Significance and Challenges of Data-driven Product Generation and Retrofit Planning

Abstract

One of the notable drivers of the fourth industrial revolution is the collection of vast amounts of data along the entire lifecycle of a product. The analysis of product lifecycle data in conjunction with product hypotheses leads to promising potentials in strategic product planning. In this thesis paper, we postulate the need for data-driven product generation and retrofit planning as an interdisciplinary field of research. We define and analyze the key concepts and derive requirements in a structured way. Based on an exhaustive research of existing approaches, we structure open research questions and propose a roadmap in order to shape future research efforts.

Full text

ScienceDirect
Available online at www.sciencedirect.com
Available online at www.sciencedirect.com
ScienceDirect
Procedia CIRP 00 (2017) 000–000
www.elsevier.com/locate/procedia
2212-8271 © 2017 The Authors. Published by Elsevier B.V.
Peer-review under responsibility of the scientific committee of the 28th C IRP Design Conference 2018.
28th CIRP Design Conference, May 2018, Nantes, France
A new methodology to analyze the functional and physical architecture of
existing products for an assembly oriented product family identification
Paul Stief *, Jean-Yves Dantan, Alain Etienne, Ali Siadat
École Nationale Supérieure d’Arts et Métiers, Arts et Métiers ParisTech, LCFC EA 4495, 4 Rue Augustin Fresnel, Metz 57078, France
* Corresponding author. Tel.: +33 3 87 37 54 30; E-mail address: paul.stief@ensam.eu
Abstract
In today’s business environment, the trend towards more product variety and customization is unbroken. Due to this development, the need of
agile and reconfigurable production systems emerged to cope with various products and product families. To design and optimize production
systems as well as to choose the optimal product matches, product analysis methods are needed. Indeed, most of the known methods aim to
analyze a product or one product family on the physical level. Different product families, however, may differ largely in terms of the number and
nature of components. This fact impedes an efficient comparison and choice of appropriate product family combinations for the production
system. A new methodology is proposed to analyze existing products in view of their functional and physical architecture. The aim is to cluster
these products in new assembly oriented product families for the optimization of existing assembly lines and the creation of future reconfigurable
assembly systems. Based on Datum Flow Chain, the physical structure of the products is analyzed. Functional subassemblies are identified, and
a functional analysis is performed. Moreover, a hybrid functional and physical architecture graph (HyFPAG) is the output which depicts the
similarity between product families by providing design support to both, production system planners and product designers. An illustrative
example of a nail-clipper is used to explain the proposed methodology. An industrial case study on two product families of steering columns of
thyssenkrupp Presta France is then carried out to give a first industrial evaluation of the proposed approach.
© 2017 The Authors. Published by Elsevier B.V.
Peer-review under responsibility of the scientific committee of the 28th CIRP Design Conference 2018.
Keywords: Assembly; Design method; Family identification
1. Introduction
Due to the fast development in the domain of
communication and an ongoing trend of digitization and
digitalization, manufacturing enterprises are facing important
challenges in today’s market environments: a continuing
tendency towards reduction of product development times and
shortened product lifecycles. In addition, there is an increasing
demand of customization, being at the same time in a global
competition with competitors all over the world. This trend,
which is inducing the development from macro to micro
markets, results in diminished lot sizes due to augmenting
product varieties (high-volume to low-volume production) [1].
To cope with this augmenting variety as well as to be able to
identify possible optimization potentials in the existing
production system, it is important to have a precise knowledge
of the product range and characteristics manufactured and/or
assembled in this system. In this context, the main challenge in
modelling and analysis is now not only to cope with single
products, a limited product range or existing product families,
but also to be able to analyze and to compare products to define
new product families. It can be observed that classical existing
product families are regrouped in function of clients or features.
However, assembly oriented product families are hardly to find.
On the product family level, products differ mainly in two
main characteristics: (i) the number of components and (ii) the
type of components (e.g. mechanical, electrical, electronical).
Classical methodologies considering mainly single products
or solitary, already existing product families analyze the
product structure on a physical level (components level) which
causes difficulties regarding an efficient definition and
comparison of different product families. Addressing this
Procedia CIRP 84 (2019) 992–997
2212-8271 © 2019 The Authors. Published by Elsevier B.V.
Peer-review under responsibility of the scientific committee of the CIRP Design Conference 2019.
10.1016/j.procir.2019.04.226
© 2019 The Authors. Published by Elsevier B.V.
Peer-review under responsibility of the scientic committee of the CIRP Design Conference 2019.
Available online at www.sciencedirect.com
ScienceDirect
Procedia CIRP 00 (2019) 000–000
www.elsevier.com/locate/procedia
2212-8271 © 2019 The Authors. Published by Elsevier B.V.
Peer-review under responsibility of the scientific committee of the CIRP Design Conference 2019
29th CIRP Design 2019 (CIRP Design 2019)
Significance and Challenges of Data-driven Product Generation and
Retrofit Planning
Melina Massmanna,*, Maurice Meyera, Prof. Dr.-Ing. Roman Dumitrescua, Sebastian von Enzbergb,
Maximilian Franka, Christian Koldeweya, Dr.-Ing. Arno Kühnb, Jannik Reinholda
aChair Advanced Systems Engineering, Heinz Nixdorf Institute, University of Paderborn, Fürstenallee 11, 33102 Paderborn, Germany
bFraunhofer IEM, Zukunftsmeile 1, 33102 Paderborn, Germany
* Corresponding author. Tel.: +49-5251-5465-346. E-mail address: msl@campus.uni-paderborn.de
Abstract
One of the notable drivers of the fourth industrial revolution is the collection of vast amounts of data along the entire lifecycle of a product. The
analysis of product lifecycle data in conjunction with product hypotheses leads to promising potentials in strategic product planning. In this
thesis paper, we postulate the need for data-driven product generation and retrofit planning as an interdisciplinary field of research. We define
and analyze the key concepts and derive requirements in a structured way. Based on an exhaustive research of existing approaches, we structure
open research questions and propose a roadmap in order to shape future research efforts.
© 2019 The Authors. Published by Elsevier B.V.
Peer-review under responsibility of the scientific committee of the CIRP Design Conference 2019
Keywords: Data-driven Product Planning; Industrial Data Analytics; Lifecycle Analytics; Generation and Retrofit Planning
1. Introduction
The Digital age is characterized by strong technical develop-
ments, automated systems and connectivity. By integrating sen-
sors, actuators and advanced information processing technol-
ogy, products evolve from mechatronic systems into cyber-
physical systems (CPS), which interact with the physical and
digital world [1], [2]. A further paradigm in this context is the
Internet of Things (IoT) which aims to make devices and ob-
jects smarter by allowing communication with each other.
Applying and integrating the ideas of CPSs and IoT to the
industrial automation domain leads to the concept Industrie 4.0,
the so called fourth industrial revolution. While there are meth-
ods to evaluate and increase a company’s performance in the
context of Industrie 4.0, as described in the research project
“INLUMIA” [3][4], the effects of the fourth industrial revolu-
tion on the product development process needs further research.
It is well known that the combination of the aforementioned
technologies as well as next-generation industrial infrastructure
with highly improved storage and computation abilities, and ad-
vanced data analytics techniques enable the collection and ex-
amination of vast quantities of data with the purpose of
knowledge discovery. Data analytics combines mathematics
and statistics, computer science, and a domain-specific field of
application. The application of data analytics in the manufac-
turing domain is called Industrial Data Analytics or Industrial
Data Science [5] and is considered one of the critical success
factors for Industrie 4.0 [6]. Industrial Data Analytics can be
applied to main company processes as well as supporting activ-
ities in the value chain, e.g. research and development [7].
Product Development in particular is attributed a growing
strategic importance, as the development of new products is as-
sociated with high uncertainty for a company, since today's
products are developed for the market of tomorrow. Moreover,
international competition and individual customer requirements
place high demands on the development of new products [8].
To enable customer oriented, time- and cost-efficient processes,
the holistic use and analysis of data along the product lifecycle
Available online at www.sciencedirect.com
ScienceDirect
Procedia CIRP 00 (2019) 000–000
www.elsevier.com/locate/procedia
2212-8271 © 2019 The Authors. Published by Elsevier B.V.
Peer-review under responsibility of the scientific committee of the CIRP Design Conference 2019
29th CIRP Design 2019 (CIRP Design 2019)
Significance and Challenges of Data-driven Product Generation and
Retrofit Planning
Melina Massmanna,*, Maurice Meyera, Prof. Dr.-Ing. Roman Dumitrescua, Sebastian von Enzbergb,
Maximilian Franka, Christian Koldeweya, Dr.-Ing. Arno Kühnb, Jannik Reinholda
aChair Advanced Systems Engineering, Heinz Nixdorf Institute, University of Paderborn, Fürstenallee 11, 33102 Paderborn, Germany
bFraunhofer IEM, Zukunftsmeile 1, 33102 Paderborn, Germany
* Corresponding author. Tel.: +49-5251-5465-346. E-mail address: msl@campus.uni-paderborn.de
Abstract
One of the notable drivers of the fourth industrial revolution is the collection of vast amounts of data along the entire lifecycle of a product. The
analysis of product lifecycle data in conjunction with product hypotheses leads to promising potentials in strategic product planning. In this
thesis paper, we postulate the need for data-driven product generation and retrofit planning as an interdisciplinary field of research. We define
and analyze the key concepts and derive requirements in a structured way. Based on an exhaustive research of existing approaches, we structure
open research questions and propose a roadmap in order to shape future research efforts.
© 2019 The Authors. Published by Elsevier B.V.
Peer-review under responsibility of the scientific committee of the CIRP Design Conference 2019
Keywords: Data-driven Product Planning; Industrial Data Analytics; Lifecycle Analytics; Generation and Retrofit Planning
1. Introduction
The Digital age is characterized by strong technical develop-
ments, automated systems and connectivity. By integrating sen-
sors, actuators and advanced information processing technol-
ogy, products evolve from mechatronic systems into cyber-
physical systems (CPS), which interact with the physical and
digital world [1], [2]. A further paradigm in this context is the
Internet of Things (IoT) which aims to make devices and ob-
jects smarter by allowing communication with each other.
Applying and integrating the ideas of CPSs and IoT to the
industrial automation domain leads to the concept Industrie 4.0,
the so called fourth industrial revolution. While there are meth-
ods to evaluate and increase a company’s performance in the
context of Industrie 4.0, as described in the research project
“INLUMIA” [3][4], the effects of the fourth industrial revolu-
tion on the product development process needs further research.
It is well known that the combination of the aforementioned
technologies as well as next-generation industrial infrastructure
with highly improved storage and computation abilities, and ad-
vanced data analytics techniques enable the collection and ex-
amination of vast quantities of data with the purpose of
knowledge discovery. Data analytics combines mathematics
and statistics, computer science, and a domain-specific field of
application. The application of data analytics in the manufac-
turing domain is called Industrial Data Analytics or Industrial
Data Science [5] and is considered one of the critical success
factors for Industrie 4.0 [6]. Industrial Data Analytics can be
applied to main company processes as well as supporting activ-
ities in the value chain, e.g. research and development [7].
Product Development in particular is attributed a growing
strategic importance, as the development of new products is as-
sociated with high uncertainty for a company, since today's
products are developed for the market of tomorrow. Moreover,
international competition and individual customer requirements
place high demands on the development of new products [8].
To enable customer oriented, time- and cost-efficient processes,
the holistic use and analysis of data along the product lifecycle

Melina Massmann et al. / Procedia CIRP 84 (2019) 992–997 993
2 Author name / Procedia CIRP 00 (2019) 000–000
offers large potentials, in particular for generation and retrofit
planning. Providing valuable information about activities in all
lifecycle phases of the product, about the users (e.g. machine
operator) and their usage of products allows receiving valuable
and objective information about the current state, failures and
incorrect operations. Using this information it is possible to de-
velop the next product generation, which is better adapted to the
actual needs and less error-prone, or to retrofit machines in use.
Using data and deriving insights out of them for product plan-
ning and product development has been a success factor in other
industries, e.g. technology and online business. Nevertheless, it
seems not to be in focus in industrial context yet. Some chal-
lenges are associated with this research field, we call data-
driven product planning or more specific data-driven genera-
tion and retrofit planning, in industrial manufacturing. These
comprise the utilization of usage data and the integration of all
the relevant and information-containing data, which are wide-
spread along different business units and the whole product life-
cycle.
2. Foundations
2.1. 4-Cycle-Model of Product Engineering
Strategic Product Planning is part of the process of Product
Engineering, which cannot be seen as a stringent sequence of
process steps but rather as an interplay of tasks that can be struc-
tured into four cycles (Fig. 1) [9].
1) In the Strategic Product Planning, the requirements for prod-
ucts are systematically identified to capture the markets of to-
morrow. It contains the three tasks: foresight, product discover-
ing, and business planning. The aim of foresight is to identify
the potentials for future success as well as the relevant business
options. The objective of product discovering is to find new
product ideas that exploit the detected potentials for success.
Business planning initially deals with the business strategy.
That means, it addresses the questions which market segments
should be covered. The product strategy is then elaborated on
this definition.
2) The Product Development includes the conceptual product
design, the domain-specific design and the appropriate prepara-
tion as well as the integration of the results of the individual
domains to a total solution.
3) The goal of the cycle Service Development is the realization
of a service idea into a market service. This is an interplay of
tasks, namely service design, service planning and service inte-
gration.
4) The starting point of the forth cycle Production System De-
velopment is the conception of the production system. Here the
four aspects of work planning, work equipment planning, work-
place planning and production logistics have to be regarded in
an integrative way.
For the concept of data-driven product planning we only
consider the first cycle – Strategic Product Planning.
2.2. Retrofit and Generation Planning
The development of new product generations is a common
approach to fulfill objectives of the Strategic Product Planning.
Product generation planning deals with the strategic planning
of new product generations. An important task of this procedure
is to elicit ideas for new product features based on an existing
reference product [10]. Another strategic approach, which is to
be on product-improving measures, is not to replace old sys-
tems but to retrofit them and hence to extend the usage phase.
In the course of the generation planning, it is therefore neces-
sary to check which features are to be retrofitted and which are
to be implemented in the new product generation.
2.3. Product Lifecycle Management and Product Lifecycle
The Product Lifecycle Management (PLM) approach aims
to manage the product related information efficiently during the
whole product lifecycle. According to Abramovici, PLM is a
strategic management approach that consists of integrated
methods and tools that are used to cooperatively create, man-
age, and apply all product-relevant engineering information
throughout the product lifecycle [11]. PLM is not to be under-
stood as an information technology (IT) system, but rather as a
management concept that uses IT systems to support it. The
goal of PLM systems is to save all necessary data about the
product in every lifecycle phase and to provide them when
needed.
The product lifecycle constitutes the basis of PLM. We con-
sider the intrinsic product lifecycle [12], [13]. Figure 2 on the
left side illustrates the product lifecycle as a cycle of successive
product lifecycle phases, from Strategic Product Planning to
product disposal. One product generation is represented here. A
retrofit is located in the usage phase and extends it.
Along the lifecycle, from the development over manufactur-
ing, usage to disposal, there are many data sources and hetero-
geneous data. Large amounts of structured and unstructured
data are abundant around the product lifecycle.
Fig. 1. 4-Cycle Model of Product Engineering [9]

994 Melina Massmann et al. / Procedia CIRP 84 (2019) 992–997
Author name / Procedia CIRP 00 (2019) 000–000 3
On the other side many assumptions in form of hypotheses
about the product usage, user behavior and product properties
exist along the lifecycle and in the functional areas of the com-
pany. In the planning and development phase product managers
and design engineers formulate many hypotheses in order to
have a foundation for relevant and concrete ideas in design
planning.
Together these foundations form the basic idea of our con-
cept of data-driven generation and retrofit planning, which is
depicted in figure 2. We derive benefit from the data along the
lifecycle by joining it and using data analytics to validate hy-
potheses and explore new ones. These processes can be com-
bined in an instrument with the ability to assist engineers in the
analysis and interpretation into new product features. We push
back these features into generation or retrofit planning.
3. Analysis of the problem
The implementation of a data-driven generation and retrofit
planning has some challenges and open questions to consider.
In the following subsections, we introduce four problem fields
and postulate requirements (R) to data-driven product planning
(Fig. 3), from which challenges and research questions can be
derived.
3.1. Grounding on Product Hypotheses
Strategic Product Planning is typically based on hypotheses
about the actual and future customer requirements. With regard
to a lifecycle data-driven product planning a systematic ap-
proach for finding and defining customer and product hypoth-
eses is necessary. This enables a consistent collection and for-
malized description for all hypotheses arriving along the
lifecycle, which is adequate for analysis purposes and interpre-
tation in the sense of a retrofit and generation planning. An ex-
ample use case is the automated selection of content-related fo-
cus for product hypotheses. In addition, decisions about simple
hypotheses or hypotheses sequences have to be made. Further-
more, the approach has to consider a prioritization concerning
the probability of occurrence and effects. The following key
requirements arise:
1. Product property and product usage hypotheses have to be
systematically identified along the lifecycle (R1).
2. Hypotheses have to be set up in a form so that engineers can
easily interpret them against the background of retrofit and gen-
eration planning (R2).
3.2. Deriving added value out of data
Product data analytics plays an essential role in data-driven
generation and retrofit planning. It can be divided into three sub
fields: data inventory, data acquisition and data analysis. These
also form three of the layers of the four-layer model for analyt-
ics use cases [14].
Data inventory in data-driven generation and product planning
addresses the problem of identifying and integrating all the rel-
evant data sources along the lifecycle in connection with data
properties, evaluation and structuring.
To be able to process the identified data efficiently, solutions
for data acquisition are prerequisite to organize, store and pro-
cess the data. There are a variety of big data analytics platforms
that attempt to translate the technical and functional require-
ments of these processes into one architecture. A data acquisi-
tion solution for data-driven product planning has to fulfill new
requirements. Features such as filtering and categorizing,
merging multiple data sources and secure storage are tailored
to the specific needs of the inventoried data. To collect missing
data, retrofit strategies for sensors have to be developed. Still
in many cases sensors for data acquisition are missing. In data
analysis for product development, two perspectives can be dis-
tinguished 1) the validation of hypotheses by using hypothesis
testing and 2) the identification of new patterns, dependencies,
correlations and hypotheses by explorative data mining.
For data-driven product planning both approaches should be
considered. A hypotheses-driven approach is common, how-
ever a large potential for added value lies in the identification
of previously unknown hypotheses and findings. Behind all an-
alytic activities, the analysis across different data sources has
to be considered. Consequently, the following requirements are
to state:
3. All Data Sources of different types have to be systematized
and integrated along the lifecycle (R3).
4. Data Sources have to be inventoried and evaluated in the
context of their processes and properties (R4).
5. Adequate technologies for the holistic collection, storage and
processing of this data have to be chosen. (R5).
6. Techniques and procedures have to be found which can pro-
vide valuable new insights and generate hypotheses in an auto-
matic way (connection to 3.1) (R6).
7. Appropriate hypothesis models for the available data and the
hypotheses to be tested are necessary (R7).
Fig. 3. Problem fields in data-driven Generation and Retrofit Planning
Fig. 2. Concept of data-driven generation and retrofit planning

Melina Massmann et al. / Procedia CIRP 84 (2019) 992–997 995
4 Author name / Procedia CIRP 00 (2019) 000–000
3.3. Renewing product planning
After analyzing the collected data along the lifecycle, data-
driven product planning requires to draw conclusions for prod-
uct planning. The results have to be interpreted and the result-
ing options for action to be evaluated in a structured way. It is
necessary for this to analyze the resulting options for action and
to develop norm strategies, which comprise guidelines for spe-
cific situations. In the course of generation planning, it has to
be checked which features are to be retrofitted and which are
incorporated in new product generations. In summary, this
leads to the following requirements:
8. Tested hypotheses have to be interpreted systematically in
the context of strategic product planning in order to derive mar-
ket-driven features (R8).
9. The effects of the results from the analysis on the products
should be studied (R9).
10. Uncovered features need to be classified into generation
planning or retrofit planning (R10).
3.4. Enabling operationalisation
In order to be able to operationalize, implement and inte-
grate data-driven product planning in the company, organiza-
tional integration is a prerequisite.
Like every new strategy, it needs a change management. This
requires appropriate approaches similar to those existing in the
context of Industry 4.0, digitization and data science [15], [16],
[167].
In addition, mechanical engineers currently only have limited
field data of the machines they supply, as the operators often
do not want to release data.
Thus, the following requirements arise:
11. The impact of establishing data-driven product planning on
the organization has to be clarified and activities to increase the
acceptance have to be implemented (R11).
12. The role and place of the data analyst in the organization
has to be defined. A classification of organizational and team
structures for types of companies is needed (R12).
13. Required Competences for companies and their employees
need to be defined (R13).
14. Operators of machines have to be motivated to provide the
data for analysis (R14).
4. State of research
After specification of the problem fields and requirements,
a clear picture of a comprehensive data-driven generation and
retrofit planning is painted. It includes an approach based on
hypotheses, a data analytics solution, an information feedback
into product planning and organizational operation – all based
on the lifecycle of the product. We have identified existing so-
lutions that cover one or more of the problem fields in an inten-
sive literature research. The following sections summarize the
main existing approaches and evaluates the requirements posed
in the section above.
4.1. Product Lifecycle Analytics
Kassner et al. proposed the product lifecycle analytics ap-
proach for the holistic integration and analysis of data from
multiple data sources around the product lifecycle [18].
Concerning our requirements R1-R14, data from all lifecycle
phases are encompassed, structured and unstructured data as
text information, are integrated (R3). Complex analytics capa-
ble of deriving novel insights with added value are applied
(R6). At first sight, one essential aspect is missing for a data-
driven product planning: the application to product planning
and product optimization, i.e. the interpretation of the analytics
results. Besides, a holistic data inventory does not occur where
data sources are identified, evaluated and structured.
4.2. Technical Inheritance
Demminger et al. regard the application of data in product
development, optimization and new product generations in the
concept of Technical Inheritance [19]. The approach is based
on the adapted principles of evolution in nature and assumes
that the efficient development of the next generation of smart
products is based on the storage and analysis of lifecycle data,
of previous generations of the product and relevant information
about their operation conditions. Here, the ideas of R3, R4 and
R8 can be recognized. However, the focus is on collection, stor-
age and exchange of the data. The concept disregards analytics
based on hypotheses (explorative as well as testing). Moreover,
the whole field of organizational implementation is neglected.
4.3. LeWiPro
In the research project „LeWiPro“ a software tool was de-
veloped that analyzes structured and unstructured data from the
product lifecycle and converts it into knowledge that can be
used in product development [20]. Nevertheless, live usage
data is not considered. Some ideas of R7 and R8 are presented,
but a complex analytics is missing, which observes different
data from different lifecycle phases simultaneously. The part of
interpretation of the analysis in the direction of product devel-
opment is completely up to the expert, no systematics is pro-
posed. The organizational implementation is not considered
such as an accompanied introduction of the tool.
4.4. WiRPro
The research project WiRPro had the goal of processing
product usage data and representing decision-relevant
knowledge for optimizing the next product generation [21].
The focus was on objective and structured data from the usage
phase. This project mainly does not address data and hypothe-
ses along the whole product lifecycle nor unstructured data as
text. Again, an organizational implementation is not regarded.
5. Research Roadmap
The literature overview reveals that, to the best of our
knowledge, no existing approach fulfills all the requirements
for a data-driven product planning. Therefore, in cooperation

996 Melina Massmann et al. / Procedia CIRP 84 (2019) 992–997
Author name / Procedia CIRP 00 (2019) 000–000 5
with the Technical University of Berlin and the University of
Applied Sciences Südwestfalen, we propose a research
roadmap in this chapter which is derived from the challenges
presented (Fig. 4). It is divided into five major cross-sectional
projects (CSP), each consisting of multiple minor projects.
First, in order to generate input for the data-driven product
planning, an inventory of product hypotheses needs to be cre-
ated. For that, hypotheses must be identified in all areas of a
company and for all phases of product usage (CSP 1.1). These
hypotheses must be described in a standardized pattern, so they
can be collected in the inventory. Next, from all the identified
hypotheses, the most suitable ones need to be extracted (CSP
1.2). This step is to be done by prioritizing on the basis of dif-
ferent factors, e.g. the strategical position of the company, suc-
cess factors and core competences. Based on the results of this
prioritization, the most suitable hypotheses will be chosen for
further data analysis.
Second, based on the chosen hypotheses, data needs to be
collected and analyzed. As hypotheses describe multiple per-
spectives of product usage, data must be gathered in all areas it
occurs and joined in a data inventory (CSP 2.1). Using a data
gap analysis, missing data for the verification or falsification of
hypotheses has to be revealed. Next, data acquisition is to be
planned (CSP 2.2). This contains the inquiry, saving and pro-
cessing of data. Furthermore, concepts for retrofitting sensors
as a consequence of the identified data gaps are to be defined.
Another important aspect is the identification of hypotheses
based on exploratory data analysis (CSP 2.3). By using these
algorithms, so far unknown connections and trends can be spot-
ted in the available data. These additional hypotheses comple-
ment the ones earlier identified and complete the product hy-
potheses inventory. Finally, combining the inventories of hy-
potheses and data, a validation of the hypotheses can be accom-
plished (CSP 2.4). This step creates the desired insights and
knowledge about the product usage based on real data.
Third, the process of Strategic Product Planning is to be sup-
ported by the results of the data analysis. Hence, these results
must be interpreted in the context of Strategic Product Planning
and consequences are to be identified in a subsequent analysis
(CSP 3.1). These consequences lead to new functions and fea-
tures which need to be derived using a systematic approach.
New functions and features can either be implemented in one
of the following product generations or in a retrofit of already
available machines. Therefore, in order to consider both tech-
nical and economical aspects as well as prevent product canni-
balization, methods for product generation (CSP 3.2) and ret-
rofit planning (CSP 3.3) need to be created. These methods
transfer the data analysis results into the process of strategic
product planning.
Fourth, the data-driven product planning must be operation-
alized. Only then companies can integrate them into their busi-
ness processes. As a start, a reference process for all the phases
from product hypotheses identification to product generation
and retrofit planning needs to be created (CSP 4.1). In order to
exploit the processes’ full potential, the company’s organiza-
tional structure must adapt to the new requirements. Therefore,
a reference organization structure and job profiles for employ-
ees in the data-driven product planning need to be defined (CSP
4.2). To be able to implement these new processes, the com-
pany must plan the acquisition of the competences described in
the job profiles (CSP 4.3). Furthermore, the company must find
ways to convince and encourage customers to share the data
measured in the machines (CSP 4.4). Considering fears and
wishes, a catalogue of actions can help companies to ensure the
availability of data. Hence, an incentive system for data acqui-
sition needs to be created.
Fifth, the operators need to be enabled to use the designed
methodologies efficiently. Therefore, a consistent tool support
is required. This includes tools for data analysis (CSP 5.1), cou-
pling of PLM and IoT-systems (CSP 5.2) as well as for product
generation and retrofit planning (CSP 5.3).
In addition to the five cross-sectional projects described, all
steps will be tested and verified in pilot projects (PP) with four
German companies: the forming technology specialist LASCO
Umformtechnik (PP 1), the Industrial Connectivity expert
Fig. 4. Research Roadmap Generation and Retrofit Planning

Melina Massmann et al. / Procedia CIRP 84 (2019) 992–997 997
6 Author name / Procedia CIRP 00 (2019) 000–000
Weidmüller Interface (PP 2), the financial and retail ATM pro-
ducer Diebold Nixdorf (PP 3) and the housing technology man-
ufacturer Westaflex (PP 4). All these pilot projects are sup-
ported by the German software developer CONTACT Software
and the German Industrie 4.0 platform developer AXOOM.
6. Summary and Conclusion
In the digital age, vast amounts of data can be collected
along the entire product lifecycle. Many potentials to create
added value by using the data have been discovered in various
industries. Most of them aim at creating new applications for
the customers. But due to increasingly stronger competition on
the international market, companies need to rethink their prod-
uct planning and development processes. By integrating data
analytics into the strategic product planning process, new and
so far unknown potentials can be revealed. Data-driven product
planning promises to unlock these potentials and adds a new
quality to product engineering. In the paper at hand, we identi-
fied challenges of data-driven product generation and retrofit
planning. Following this, we derived a research roadmap from
them. Five cross-sectional research projects cover the identifi-
cation of product hypotheses, data analytics for hypotheses ver-
ification, strategic product generation and retrofit planning, the
operationalization of the methodology as well as tool support.
Furthermore, the results will be verified in cooperation with
four companies.
7. Acknowledgement
This contribution was developed in the research project
“DizRuPt – Datengestützte Retrofit- und Generationenplanung
im Maschinen- und Anlagenbau” which is funded by the Fed-
eral Ministry of Education and Research (Bundesministerium
für Bildung und Forschung).
The foundation of this contribution was developed in the re-
search project “INLUMIA – Instrumentarium zur Leistungs-
steigerung von Unternehmen durch Industrie 4.0” which is
funded by the European Regional Development Fund.
References
[1] Broy M. Cyber-Physical Systems – Innovation durch Software-intensive
eingebettete Systeme, Springer, Berlin, Heidelberg 2010.
[2] Geisberger E, Broy, M. agendaCPS – Integrierte Forschungsagenda Cyber-
Physical Systems. Springer-Verlag, Berlin, Heidelberg, 2012
[3] Pierenkemper, C.; Drewel, M.; Gausemeier, J.:Zukunftsorientierte
Leistungssteigerung von Unternehmen durch Industrie 4.0. In: Smajic, H.
– Tagungsband AALE 2018, VDE Verlag, Berlin 2018
[4] Knospe, O.; Drewel, M.; Mittag, T.; Pierenkemper, C.; Hobscheidt, D.:
ZWF Zeitschrift für wirtschaftlichen Fabrikbetrieb, 2018 Band 113,
Ausgabe 1-2, S. 83-87
[5] Reinhart F. Industrial Data Science – Data Science in der industriellen
Anwendung. Industrie 4.0 Management, 32:27-30, 2016.
[6] Gausemeier, J.; Klocke, F.; Dülme, C.; Eckelt, D.; Kabasci, D.; Kohlhuber,
M.; Schön, N.; Schröder, S.: Industrie 4.0 – Internationaler Benchmark,
Zukunftsoptionen und Handlungsempfehlungen für die Produktions-
forschung. Paderborn und Aachen: Heinz Nixdorf Institut, Universität
Paderborn und Werkzeugmaschinenlabor WZL der Rheinisch-
Westfälischen Technischen Hochschule Aachen.
[7] Luet h K, Patsioura, C, Williams, Z, Kermani, Z. Industrial Analytics
2016/2017 – The current state of data analytics usage in industrial
companies. IoT Analytics, Hamburg, 2016.
[8] Dienst S. Produktverbesserung durch die Anwendung eines Feedback
Assistenz Systems. Dissertation, Naturwissenschaftlich-Technische
Fakultät, Universität Siegen, Universitätsbibliothek Universität Siegen,
Siegen, 2014
[9] Gausemeier J, Czaja A, Dülme C. Innovationspotentiale auf dem Weg zu
Industrie 4.0. In: Wissenschafts- und Industrieforum Intelligente
Technische Systeme 2015, Heinz Nixdorf Institut, (2015).
[10] Albers A. A., Bernijazov R., Kaiser L.,Dumitrescu R. Internet of Things
Canvas for Ideation in Model-Based Product Generation Planning. In 2018
13th Annual Conference on System of Systems Engineering (SoSE) (pp.
327-334). IEEE
[11] Abramovici M, Schulte, S. Best Practice Strategien für die Einführung von
Product Lifecycle Management, in Industrie Management 2/2005, GITO-
Verlag, 2005.
[12] Feldhusen J., Grote, K.H. Pahl/Beitz Konstruktionslehre: Methoden und
Anwendung erfolgreicher Produktentwicklung. Berlin Heidelberg New
York: Springer-Verlag, 2013.
[13] Gausemeier J, Kokoschka M. Schutzmaßnahmen vor Produktpiraterie. In:
Gausemeier,J.;Glatz,R.;Lindemann,U.: Präventiver Produktschutz – Leit-
fäden und Anwendungsbeispiele, Carl Hanser Verlag, München, 2012
[14] Kühn, A., Joppen, R., Reinhart, F., Röltgen, D., von Enzberg, S., &
Dumitrescu, R. (2018). Analytics Canvas–A Framework for the Design
and Specification of Data Analytics Projects. Procedia CIRP, 70, 162-
167.
[15] Martinez M, Di Nauta P, Sarno D. Real and apparent changes of
organizational processes in the era of big data analytics, Studi
organizzativi, 2017.
[16] Porter M. E., Heppelmann J. E. How smart, connected products are
transforming companies. Harvard Business Review, 93(10), 96-114,
2015.
[17] Butschan J. et al. Kompetenzaufbau zur Umsetzung von Industrie 4.0 in
der Produktion. Betriebswirtschaftliche Aspekte von Industrie 4.0.
Springer Gabler, Wiesbaden, 2017. 75-110.
[18] Kassner L, Gröger C, Mitschang B, Westkämper E. Product life cycle
analytics–next generation data analytics on structured and unstructured
data. Procedia CIRP, 33, 35-40, 2015.
[19] Demminger C, Mozgova I, Quirico M, Uhlich F, Denkena B, Lachmayer
R, Nyhuis P. The Concept of Technical Inheritance in Operation: Analysis
of the Information Flow in the Life Cycle of Smart Products. Procedia
Technology, 26, 79-88, 2016.
[20] Bentlage A, Zenker M. Data Mining of lifecycle information
https://www.ast.uni-hannover.de/fileadmin/ast/16.10.2014_Data-
Mining_of_lifecycle_information_Bentlage.pdf
[21] Abramovici M, Lindner A, Walde F, Fathi, Madjid, Dienst S. Decision
support for improving the design of hydraulic systems by leading
feedback into product development. ICED 11 - 18th International
Conference on Engineering Design - Impacting Society Through
Engineering Design. 9. 1-10, 2011