Available online at www.sciencedirect.

com

ScienceDirect
Procedia CIRP 16 (2014) 205 – 210

Product Services Systems and Value Creation. Proceedings of the 6th CIRP Conference on Industrial
Product-Service Systems

A Method for Impact Analysis of Cyclic Changes within Innovation

Processes of PSS
Daniel Kasperek*, Nepomuk Chucholowski, Sebastian Maisenbacher, Udo Lindemann,
Maik Maurer
Institute of Product Development, Technische Universität München, Munich, Germany

* Corresponding author. Tel.: +49-89-289-15140; fax: +49-89-289-15144. E-mail address: Kasperek@pe.mw.tum.de

Abstract

Innovation processes are subject to various internal and external cycles. Changes of the environment of the system from a
business perspective such as the readiness of new technologies or shifting requirements due to changing customer demands, or
other cycles can affect the system as they significantly influence its innovation process. Especially companies providing PSS are
affected by these cycles because physical product and service aspects as well as a high number of stakeholders have to be
integrated during the design and development phase. While in conventional development processes dynamics caused by the
occurrence of cycles already play a major role, the influence of dynamic aspects in the development of PSS is further enhanced.
In particular, the high degree of crosslinking between the different phases of the innovation process leads to highly dynamic
responses within the innovation process. A lack of knowledge and the challenging handling of these cycles and their
dependencies and effects lead to challenges. In highly dynamic systems, the classical approaches of process modelling often
reach their limits, as the depicted elements and relations are usually assumed to be static. Consequently, it is hard to predict the
dynamic impact of cyclic changes on the innovation process of PSS. There are approaches to model the dynamics of systems
such as System Dynamics, agent-based modelling or discrete event simulation. Especially System Dynamics is used throughout
the public and private sector for policy analysis and design. It is a method to analyse and simulate the dynamic behaviour of
systems on a high level of abstraction. Therefore, within this paper, a System Dynamics based impact analysis of cyclic changes
within innovation processes of PSS is suggested. We provide background information on the innovation process of PSS, how
they are challenged by cycles and how this approach can be used for modelling innovation processes and on the influence of
cycles on these processes. An exemplary case study is presented to show the application of our approach on an innovation
process of a PSS.
© 2014
© 2014 The Authors.
Elsevier B.V.Published
This is anbyopen
Elsevier B.V.
access article under the CC BY-NC-ND license
Selection and peer-review under responsibility of the International Scientific Committee of “The 6th CIRP Conference on Industrial Product-
(http://creativecommons.org/licenses/by-nc-nd/3.0/).
Service Systems”
Selection in the person
and peer-review of the
under Conference of
responsibility Chair
the Professor HodaScientifi
International ElMaraghy.
c Committee of “The 6th CIRP Conference on Industrial
Product-Service Systems” in the person of the Conference Chair Professor Hoda ElMaraghy”
Keywords: product-service system; system dynamics; innovation process; impact analysis; cycle

* Corresponding author. Tel.: +49 89 289 15140; fax: +49 89 289 15144.
E-mail address: Kasperek@pe.mw.tum.de

2212-8271 © 2014 Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/3.0/).
Selection and peer-review under responsibility of the International Scientific Committee of “The 6th CIRP Conference on Industrial
Product-Service Systems” in the person of the Conference Chair Professor Hoda ElMaraghy”
doi:10.1016/j.procir.2014.02.008
206 Daniel Kasperek et al. / Procedia CIRP 16 (2014) 205 – 210

1. Introduction particular interest.”

Fig. 1. illustrates the components of PSS.
Various company-internal and -external influence factors
affect the innovation processes of product-service systems by
causing changes to the process and its results [1,2]. Examples
for these factors are changed staff, changed organizational
structures, new company strategies, revised laws or
regulations, new available technologies or changed customer
Product Service
requirements. However, it is not clear how exactly these
influence factors affect the innovation process and the related PSS
effects of changes within the innovation process are unknown. Fig. 1. Scheme of a PSS [1]
Hence, especially companies providing PSS face a lot of
uncertainties when dealing with such impacts on the process, Tukker [4] differentiates three categories of PSS:
since the high degree of interdependencies between product
x Product-oriented service: the business model focuses
and service aspects, as well as the processes to generate them,
on a conventional sale of a product with some extra
are hard to overlook. services added.
The lack of knowledge about influence factors and about
their impact on innovation processes makes it difficult to x Use-oriented service: the product is owned by the
effectively deal with them. Yet, most of these impacts occur provider and sometimes shared by a number of users,
in reoccurring patterns, referred to as ‘cycles’ [1]. By while it still plays a central role in the business model.
analyzing innovation processes from the past, it is possible to x Result-oriented services: Customers and clients agree
derive probabilities for the time of occurrence of a cyclic on a result without the involvement of a predetermined
influence (e.g. the often cited Kondratiev-cycles [3]) and to product.
identify the aspects of the innovation process that have been
We use the term innovation process in the context of PSS
affected by these impacts, respectively. However, in order to
based on the basic model of innovation processes of [2] which
estimate the potential effects of cyclic influences on
is illustrated in Fig. 2. The model subdivides the innovation
innovation processes prospectively, the dynamic response of
process into seven discrete phases, embracing the appearance
the tightly cross-linked processes has to be simulated, whereas
of customer demand, its analysis, idea generation, product and
all conceivable points in time for the occurrence of an impact
service design, production, introduction and usage of the PSS.
have to be considered. For this purpose, today’s classical
The double-tracked bus, connecting the phases, emphasizes
approaches of process modelling, which usually regard
the cyclic character of innovation processes.
process elements and relations to be static are insufficient.
This is why we propose to use a modelling technique that
Demand
enables the consideration of a dynamic model behavior for the
modelling of PSS innovation processes.
Analysis
Before we introduce our System Dynamics PSS model in
section 3, we provide some background information on the
innovation processes of PSS, cyclic influence factors and the Idea generation
modelling technique that we used called System Dynamics in
section 2. Section 4 describes the effects of a cyclic impact Design
can be analyzed with the help of our model. An exemplary
simulation of a cyclic impact on our System Dynamics PSS Production
model is presented in section 5. A discussion of our research
results so far is given in section 6, before section 7 concludes Introduction
the paper and gives an outlook for the next research steps.
Utilization
2. Background information
Fig. 2. Basic model of the innovation process [2]
2.1. Innovation processes of PSS
2.2. Cyclic influences on innovation process of PSS
We use a definition of the term PSS that especially focuses
on the relevant aspects in the context of cycle management
Schenkl et al. [1] define cycles as temporally or
from [1] who derived this definition based on the definitions
structurally reoccurring patterns which may be divided into
of [4,5]:
phases and gives a list of relevant cycles within the innovation
“A Product-Service System (PSS) integrates product and
process of product-service systems. Furthermore [1]
service components, whereby the product components may
consist of mechanics, electrics/electronics and software. In the characterizes cycles by the features of repetition, phases,
context of cycle management for innovation processes, the duration, trigger and effects. Triggers are thereby seen as
dynamics and complexity involved in understanding, initiating events for the repetition of a phase and result from a
planning, and managing Product-Service Systems are of
 Daniel Kasperek et al. / Procedia CIRP 16 (2014) 205 – 210
deviation of the actual state from the nominal state of an
object.
This repetition requires a specific time and causes individual
effects, which may lead to yet another cycle [2,6]. The terms
of the Fig. are explained below (cp. [6]):
Repetition
Trigger Phase 1 Phase 2 Phase n Effects
Duration
Fig. 3. Scheme of a cycle [1]
x Trigger: intentional or unintentional event(s) that
initiate(s) an iteration of a cycle
x Repetition: cycles represent iterative, reoccurring
patterns
x Phase: a cycle may be divided into several more or
less distinctive phases
x Effect: cycles have one or multiple effects
x Duration: needed time for one iteration of a cycle
2.3. System Dynamics modelling
The System Dynamics methodology is a mathematical
modelling technique to study the dynamic behavior of
systems utilizing a visual modelling and simulation technique
[7,8,9]. Forrester [10] developed System Dynamics (SD) in
the late 50s. Currently it is used throughout the public and
private sector for policy analysis and design. An overview of
the SD methodology can be found in [11]. Next to SD, there
are various other methods for analyzing and simulating
dynamic systems, such as agent-based modelling [7,12,13]
and discrete event modelling [13,14]. While discrete event
modelling is more suitable for simulations at the operational
level, SD is especially suitable for simulation at high levels of
abstraction and agent-based modelling can be used at all
levels of abstraction [13,15]. Due to the interdisciplinary
approach of SD [7], it is an adequate simulation method for
modelling and simulating models which are positioned at the
strategic level. [9,16]
SD is based on the simple graphical notations of causal
loop diagrams and stock and flow diagrams to model systems
[9]. The first step of the SD modelling process is the
development of qualitative influence models. As the
qualitative model is based on a graph notation of edges and
nodes it can theoretically be derived from models in other
notations that base on the same concept, such as graph or
dependency models [17]. The step of quantitative modelling is
then followed by the development of quantitative simulation
207
models. These models allow for a flexible illustration of
complex scenarios, like modelling different influencing
factors that occur at the same time during the engineering
design process activity. Thereby, the results of the simulation
are generated patterns over-time. [18]
Within SD models, decisions can be devised and tested in a
safe environment. Thus, they work as experimental
management laboratories. Their feedback perspective and
“what-if” capability enables systematic problems to be
identified, understood and managed. [18]
Le [19] gives an overview about SD models in product
development to illustrate the iterative nature of these
processes. As innovation processes include development
processes the overview can be also applied on innovation
processes. There already exist a couple of System Dynamics
models to display iterations in development processes. Mainly
they are based on a generic structure that models flows of
activities between stocks that represent the current state of
execution. The main component of System Dynamics models
in iterative processes is the rework cycle, first mentioned by
[20]. Later authors have developed variants of the rework
cycle, such as [21]. In Fig. 4, an adapted single phase version
of the rework cycle is shown.
The rework cycle operates as follows: The variable
definition A defines the initial value of Work Remaining A.
The work tasks then flow as a parallel flow of work with
errors (amount defined by work quality A) accumulating into
Undiscovered Rework A. Only if these errors are detected,
rework will be necessary. The work that has to be reworked is
described by the rework discovery A rate which processes the
work to the Work Remaining A stock. and a flow back from
Work Accomplished A into Work Remaining A. An additional
feedback is implemented from Work Accomplished A to Work
Remaining A. The feedback is controlled by the corrupt A rate
which is triggered by external events. The modelling construct
with these two feedback loops allows to differentiate between
internal rework within the phase (controlled by rework
discovery A), such as conventional iterations during the design
phase, and rework due to external events, such as the
occurrence of cyclic impacts. The rework cycle finishes if the
amount of accomplished work is equal to the sum of initial
work to do and externally triggered work.
Fig. 4. Adapted single phase version of the rework cycle
208 Daniel Kasperek et al. / Procedia CIRP 16 (2014) 205 – 210
3. The System Dynamics PSS model
Fig. 5 illustrates the System Dynamics model of the
innovation process of product-service systems as suggested by
the authors. Seven rework cycles are used to model the seven
phases of innovation processes as proposed by [2]: Demand,
analysis, idea generation, design, production, introduction and
utilization. The strength of the rework cycle construct is to
illustrate the iterative nature of the innovation process and its
phases. It is used as it is not only capable to illustrate the
inherent iterations of the different phases but also to illustrate
the rework originating from the cyclic impacts to be modeled.
As the degree of iterations to be expected differs for the
different phases of the innovation process, the rework cycles
for the particular phases can be customized to the average
amount of rework necessary for the particular phase. While
especially the phases of idea generation and design usually are
subject to a high amount of iterations, phases such as demand
or production are usually not subject to much iteration.
Nevertheless cyclic changes can cause iterations in every
phase and therefore each phase is modeled by the composite
construct of a rework cycle.
Fig. 5. System Dynamics model for innovation processes of product-
service systems
The rework work cycles are sequentially connected to each
other: When the demand phase is finished, the analysis phase
starts. Degrees of parallelism, as common for any innovation
process, can also be easily incorporated in the model: The
starting condition of a task can be changed in the range from
“previous task 100% done” to “previous task 0% done (start
with start of previous tasks)” to model all possible degrees of
parallelism.
The model as illustrated in Fig. 5 is a base model for
innovation processes of PSS which can be further customized
for each application. The base model does not include
problem specific factors such as the motivation or workload
of the involved persons. Depending on the model purpose,
such factors can play an important role and might also need to
be included in the model depending on the particular
application. For these cases the model presented here can be
used as a base model.
4. Impact analysis of cyclic changes
The System Dynamics model presented in Fig. 5 is used
for the impact analysis of cyclic changes. Cyclic changes can
theoretically be triggered at any point of time by an initiating
event. Depending on the current phase of the innovation
process of the PSS the occurring cycle can have different
effects respectively impact on the system. For example a
cycle of technology change that occurs before the design and
especially the production phases start will probably have less
impact on the innovation process as if the technology change
cycle occurs after the production phase has started. The
reason is that the effects of the cycle can be incorporated in
the initial design of the PSS production if the technology
change occurs prior to the design of production. If the
innovation process is in post-production (respectively
production design) phases, the new technology that needs to
be incorporated in the PSS production will cause a feedback
of the innovation process to the design (of production) phase.
5. Evaluation of applicability
To illustrate the applicability of the proposed procedure a
fictitious academic example is chosen: It represents an
innovation process of a PSS with the mentioned seven phases.
The authors are aware of the fact that the evaluation does not
prove the applicability of the procedure on an industrial level,
but want to show the general idea behind it and therefore
choose a simple example.
Fig. 6 shows the System Dynamics model for innovation
processes of PSS extended by a cycle of technology change
affecting the production phase of the PSS.
 Daniel Kasperek et al. / Procedia CIRP 16 (2014) 205 – 210
Fig. 6. System Dynamics model for innovation processes of PSS
extended by a cycle of technology change affecting the production phase of
the PSS
Fig. 7 provides a zoom in the particular area of the System
Dynamics model where the cyclic impact is realized.
For the example, each of the seven innovation process
phases is assumed with equal duration if no cyclic impacts
occur. The technology change cycle is implemented as a step
function that adds additional work-to-do to the remaining
work in this phase. For the cycle of technology change four
different scenarios are examined:
x The cycle does occur prior to the production phase start
x The cycle occurs while the production phase is running
x The cycle occurs after the production phase is finished
x The cycle does not occur
Fig. 7. Zoom into the production phase including the impact of a cycle of
technology change
These four scenarios are implemented by different shapes
of the used step function. The shape of the step function can
be changed in three ways: by the height of the step (amount of
209
additional work to do), by the discrete time of the occurrence
(occurrence date of the cycle), and by the length of the
occurrence (occurrence duration of the cyclic impact). For the
four scenarios only the discrete time of the occurrence was
varied and the over parameters were kept constant. The
impact of the different scenarios on the accomplished work
within the production phase is illustrated in Fig. 8. The left
axis of the illustrated table shows the amount of finished work
packages. The initial value of work packages was 1000, so
that the whole production phase cycle was assumed to be
finished if the work accomplished value reached this value.
While the overall duration of the innovation process was 73
months, the production phase was assumed to be finished
after 61 months without the occurrence of any cyclic impacts.
Fig. 8 shows the overall duration of the production phase for
the four examined scenarios.
Fig. 8. Impact of the different scenarios on the accomplished work within
the production phase into the production phase cycle
It can be seen that if the cycle occurs before the production
phase, as well as if the cycle occurs while the production
phase, the overall duration of the phase is smaller than if the
cycle of technology change occurs after the actual end of the
production phase. This can be explained by the additional
rework that is necessary to change the PSS after the
production is finished. The fact that the production phase
finishes at the same time – whether the cycle occurs in front
or within the affected phase – can be explained by the
modelling assumption that it is not differentiated at what point
of time within the particular phase the impact occurs. For
instance, it is assumed that the amount of additional rework is
the same whether the impact occurs before, at the beginning,
or at the end of the production phase as long as it occurs
before the phase is finished.
6. Discussion
Even though the presented exemplary case study is
strongly simplified, the application of our approach to apply
System Dynamics models of the PSS innovation process to
analyze the impact of cyclic changes shows the potential of
dynamic structure modelling techniques for the behavioral
analysis. System Dynamics is a powerful tool which enables
to further develop the PSS innovation process model. As this
publication is a result of ongoing research activities, there is
more research to conduct to reach a level of industrial
210 Daniel Kasperek et al. / Procedia CIRP 16 (2014) 205 – 210
applicability: It might be necessary to take more boundary
conditions, such as different attributes of the particular phases
into account. Right now, the process is assumed as linear with
a certain degree of parallelization, but without feedback loops
between the different phases. Furthermore, a lot more
potential influence factors and more various impacts are
conceivable. With System Dynamics, more interdependencies
and influences can be added to the process model.
As for any model, the quality of the System Dynamics
model is dependent on the underlying assumptions. The
assumptions used within this model can be further improved
to closer approximate the real system.
7. Conclusion and outlook
The paper presents a method for the impact analysis of
cyclic changes within innovation processes of product-service
systems. The method proposes System Dynamics in order to
model and simulate the impacts of cyclic changes on product-
service systems. A base model for modelling the impacts on
the innovation process is provided. This base model can be
further customized for the particular application of impact
analysis of cyclic changes within PSS innovation processes.
An exemplary case study shows the applicability of the model
and illustrates potential further use cases of the suggested
method. The next steps will be to model different PSS
innovation processes from real life examples on the one hand,
and to model different impact factors on these processes on
the other hand. Based on descriptive study on how the
different aspects within innovation processes of PSS are
interrelated the model will be further enhanced. With an
established System Dynamics model at hand, it is possible to
analyze several scenarios of cyclic influences on an
innovation process by simulation, which will help to further
improve the planning and management of innovation
processes of PSS by providing a tool to handle cyclic impacts.
Acknowledgements
We thank the German Research Foundation (Deutsche
Forschungsgemeinschaft – DFG) for funding this project as
part of the collaborative research centre
‘Sonderforschungsbereich 768 – Managing cycles in
innovation process – Integrated development of product-
service systems based on technical products’.
References
[1] S. a. Schenkl, F. G. H. Behncke, C. Hepperle, and S. Langer, “Managing
Cycles of Innovation Processes of Product-Service Systems,” 2013 IEEE
Int. Conf. Syst. Man, Cybern., pp. 918–923, Oct. 2013.
[2] S. Langer and U. Lindemann, “Managing cycles in development
processes-analysis and classification of external context factors,” in 17th
International Conference on Engineering Design, 2009, pp. 1–550.
[3] G. Garvy, “Kondratieff’s theory of long cycles,” Rev. Econ. Stat., vol.
25, no. 4, pp. 203–220, 1943.
[4] A. Tukker, “Eight types of product–service system: eight ways to
sustainability? Experiences from SusProNet,” Bus. Strateg. Environ.,
vol. 13, no. 4, pp. 246–260, Jul. 2004.
[5] T. S. Baines, H. W. Lightfoot, S. Evans, A. Neely, R. Greenough, J.
Peppard, R. Roy, E. Shehab, A. Braganza, A. Tiwari, J. R. Alcock, J. P.
Angus, M. Bastl, A. Cousens, P. Irving, M. Johnson, J. Kingston, H.
Lockett, V. Martinez, P. Michele, D. Tranfield, I. M. Walton, and H.
Wilson, “State-of-the-art in product-service systems,” Proc. Inst. Mech.
Eng. Part B J. Eng. Manuf., vol. 221, no. 10, pp. 1543–1552, Jan. 2007.
[6] S. Langer, C. Knoblinger, and U. Lindemann, “Analysis of dynamic
changes and iterations in the development process of an electrically
powered go-kart,” in 11th International Design Conference DESIGN
2010, 2010.
[7] J. Sterman, Business dynamics:Systems thinking and modeling for a
complex world. Boston: Irwin-McGraw-Hill, 2000, p. 890.
[8] J. Morecroft, Strategic modelling and business dynamics: A feedback
systems approach. Wiley, 2007.
[9] H. Meier and M. Boßlau, “Design and Engineering of Dynamic Business
Models for Industrial Product-Service Systems,” in in The Philosopher’s
Stone for Sustainability, Berlin: Springer, 2013, pp. 179–184.
[10] J. W. Forrester, Industrial dynamics, vol. 2. Cambridge, USA: MIT
Press, 1961.
[11] A. G. Rodrigues, “The application of system dynamics to project
management: an integrated methodology (SYDPIM),” University of
Strathclyde, 2000.
[12] T. Terano, H. Kita, and T. Kaneda, Real-world Applications: Post-
proceedings of the Third International Workshop on Agent-Based
Approaches in Economic and Social Complex Systems 2004, vol. 1.
Tokyo; Japan: Springer, 2005.
[13] O. Grasl, “Professional service firms : business model analysis ; method
and case studies,” University St. Gallen, Switzerland, 2009.
[14] M. Pidd, Computer Simulation in Management Science. New York,
USA: John Wiley & Sons, 1992, p. 372.
[15] A. Borshchev and A. Filippov, “From System Dynamics and Discrete
Event to Practical Agent Based Modeling: Reasons, Techniques, Tools,”
in Proceedings of the 22nd international conference of the system
dynamics society, 2004, no. 22.
[16] D. Kasperek, S. Maisenbacher and M. Maurer, “Structure-based
Analysis of Dynamic Engineering Process Behavior” in 2014 IEEE
International Systems Conference (SysCon 2014), 2014.
[17] D. Kasperek and M. Maurer, “Coupling Structural Complexity
Management and System Dynamics to represent the Dynamic Behavior
of Product Development Processes,” in 2013 IEEE International Systems
Conference (SysCon 2013), 2013, pp. 414–419.
[18] A. G. Rodrigues, “Managing and Modelling Project Risk Dynamics A
System Dynamics-based Framework Risk Management in Projects
Overview Current Framework for Project Risk Management,” in Fourth
European Project Management Conference: PMI Europe, 2001, vol. 7,
pp. 6–7.
[19] H. Le, D. C. Wynn, and P. Clarkson, “Re-designing PD process
architecture by transforming task network models into system dynamics
models,” 11TH Int. Des. Conf. (DESIGN 2010), vol. 1–3, pp. 319–330,
2010.
[20] K. Cooper, “Naval ship production: A claim settled and a framework
built,” Interfaces (Providence)., vol. 10, no. 6, pp. 20–36, 1980.
[21] D. N. Ford and J. D. Sterman, “Dynamic modeling of product
development processes,” Syst. Dyn. Rev., vol. 14, no. 1, pp. 31–68,
1998.