=Paper=
{{Paper
|id=Vol-2900/WS9Paper1
|storemode=property
|title=Digitization Toolkit for SMEs
|pdfUrl=https://ceur-ws.org/Vol-2900/WS9Paper1.pdf
|volume=Vol-2900
|authors=Patrick Gering
|dblpUrl=https://dblp.org/rec/conf/iesa/Gering20
}}
==Digitization Toolkit for SMEs==
https://ceur-ws.org/Vol-2900/WS9Paper1.pdf
Digitalisation toolkit for SMEs
Gering, Patricka
a
Fraunhofer Institute for Production Systems and Design Technology, Pascalstraße 8-9, 10587 Berlin, Germany
Abstract
The digitalisation toolkit combines technological and methodological competencies tailored
for small and medium-sized enterprises. In a series of workshops, participants learn the
fundamentals of process optimisation and digitalisation through a learning game and
prototypical digitalisation of own use cases. Proving the potential for improvement of
optimisation measures with concrete measured values is particularly difficult in flexible
production environments without it support. This paper presents a technical solution that
provides the necessary data for a performance analysis with process mining methods without
technical preconditions.
Keywords
Industrie 4.0, Digital Transformation, Process Mining, Learning Games
1. Introduction
High investment costs, lack of technology standards and a lack of know-how and resources are
obstacles towards a digital transformation SME in Germany have to face. [1]
Increased amount of data in production systems offer new possibilities for process optimization,
though. Real-time acquisition, analysis and evaluation of process data offer the possibility of efficient
control of production processes based on networked data sources [2, 3] and might justify the investment
costs in new technologies. Based on this basic consideration, “Industry 4.0 out of the box” has emerged,
a mobile solution for the fast digitalization of processes with the help of sensor technology [4].
However, the digital transformation cannot be managed with technologies alone. Know-how and a
clear value proposition are equally important. The digitalisation toolkit combines technological and
methodological competencies tailored for small and medium-sized enterprises. In a series of workshops,
participants learn first the fundamentals of process optimisation and digitalisation through a learning
game. Furthermore, based on company-internal use cases of the participants the benefits of digital
transformation are demonstrated through a prototypical digitalisation of processes using sensor
technology and smart services during a period of several weeks. This paper focuses on a technical
solution for the necessary data gathering to evaluate the process improvements, which is part of the
digitalisation toolkit. The learning game is used as an example.
2. Learning Game
The learning game depicts a traditional drilling machine factory. At different workstations drilling
machines are manufactured in different variations. Figure 1 illustrates the manufacturing process of
drilling machines the game covers, the process inputs and outputs as well as the work stations the
processes are fulfilled at. The process begins with prefabrication of parts till the provisioning of the
drilling machines to the customer. The participants take on the role of a worker and are distributed
among the works stations.
Proceedings of the Workshops of I-ESA 2020, 17-11-2020, Tarbes, France
EMAIL: patrick.gering@ipk.fraunhofer.de
©️ 2020 Copyright for this paper by its authors.
Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).
CEUR Workshop Proceedings (CEUR-WS.org)
Figure 1: Example figure
Starting with a chaotic and complex production, the drill factory is transferred to a lean production
system together with the participants. In the course of the training, they learn methods and tools to
independently optimise the factory. The game already starts with a running factory. An error-prone
production control and a function-oriented layout quickly lead to a chaotic state. Key figures such as
delivery reliability or errors illustrate the effects on the profitability of the company. In a second round
of the game, the participants themselves propose and implement optimisation measures. Usually,
however, these measures do not lead to the desired improvements, sometimes the result even
deteriorates. Using an integrated business process model, the trainer finally illustrates the complex
control and order flow and the problematic dependencies in the existing layout. Now an optimised
model is developed step by step together with the participants and its implementation is realised.
3. Technical Solution
In order to make the success or failure of the optimization measures clear to the participants, it is
essential to underpin the individual game levels with key figures. Thus, the learning game differs from
typical optimization and digitalisation projects in the industry only in its scope. Here, too, it is necessary
to back up measures with facts to enable an evaluation. With the goal to gather process oriented-
knowledge, a lot of research in the last decade focused on the application of process mining techniques
in real business process analysis. By analysing event logs of information systems, such as
manufacturing execution systems (MES), enterprise resource planning (ERP) tools or workflow
management systems (WFMS), process mining can deliver useful insight into the actual behaviour of
processes [5]. The three major analysis perspectives in process mining are process discovery,
performance analysis and conformance checking [6]. However, widespread IT systems are not suitable
for the constantly changing processes from the learning game, which reflects the challenges producing
companies are facing, as order specific procedures become increasingly necessary [7]. This paper
proposes a technical solution as part of the Digitalisation Toolbox to create event logs in flexible
production processes to enable process mining techniques in order to analyse process performance.
The process in Figure 1 is modelled using the Integrated Enterprise Modelling [8]. During the
different levels of the learning game, the participants adapt the model according to the proposed
optimizations. Due to the possibilities of the IEM, the model includes not only the process flow itself,
the model can be enriched by information, such as the work station, the manufacturing step is supposed
to take place, the role who is responsible, the used machine or tools, etc.. Not only that, the information
flow as well as the input and output relationships can be mapped. Usually configured elaborately in
ERP or MES systems, the proposed technical solution reuses these already available information by
adding real time events. The core of the solution is the process assistant (PA). The PA is a tool that
provides the internal processes, dependencies, responsibilities based on IEM models of process-oriented
companies to various user groups in a format that meets their needs [9]. All the information an IEM
model consists of is available and will be used to create the logs in order to enable process mining
analysis perspectives. Therefore, the PA was extended by a new Application Programming Interface
(API). With this API, it is possible to record events that occurred during a process (see Figure 2). REST,
which stands for REpresentational State Transfer, is an architectural style for distributed systems, was
first presented by Roy Fielding [10]. In this concept, it is used as an API for AutoID Systems, such as
barcode or qr code readers.
Figure 2: Example figure
This new API for the PA will be used, to track orders along the manufacturing process. It expects
two values, a ReaderID, to identify the AutoID device the API is triggered by, as well as the OrderID,
which identifies the product/product part to manufacture. The PA creates logs for every event triggered
by the API and stores them in a MongoDB [11] database. A first event of a ReaderID and OrderID
combination stores the OrderID along with timestamp in the database. A second event of that same
ReaderID and OrderID combination is handled as a stop event, in order to calculate the process duration
as the difference between the two events. Intermediate scans are treated as process interruptions. By
facilitating the information of the IEM model, these events will be enriched by their context. As shown
in Figure 2, every action in this example has connected resources, such as a work station, the action is
supposed to run on as well as a specific AutoID system, which has a unique ReaderID as its metadata.
That ReaderID connects the information within the model with the events triggered by the API. Using
that information, it is now possible to link each unique event identified by OrderID and ReaderID with
its corresponding action and its connected IEM objects. Through barcode readers at each workstation
of the learning game it is now possible to track detailed process times of each step for each job. This
information is made available via a dashboard and now enables statements to be made about the benefits
of optimization measures.
4. Conclusion
Digital transformation can only be motivated by concrete benefits. Performance analyses with
process mining methods can prove concrete benefits by before/after comparisons. The necessary data
is often not available, especially in flexible manufacturing environments such as those found in SMEs.
Proven in an educational game, the presented solution offers an easy way to collect the necessary data
without technical preconditions.
5. Acknowledgments
This research and development project is funded by the european regional development fund. The
author is responsible for the contents of this publication.
6. References
[1] Pierre Audoin Consultants, Ed., “IT-Innovation-Readiness-Index,” Munich, Germany, 2015.
[2] R. Kiem, Qualität 4.0: QM, MES und CAQ in digitalen Geschäftsprozessen der Industrie 4.0.
München: Hanser, 2016.
[3] A.-W. Scheer, Industrie 4.0 - Wie sehen Produktionsprozesse im Jahr 2020 aus?, 2013.
[4] P. Gering and P. Drange, “Industry 4.0 Out of the Box,” in Enterprise Interoperability VIII:
Smart Services and Business Impact of Enterprise Interoperability, Cham, 2019, pp. 45–53.
[5] SookYoung Son, Bernardo Nurgroho Yahya, Minseok Song, Sangsu Choi, and Nakyun Sung,
“Process Mining for Manufacturing Process Analysis: A case Study,” in 2014. [Online].
Available: https://www.researchgate.net/publication/271910986_Process_Mining_for_
Manufacturing_Process_Analysis_A_case_Study
[6] M. Park, M. Song, T. H. Baek, S. Son, S. J. Ha, and S. W. Cho, “Workload and Delay Analysis
in Manufacturing Process Using Process Mining,” in Asia Pacific Business Process
Management: Third Asia Pacific Conference, AP-BPM 2015, Busan, South Korea, June 24-26,
2015, Proceedings, Cham, 2015, pp. 138–151.
[7] M. Mörtl, Ressourcenplanung in der variantenreichen Fertigung. Zugl.: München, Techn.
Univ., Diss., 2008. München: Utz, 2008.
[8] G. Spur, K. Mertins, and R. Jochem, Integrierte Unternehmensmodellierung, 1st ed. Berlin:
Beuth, 1993.
[9] Prozessasssistent. [Online]. Available: https://prozessassistent.ipk.fraunhofer.de/ (accessed:
Nov. 9 2020).
[10] R. T. Fielding, Architectural styles and the design of network-based software architectures,
2000.
[11] MongoDB, The most popular database for modern apps. [Online]. Available: https://
www.mongodb.com/ (accessed: Nov. 10 2020).