Weg in ein
wertschaffendes Produktionsparadigma, in: Bauernhansl, T.; Hom-
pel, M.; Vogel-Heuser, B. (Eds.): Industrie 4.0 in Produktion, Au-
tomatisierung und Logistik, Wiesbaden: Springer, 2014, pp 5–35
3. Gutenberg, E.: Grundlagen der Betriebswirtschaftslehre. Band 1:
Die Produktion, Berlin: Springer, 1951
4. Sethi, A. K.; Sethi, S. P.: Flexibilty in Manufacturing: A Survey, Inter-
national Journal of Flexible Manufacturing Systems, 2 (1990), No 4,
pp 289–328
5. Koste, L. L.; Malhotra, M. K.: A theoretical Framework for analyzing
the Dimensions of Manufacturing Flexibility, Journal of Operations
Management, 18 (1999), no. 1, pp 75–93
6. Wolfsgruber, C.: Informatization in Production Planning and Con-
trol: A Simulation based Evaluation of the Impacts in Flow-Shop
Production Systems, Doctoral Thesis, Graz, University of Technol-
ogy, 2016
7. Schwarz, L. B.: A New Teaching Paradim: The Information/Control/
Buffer Portfolio, Production and Operations Management, (1998),
no. 2, pp 125–131
8. Wolfsgruber, C.; Vössner, S.: Informatisierung in der Produktions-
planung und -steuerung: Eine simulationsbasierte Analyse der
Auswirkungen am Beispiel der Fließfertigung, Leoben, 2015
224 Wolfsgruber & Lichtenegger © The Author(s) BHM, 161. Jg. (2016), Heft 5


Optimal Configuration in Production Planning and Control

Boano, Carlo Alberto

Römer, Kay

Bloem, Roderick

Witrisal, Klaus

Baunach, Marcel

Horn, Martin

Dependability for the Internet of Things—from dependable networking in harsh environments to a holistic view on dependability

Seifriedsberger, Johannes

Rumplmayr, Peter

Evaluation of the combustion noise of passenger car diesel engines

Schwarz, Andreas

Ofner, Patrick

Pereira, Joana

Müller-Putz, Gernot R.

Time domain classification of grasp and hold tasks 
Andreas Schwarz1, Patrick Ofner1, Joana Pereira1 and Gernot R. Müller-Putz1,* 
1Institute of Neural Engineering, Graz University of Technology 
*Stremayrgasse 16 / IV, 8010 Graz, E-mail: gernot.mueller@tugraz.at
Introduction: Brain-Computer Interfaces (BCIs) enable its users to interact with their environment only by 
thought. Earlier studies indicated [1, 2] that BCI might be a suitable method for controlling a neuroprostheses, 
which could assist people with spinal cord injuries (SCI) in their daily life. One drawback for the end user is that 
only simple motor imaginations (MI) are available for control e.g. MI of both feet to control ones arm is abstract 
and in contradiction to an associated natural movement. Therefore we are looking for means to design a more 
natural control modality. One promising scenario would be to use MI of different grasps to actually control 
different grasps of the neuroprosthesis. In this study we attempt to classify the execution of different grasp types 
in low-frequency time-domain EEG signals. 
Methods: Fifteen healthy participants from age 23 to 37 participated in the experiment. In a cue guided paradigm
(see figure 1), subjects were instructed to perform 3 different reach-grasp-hold tasks on 3 different objects: 
palmar grasp (cylinder), pincer grasp (needle) and key grasp (key). To introduce a control condition, one spot 
was deliberately left empty and users were asked to not perform any movement. We recorded 72 trials per 
condition (288 in total) over 8 runs and varied the position of the objects so that every object was positioned 
equally often on each position. We recorded 61 active electrodes (g.tec, g.GAMMAsys) as well as data from a 
data glove (5DT) and a switch button to obtain the movement onset. We rejected artifact contaminated trials and 
channels using a statistical outlier rejection. We down-sampled the EEG to 16 Hz and applied a bandpass-filter 
between 0.3 and 3 Hz (4th order, Butterworth, zero-phase) to extract the low-frequency signal. Using 5 fold 
crossvalidation to avoid overfitting and a random forests classifier [3], we investigated all grasp versus grasp 
combinations. To score significantly higher than chance level (p = 0.05, Bonferroni corrected for multiple 
comparisons over trial time), the accuracy level had to be higher than 64.7 %. Table 1 displays the results.
Figure 1: Paradigm: Participants were instructed to rest the hand comfortably on a pressure button. At second 0, a cross appeared on the 
screen to focus users’ attention. At second 2, one of the objects was highlighted in white for a random time period. As soon as the 
highlighting turned green, participants performed the reach and grasp tasks and held the object as long as the green highlighting remained. 
Thereafter participants returned their hand to the pressure button.
Grasp vs Grasp S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 S12 S13 S14 S15
Pal vs Pin 61.5 78.8 61.7 73.0 76.8 69.7 65.9 69.9 69.6 60.6 71.2 68.0 74.3 63.7 69.2
Pal vs Key 70.0 74.4 64.1 76.9 63.6 68.5 64.6 63.6 66.2 65.7 70.7 71.1 66.2 66.7 67.5
Pin vs Key 64.4 68.9 63.0 67.1 66.7 61.3 66.1 63.0 66.9 65.7 69.7 66.4 62.2 70.7 69.3
Table 1: Peak accuracies after movement onset over all trials in percent. (Pal = Palmar, Pin = Pincer, Key = Key Grasp). Bold values 
indicate performance levels significantly higher than chance. 
Discussion: We could confirm that grasp versus grasp classification in the low-frequency time-domain is 
possible. Fourteen out of 15 participants scored significantly better than chance in at least one combination, 
whereas 8 participants’ performance topped 70%. Peak performances occurred within the first one and a half 
seconds after movement onset, but different for each subject. We believe this is due to the varying movement 
speed towards the object. No significant predictions could be made before actual movement onset. So far these 
results only reflect motor execution of a grasping task – there is still need to investigate whether these results can 
be achieved with motor imagery. Furthermore it is still unknown whether user can be trained to boost 
classification to a robust level. 
Significance: We could show that executed grasp versus grasp classification is possible. We believe that these 
findings will contribute to a more intuitive and natural form of control for neuroprostheses. 
Acknowledgements: This work is supported by the EU ICT Programme Project H2020-643955 MoreGrasp and 
the ERC Consolidator Grant “Feel your reach”. 
References
[1] Pfurtscheller G, Müller GR, Pfurtscheller J, Gerner HJ, Rupp R. “Thought”-control of functional electrical stimulation to restore 
handgrasp in a patient with tetraplegia. Neuroscience Letters, Vol.351 33-36, 2003. 
[2] Rupp R, Rohm M, Schneiders M, Kreilinger A, Müller-Putz G.R. Functional Rehabilitation of the Paralyzed Upper Extremity after 
Spinal Cord Injury by Noninvasive Hybrid Neuroprostheses. Proceedings of the IEEE Vol.103(6) 954-968, Jun 2015. DOI: 
10.1109/JPROC.2015.2395253
[3] Steyrl D, Scherer R, Faller J, Müller-Putz GR. Random forests in non-invasive sensorimotor rhythm brain-computer interface: A 
practical and convenient non-linear classifier. Biomedical Engineering. April 2015. DOI: 10.1515/bmt-2014-0117
DOI: 10.3217/978-3-85125-467-9-76Proceedings of the 6th International Brain-Computer Interface Meeting, organized by the BCI Society
Published by Verlag der TU Graz, Graz University of Technology, sponsored by g.tec medical engineering GmbH 76


Time domain classification of grasp and hold tasks

Hafner, Norbert

Lottersberger, Florian

Intralogistics systems - optimization of energy efficiency

Municchi, Federico

Radl, Stefan

Goniva, Christoph

Development of closure models for unresolved particle simulations of heat and mass transfer in dense particle beds using CPPPO: a library for fast parallel data filtering

Gilch, Lorenz Alexander

Asymptotic Entropy of Random Walks on Regular Languages over a Finite Alphabet

Renner, Herwig

Erbringung von Systemdienstleistungen in zukünftigen Übertragungs- und Verteilnetzen

Aichhorn, Markus

Pourovskii, Leonid

Seth, Priyanka

Vildosola, Veronica

Zingl, Manuel

Peil, Oleg

Deng, Xiaoyu

Mravlje, Jernej

Kraberger, Gernot J.

Martins, Cyril

Ferrero, Michel

Parcollet, Olivier

TRIQS/DFTTools: A TRIQS application for ab initio calculations of correlated materials

Graz University of Technology

Not a member yet