Modeling of Intralogistic Processes for the Implementation of

Warehouse Management Systems

Markus Rabe and Felix Stadler

Department IT in Production and Logistics, TU Dortmund University, Leonhard-Euler-Str. 5, 44227 Dortmund, Germany

Keywords: Modeling Method, Process Modeling, Intralogistics, Warehouse Management Systems.

Abstract: Due to the growing complexity of intralogistics systems, the use of warehouse management systems (WMS)

is becoming increasingly attractive for companies. As an often business-critical management system of

internal material flows, however, their implementation or change is complex and carries risks. Especially the

insufficient knowledge of companies about their own processes leads to a high capacity and cost burden due

to the time-consuming involvement of their own experts and, often, also contracted WMS consultants. In this

context, models and modeling methods are gaining additional importance. But, particularly in intralogistics,

with its special demands and characteristics, there is a lack of methodological support for mapping and

transferring process knowledge appropriate for the WMS implementation. The consequences, besides a low

level of acceptance among the affected employees, are project aborts and production downtimes. This paper

discusses experiences from industrial practice during the implementation of WMS and takes the position that

a supporting method is urgently required in this context. Therefore, we propose the development of a modeling

language for mapping intralogistic processes in line with the requirements for the implementation of WMS

as well as procedural method components that support the generation and transmission of the process

knowledge.

1 INTRODUCTION

Intralogistic systems are confronted with increasing

complexity (Motschenbacher & Felch, 2020). To

master this complexity, process standardization,

automation, and the use of data processing systems

are gaining additional importance. In intralogistic

systems, warehouse management systems (WMS)

play an important role by controlling, monitoring, and

optimizing the internal material flow and storage

systems for the economic operation of a company (ten

Hompel & Schmidt, 2010).

As the complete flow of materials of a company

is influenced by WMS, the implementation or change

of such a system is often complex and fraught with

risk. A main problem, according to the Fraunhofer

IML study from 2018, is the insufficient knowledge

of the companies about their own intralogistic

processes and its deficient transfer between the

involved parties (Fraunhofer IML, 2018). This

market study for WMS shows that with 76% the lack

of process knowledge on the customer side and with

40% the miscommunication of the requirements are

main reasons for in-deficit WMS implementations

(Fraunhofer IML, 2018).

Thus, in many cases, the process knowledge is

developed for the first time during the

implementation of a new system, leading to the

expensive involvement of both the company experts

and the external WMS consultants, who are often

hired to compensate the lack of WMS knowledge.

This leads to a high impact in terms of capacity and

cost on the company side, which puts a strain on the

success of the project. Amplified by inadequate and

misleading communication between the stakeholders,

this can result, according to Hartel (2019), not only in

a lack of acceptance of the WMS among the

employees, but also in project aborts and production

downtimes.

To prevent this, the use of models and modeling

methods can be a solution approach by supporting the

development, documentation, and communication of

process knowledge (Becker et al., 2017). For

intralogistic processes and the implementation of

WMS, however, a deficit in the methodological

support was evident while accompanying various

WMS implementations and can be considered

contributory to the results of the Fraunhofer study.

This deficiency led to serious problems during and

after the implementation of the WMS.

Rabe, M. and Stadler, F.

Modeling of Intralogistic Processes for the Implementation of Warehouse Management Systems.

DOI: 10.5220/0010784300003119

In Proceedings of the 10th International Conference on Model-Driven Engineering and Software Development (MODELSWARD 2022), pages 167-173

ISBN: 978-989-758-550-0; ISSN: 2184-4348

167

In order to solve this, in this paper examples and

essential requirements for the modeling of

intralogistic processes regarding the implementation

of WMS are discussed and characteristics of a

prospective supporting method are proposed.

Thus, the remainder of the paper is organized as

follows: Section 2 begins by introducing

characteristics of a WMS implementation. In

particular, the challenges that arise and the interaction

of the parties involved are discussed, here, in the

context of modeling. Following on from this, the

second part of the section gives indications for the

weak methodological support and derives implications

from the experiences of the WMS implementations

that were accompanied. Based on this, the third section

explains the position on how WMS implementations

could benefit from methodological support and

outlines two main aspects of this. In the last section,

conclusions and a possible research plan are

established to follow up and evaluate the position. The

considerations made are, thereby, illustrated by

impulses from industrial applications.

2 EXPERIENCES ON

IMPLEMENTING WMS

The implementation of a WMS as a far-reaching and

business-critical management system is afflicted with

risk and influenced by various framework conditions

(Fraunhofer IML, 2018). In addition to the

complexity of the underlying intralogistic processes

as a major expense driver, the implementation of a

WMS is also influenced by the type of software

migration. Depending on the objectives of the WMS

implementation (e.g., regarding the degree of

individualization), the planned effort, and the

possibility of adapting the existing processes, there

are several types of migration to choose from (Coyle

et al., 2017). The most common form for WMS

implementations is customizing a standard WMS

software system (2018: 77 %) (Fraunhofer IML,

2018). Here, the standard software is adapted to the

scope of functions and structures required by the

WMS customer within the framework of predefined

configuration and parameterization options (Mertens

et al., 2017). For this purpose, the customization

process of the software requires a comparison of the

customer's processes with the functional scope of the

standard software. (Hesseler & Görtz, 2014). To do

so, the customizing process has to be accompanied by

several coordinating, iterative, and creative

conceptual activities performed by the persons

involved that aim at converging the customer and

software processes. Thus, the course of these

activities significantly determines the scope, time,

and cost of the implementation of a WMS (ten

Hompel & Schmidt, 2010).

In this alignment process with its diverse

transformation tasks, various barriers can arise and

lead to problems in the customization of standard

software. In WMS implementation, a lack of process

knowledge and its insufficient transfer is a main

problem.

This leads to a resulting need for subsequent

development and transfer of process knowledge by

the involved persons, which ties up both the cost-

intensive WMS experts and the experts on the

company side (Hartel, 2019). Besides the ensuing

frustration of the participants during and after the

implementation of a WMS due to this, as well as the

time-consuming and cost-intensive rework, this can

jeopardize the intended operation of the WMS.

These deficits were also reflected during the

implementation of a WMS for a medium-sized

machine manufacturing company, which was

characterized by misunderstandings between the WMS

consultants and the company experts resulting in

frustration on both sides. There, for almost half a year,

the experts from both domains discussed and debated

without being able to agree on a common picture.

Based on the insufficient documentation, the deficient

transfer of process knowledge and the lack of experi-

ence of the WMS consultants in the specific industry

sector, process aspects were unnecessarily discussed

repeatedly, and specifications were regularly revised.

Subsequent investigations revealed that this was

caused by several reasons. One aspect, for example,

was that the process knowledge was spread across

several departments and among several experts and

not properly consolidated. This led to the WMS

experts sporadically obtaining the relevant

information from the respective company experts and

formulating their own WMS concept. This concept,

however, was not comprehensible for the company

experts due to the technical terms and, in their eyes,

inappropriate visualization (as an EPC model). But,

intimidated by the complexity of the models, the

company experts endorsed the proposals of the WMS

consultants with little reflection. The results were

undiscovered uncertainties on both sides. The

situation escalated when the company experts noticed

in the late tests that business-critical side processes

were unknown to the WMS experts.

Besides a general culture problem in the project,

a structured overview and appropriate documentation

of the processes in a comfortable language for the

MODELSWARD 2022 - 10th International Conference on Model-Driven Engineering and Software Development

168

company experts was lacking, as was an organized

transfer of knowledge. The consequences for the

project were significantly higher costs and runtime

due to the necessary rework.

A similar situation emerged for another WMS

implementation that was accompanied. Here, the

unstructured and deficient coordination between the

WMS consultants and the project team almost led to

a duplication of the project duration compared to the

original plan. This was causally induced, for example,

by the fact that the same process sections were

repeatedly discussed in isolation over a period of

months, instead of going through the process in a

structured manner, for example top-down and line

forward. As there was also no commonly accepted

and transparent documentation of the processes, a

consistent overall picture among the participants was

lacking. For the company experts, there was no

assurance that the WMS experts had understood and

paid attention to their concerns and process demands.

Due to the resulting perceived cost pressure on

the management, the implementation was carried out

even though concerns of the company experts

remained. The result was a production shutdown of

nearly two weeks and an adjustment of the system for

four months during ongoing operation with reduced

production output.

Solution approaches to avoid this can be the

enabling of preparatory work on the company side

and, above all, methodological support of the project

participants in the generation, documentation, and

transfer of process knowledge for a structured

collaboration of the domains involved (Groß and

Pfenning, 2017).

Here, models and modeling methods are gaining

importance, which can not only depict complex

problems, but also determine their solutions and

create a uniform basis for communication

(Haberfellner et al., 2019). However, if no suitable

method is available or known, companies can only

attempt to carry out the modeling based on individual

experience or hire cost-intensive external personnel.

This can have serious consequences for the quality of

WMS implementations and, thus, for the productivity

of the company as shown by the accompanied

implementations. Particularly in intralogistics, where

specialists and managers often have little or no

academic background, there is a lack of

methodological support for appropriate modeling and

transfer of process knowledge to implement WMS.

This is reflected by the fact, that hardly any of the

managers in the implementations we accompanied

had applicable experience with modeling languages,

and it was visibly difficult for the experts to identify

and prove their documented statements. The

modeling languages used (such as BPMN 2.0 or EPC)

were often perceived as unnecessarily complicated

with too many elements on one hand and on the other

hand not designed to map the relevant data of

intralogistic processes (such as the different

conveyors, packaging, and storage equipment

relevant for WMS implementation). Also, a lack of

orientation in the process model was complained due

to the missing intralogistic structural elements in the

models (such as areas), which they were used to.

Other languages that seemed more accessible to them

(such as value stream design), on the other hand, were

too unspecific for the WMS experts.

This became also apparent in almost all WMS

implementations we accompanied, where, as a likely

consequence of this deficit, no, only rudimentary or

severely outdated process documentation existed

prior to the implementation. Since no standard was

accepted by the intralogistic personnel, the few

models available were often in different modeling

languages and different information levels were used.

The more standardized modeling languages such as

BPMN 2.0 or EPC were used by, if at all, those with

an IT background, but even there they were reduced

to fewer model elements due to the deemed excessive

complexity of the languages. However, intralogistics

personnel often considered these models to be

unsuitable and were unable to accept them, as they

failed to depict essential process elements from their

perspective. Figure 1 illustrates this, showing a small

section of an intralogistic process modeled with EPC.

Figure 1: Intralogistic process visualized with EPC.

Receipt of

the transport

order

Drives to the

pick up

location

Transport

goods

identified

Scanning the

barcode

Forklift driver

ID of the

pallet

Forklift 001

Area 01

Scanner 001

Area 02

ID of

Pick up

location

Pallet 001

ID in System

Modeling of Intralogistic Processes for the Implementation of Warehouse Management Systems

169

The process has been extended by operating

resources, which are shaded in gray. The logistic

areas, which are of particular importance for the

intralogistic personnel, were exposed on the right.

Relevant for the intralogistic stakeholders are mainly

the activities (green) and the resources used (gray and

yellow).

Overall, in all cases observed and in line with the

results of the Fraunhofer study (2018), the lack of

methodological support for modeling intralogistic

processes seems to lead to insufficient creation,

documentation, and thus communication of the

process knowledge between the participants.

3 METHODOLOGICAL

SUPPORT AS A SOLUTION

A promising approach to improve this situation can

be to provide methodological support for both

problem fields. On the one hand, the appropriate

documentation of the process knowledge should be

supported, and on the other hand, there should be a

determined procedure for the structured elaboration,

communication, and transformation of process

knowledge among the experts involved.

A wholesome methodological concept that fulfills

these requirements in the context of modeling

intralogistic processes for the implementation of

WMS could not be identified. Therefore, the position

is taken that a new or significantly evolved method is

needed to support companies during their

implementation of a WMS regarding the modeling of

the relevant processes. This method should comprise

a language aspect for the mapping of the process

knowledge according to the requirements and a

process aspect for the structured elaboration and

transfer of the process knowledge.

3.1 Language Aspect

The modeling language should be appropriate to

depict the characteristics of intralogistic processes

pertinent to the implementation of WMS. Logistical

areas, for example, form the basis of the WMS stock

structure and can provide intralogistics personnel

with the required orientation in the model (Sachan &

Jain, 2020). Thus, these process elements should have

a higher-ranked status in intralogistic models and

could be used, for example, as structuring elements.

Figure 2 shows a proposal in which the process as a

sequence of activities is visually structured by areas.

Figure 2: Process structuring using logistical areas.

As most existing modeling approaches cannot

map these and other process characteristics (such as a

large number of operating resources) appropriately,

different method concepts are adapted and different

process content is depicted, depending on the

individual experience of the modeler. According to

Becker et al. (2012) failures in this process lead to

inconsistencies, a lack of information or over-

information, and higher modeling efforts, as well as a

low acceptance of the models.

On the other hand, the language should also

support a flexible and low-effort adaptation of models

during the customizing iterations. Therefore, simple

but stringent syntactic rules should be supported by

the language. As a result, a consistent modeling

language, with a notation for the concrete syntax, a

language-based metamodel for the specification of

the abstract syntax, and semantic as well as pragmatic

descriptions should be developed.

Here, it should be examined whether an existing

language can be sufficiently adapted or whether a

new language needs to be developed. However, the

language should be less formalized and aim to make

the company experts feel comfortable using it. In the

first step, therefore, the language does not need to be

automatable or machine-readable, but simply

accepted by the company experts and understood by

the WMS experts.

In addition, principles from Gestalt psychology

(Fitzek, 2014) should be considered in the design of

the notation. These ensure the purpose-oriented

perception of the models as well as the low-

complexity design of the models and model elements

by means of Gestalt laws (Desolneux et al., 2008).

However, in the Gestalt-psychological examination

of existing methodological modeling concepts, it was

found that these aspects are often neglected. Yet, they

contribute to its perception and interpretation,

especially in the case of more comprehensive models.

It became apparent in the WMS implementations

we accompanied that a modeling language adapted to

the specific requirements of intralogistic processes

and to the needs of the employees involved

encountered great acceptance. By also applying the

principles of Gestalt psychology to the design of the

notation, the perceived complexity of the models

Activity Activity Activity Activity Activity

Activity

Area 2Area 1 Area 3

MODELSWARD 2022 - 10th International Conference on Model-Driven Engineering and Software Development

170

could be significantly reduced. An example of this is

the common visualization of a few but essential

process attributes in the main model. This simplified

the required understanding of the process models

considerably and made it possible to reduce the

number of different models. Figure 3 demonstrates

this by taking the process section shown in Figure 1

and mapping it in an alternative way, more

appropriate to the intralogistic requirements.

Figure 3: Proposal of an alternative process visualization.

3.2 Process Aspect

Furthermore, the processes for the elaboration,

documentation and transfer of the process knowledge

should be supported by the method with appropriate

procedural components. For this purpose, the

procedural support should be considered

comprehensively and on several levels. Here,

process-related components should be developed that

provide support and orientation for the persons

involved at the operational, tactical, and strategic

levels of modeling.

For example – alongside a higher-level

procedural model for describing the development,

structuring and transfer of process knowledge in

general – more specific activities for the structured

modeling of individual complex intralogistics

processes should also be determined as well as

concrete operational steps, how model elements are

set. To illustrate this, the following are some

suggested questions to be answered at each

procedural level.

At the strategic modeling level, for example, the

following questions should be answered:

- How does the process structure to be modeled

look like and how are the processes delimited from

each other (e.g., processes are structured

hierarchically according to a top-down approach and

delimited based on physical buffers or areas)?

- How can preliminary work be realized, and how

can the intralogistics personnel build up their own

understanding of the process (e.g., by two modeling

cycles: one to support domain-oriented modeling for

low-cost preparatory work on the company side and

one to support WMS-oriented modeling for adapting

the business unit models to WMS requirements with

the involvement of cost-intensive consultants)?

Figure 4 shows a possible approach of a simple phase

model, which takes this into account.

Figure 4: Proposal for a phase model.

On the tactical level, on the other hand, the

following questions are in the focus:

- How are the models constructed (e.g., models

are built in three stages: Stage One is the mapping of

the basic process structure, in Stage Two, the

intralogistics resources are added and in Stage Three

other process attributes, such as process times, are

supplemented)?

- By which techniques are the model data

collected and how are models investigated, validated,

and verified (e.g., via observation and additional

questioning of the experts)?

Finally, at the operational level, the modeling

rules are defined, and practical questions are

answered, for example, about the concrete procedure

of modeling workshops:

- In which order should the model elements be

placed (e.g., first process-external elements as

starting points, then the following process element,

then the connector between both)?

- How are modeling workshops conducted? (e.g.,

first structured walk- and talk-through of the entire

process, then talk-through line-forward process

Drive to

pick up location

Scan

pallet barcode

Forklift driver

Forklift 001

Forklift driver

Pallet 001

Scanner 001

Definition of processes

and modeling demand

Domain-oriented

modeling

Domain-oriented

evaluation

WMS-oriented

modeling

WMS-oriented

modeling

Assessment and release

of modeling results

WMS-oriented

modeling cycle

Domain-oriented

modeling cycle

Modeling of Intralogistic Processes for the Implementation of Warehouse Management Systems

171

section by process section; notes are taken directly in

the model; moderation by domain-neutral and trained

modeler)?

These questions need to be answered in

procedural models. Here, it is important that these

process-related specifications are aligned with the

characteristics of the intralogistics modeling

language and the specifics of the WMS

implementation process. An example for this is the

differentiation in two modeling cycles. The first cycle

helps to establish a common understanding of the

processes within the company. The focus in the WMS

modeling cycle, as the second cycle, is on supporting

the customizing interactions among the participants

and in building a common understanding of what the

processes in the WMS will look like.

With these two aspects, the language aspect, and

the procedural aspect, both closely aligned and

intertwined, a methodological support for the

implementation of WMS would be given and risks

occurring there could be reduced.

4 CONCLUSION

Our research indicates that the use of appropriate

models and modeling methods for intralogistic

processes to initiate WMS has a significant impact on

the economic implementation of WMS. Therefore, a

method is required that covers both the language and

the process aspects of modeling intralogistic

processes needed for the implementation of WMS.

The overall research goal should be a methodological

support for modeling intralogistic processes

regarding the implementation of WMS.

Future work and research in this field include the

detailed specification of the requirements for a

modeling method from both a language and a process

point of view. Afterwards, a suitable development

process for the method has to be defined, which

ensures its high-quality and consistent structure. On

this basis, the development of the modeling method

can be carried out and evaluated step by step, ideally

accompanied by a practical application in the idea of

prototyping.

The concrete development of the method should

start with the definition of perspectives that need to

be taken for an appropriate modeling of the

intralogistic processes. Subsequently, the modeling

language with its three aspects the syntax, the

semantics, and the pragmatics can be built. It can be

expected that at least basic aspects of existing process

modeling languages can be adopted due to generally

valid process characteristics.

For the syntax, a language-based metamodel

should be constructed that specifies the language

elements and their relationships among each other.

Here, a small number of language elements should be

aimed for to keep the complexity of the modeling

method low for the intralogistics experts from the

outset. For this purpose, discoveries from Gestalt

psychology should be used. Here, the intralogistics

areas should take on a superordinate and process-

structuring significance.

Gestalt psychology should also be considered

when designing the representational forms of the

language elements. This should support an easily

perceivable and interpretable shape of the forms.

Overall, the focus should be on supporting the

company experts with a suitable modeling language

to be able to adequately map their process knowledge

and adapt it according to the WMS requirements.

Furthermore, procedural method components also

need to be developed, which determine the modeling

process and structure the interaction among the

involved domains. Here, the main goal should be to

support the business experts and convey their

modeled process knowledge to the WMS consultants

during the customizing process. In particular,

preparatory work by the company should be enabled

and the target-oriented exchange between the parties

simplified. Two modeling cycles should be developed

for this purpose: one to support business-unit-

oriented modeling and one to support WMS-oriented

modeling. Furthermore, the process-related

components should cover the operational and tactical

as well as the strategic modeling level to support the

participants. Next, to support the involved

stakeholders, it should be examined to what extent the

modeling can be supported by information

technology tools. For this, the modeling language

should be supported by a modeling tool or a

visualization program.

Finally, the methodological construct should be

evaluated in practice.

REFERENCES

Becker, J., Bernhold, T., Knackstedt, R. & Matzner, M.

(2017). Planung koordinierter

Wertschöpfungspartnerschaften. Springer, Berlin

Becker, J., Probandt, W. & Vering, O. (2012). Grundsätze

ordnungsmäßiger Modellierung. Springer, Berlin

Coyle, J. J., Langley, C. J., Novack, R. A. & Gibson, B. J.

(2017). Supply Chain Management. A Logistics

Perspective. Cengage Learning, Boston. 10

edition

MODELSWARD 2022 - 10th International Conference on Model-Driven Engineering and Software Development

172

Desolneux, A., Moisan, L. & Morel, J.-M. (2008). From

Gestalt Theory to Image Analysis. A Probabilistic

Approach. Springer, New York

Fitzek, H. (2014). Gestaltpsychologie kompakt.

Grundlinien einer Psychologie für die Praxis. Springer,

Wiesbaden

Fraunhofer IML. (2018). WMS Marktreport Kompakt 2018.

Trends und Entwicklungen auf dem Markt für

Warehouse Management Systeme. Fraunhofer IML,

Dortmund. 5

edition

Groß C. & Pfenning R. (2017). Professionelle

Softwareauswahl und -einführung in der Logistik.

Springer Gabler, Wiesbaden

Haberfellner R., de Weck O., Fricke E. & Vössner S.

(2019). Systems Engineering. Fundamentals and

Applications. Springer Nature Switzerland AG, Cham

Hartel, D. H. (2019). Projektmanagement in Logistik und

Supply Chain Management. Springer Gabler,

Wiesbaden. 2

edition

Hesseler, M. & Görtz, M. (2014). Basiswissen ERP-

Systeme. Auswahl, Einführung & Einsatz

betriebswirtschaftlicher Standardsoftware. W3L,

Dortmund. 3

edition

Mertens, P., Bodendorf, F., König, W., Schumann, M.,

Hess, T. & Buxmann, P. (2017). Grundzüge der

Wirtschaftsinformatik. Springer Gabler, Berlin. 12

edition

Motschenbacher, S. & Felch, V. (2020). Intralogistik 4.0:

Die unternehmensinterne Logistik im Kontext der

Digitalisierung und Industrie 4.0. In Werner, J.,

Biethahn, N., Kolke, R., Sucky, E. & W. Honekamp

(eds.). Mobility in a Globalised World 2019 (pp. 163–

186). Univ. of Bamberg Press, Bamberg

Sachan, N. & Jain, A. (2020). Warehouse Management with

SAP S/4HANA. Embedded and Decentralized EWM.

Rheinwerk Publishing, Boston. 2

edition

ten Hompel, M. & Schmidt, T. (2010). Warehouse

Management. Organisation und Steuerung von Lager-

und Kommissioniersystemen, Springer, Heidelberg. 4

edition

Modeling of Intralogistic Processes for the Implementation of Warehouse Management Systems

173