Towards Simple and Robust Automation of
Sustainable Supply Chain Communication
Gregor Grambow, Nicolas Mundbrod, Jens Kolb and Manfred Reichert
Institute of Databases and Information Systems
Ulm University, Germany
{gregor.grambow,nicolas.mundbrod,jens.kolb,manfred.reichert}@uni-ulm.de
http://www.uni-ulm.de/dbis
Abstract. In supply chains, companies are forced to produce in a
more sustainable way. Amongst others, this involves the reporting of
various sustainability indicators to legal entities. To gather necessary
data from their suppliers, companies must employ long-running, cross-
organizational data collection processes. Being dependent on many com-
panies and contextual factors such processes imply great variability, are
manually managed, and tend to be tedious and error prone. This paper
proposes an approach for automated contextual process configuration
that also supports humans with lightweight, correct modeling and exe-
cution of configurable processes.
Key words: Process Configuration, Business Process Variability, Data Col-
lection, Sustainability, Supply Chain
1 Introduction
In many domains of modern industry, sustainability has become increasingly
important. Companies are forced by legal regulations and customer demands to
a more sustainable production. This involves the reporting of various sustain-
ability indicators, such as the amount of lead in a certain product. However, in
a supply chain, this becomes a challenge as most products are the result of the
collaboration of multiple companies. Consequently, reporting companies must
employ long-running, cross-organizational data collection processes to obtain re-
quired indicators from their suppliers as well. Furthermore, as sustainability is
an emerging topic, most companies do not have a standardized approach to such
reporting. Thus, the latter still is tedious and error prone.
In the SustainHub1project, we have identified a set of core challenges for
such data collection processes by investigating use cases from companies in the
electronics and automotive domain. The first challenge concerns the selection
of the parties involved. As reporting is manually managed, it is often not clear,
which suppliers or service providers need to be involved for retrieving a certain
indicator. Furthermore, as different companies employ different approaches, data
1SustainHub (Project No.283130) is a collaborative project within the 7th Framework
Programme of the European Commission (Topic ENV.2011.3.1.9-1, Eco-innovation).
formats, and reporting tools, getting access to the data is often problematic. Be-
ing dependent on a myriad of contextual influences, each data collection process
is almost unique. Thus, usually, many variants of the same request exist usually
that cannot be reused. Moreover, the meta data, on which these variants depend,
is mostly managed implicitly and thus obscure (see [1] for more information on
these challenges).
2 A Process-based Approach for Data Collection
In the SustainHub project, we are developing an approach to support such data
collection in a semi-automated fashion. As basis for this approach, we realize
the data collection processes through explicitly modeled processes enacted in
a PAIS (Process-Aware Information System). Thus, a set of issues can already
be dealt with as the processes are now explicit and repeatable, and the PAIS
automatically manages the different manual and automatic activities involved.
However, still many open issues remain. Therefore, we build a system comprising
additional components as illustrated in the upper part of Fig. 1.
The central component of our approach is a process configuration component
that extends pre-defined base processes with pre-defined process fragments. To
enable this in alignment with the context and meta data of the respective sit-
uation, we add a context mapping component as well as a comprehensive data
model. Thus, it becomes possible to map various contextual factors into the
SustainHub system and use them as parameters for process configuration to au-
tomatically generate process instances matching the properties of the respective
situation.
However, to enable such a flexible automated approach, users must be able
to pre-model many aspects involved in process configuration as illustrated in
the lower part of Fig. 1. This comprises concepts for the context mapping: we
apply context factors to capture contextual data. In turn, these can be mapped
via context rules to a stable set of process parameters used as input for process
configuration. The process configuration component employs bases processes and
process fragments. The former are used as basis for a data collection process,
while the latter are applied to extend that process situationally (see [2] for more
details).
All entities based on these concepts have to be modeled correctly by users.
To support this, we have introduced comprehensive correctness checks and put
focus on keeping our approach as simple as possible. The first building block to
achieve this is keeping the modeling of both base processes and process fragments
directly in the PAIS. So the modeling is understandable and the processes can
be checked by the PAIS automatically. In addition, users only have to model
entities based on the three introduced context mapping concepts a small number
of other entities for process configuration that govern where exactly to insert the
fragments into the base processes. For both parts, we have also added correctness
checks to ensure that only correct models come to execution.
Context
Mapping
Process
Configuration
Configured Process Instance
Base Process
Process
Fragments
Data Model
Contextual
Influences
Product
Customer
Customer
Relationship
SustainHub
Users and
other
Systems
Correctness
Checks
Correctness
Checks
Correctness
Checks
Correctness
Checks
P2=
1
P1=
1
P1=
2
CF2 P2=2
CF3 P2=1
CF1 P1=1
Context
Factors Context
Rules
Process
Parameters
CF3
CF2
CF1
CF1: Contact Person X; CF2: Tool Connector Y; CF3: Certification missing
P1=1: Manual Data Collection; P1=2: Automatic Data Collection; P2=1: Validation needed
Configure Data
Collection Aggregate
Data
Deliver
Data
Inform
Requester
Inform
Person Collect Data
Manually
Collect Data
Automatically Request
Validation
Convert
Data
Validate
Data
Configure Data
Collection Aggregate
Data
Deliver
Data
Inform
Requester
Base Process
Configured Process Instance
Collect Data
Automatically
Process Fragment 1
Convert
Data
Process Fragment 4
Validate
Data
Process Fragment 3
Request
Validation
Inform
Person Collect Data
Manually
Process Fragment 2
Application Example SustainHub Approach
Fig. 1: Configurable Data Collection Approach in SustainHub
The lower part of Fig. 1 shows a simplified example of such a configuration
use case from the industry: An automotive company wants to collect sustain-
ability data relating to the quantity of lead contained in a specific product. This
concerns two of the companies suppliers that deliver parts of that product. One
is a bigger company with a dedicated system for managing sustainability data in
place. The other is a smaller company with no system and no dedicated respon-
sible for sustainability. For the smaller company, a service provider is needed
that will validate the manually collected data ensuring its compliance with legal
regulations. The system of the bigger company has its own data format that has
to be explicitly converted in order to be usable. As depicted in Fig. 1, we apply
three context factors to model that situation. These are mapped to values for
two process parameters leading to the selection of four process fragments. Two
of these are used for automatic and manual data collection while the two oth-
ers enable external validation and data conversion. All fragments are integrated
automatically into the base process to obtain a configured process instance. The
latter first executes the data collection activities for both companies in parallel.
After that, again in parallel, the data of one company is converted, while the
data of the other company is validated externally.
3 Related Work
Various approaches for process variability exist. One aspect is the extension of
process modeling languages to enable configurable process models [3]. Other ap-
proaches like [4] target meta-modeling of various different aspects for process
variability. Process fragments and their modeling, in particular, have been the
focus of research, as in [5]. Finally, the automated composition of executable
processes from pre-defined process fragments has also been shown in approaches
like [6]. These approaches cover many areas of the process variability topic com-
prehensively. However, none of them has taken the context and users as our
approach into account. The latter not only enables automated contextual pro-
cess composition but also easy and error free modeling.
4 Conclusion
In this paper, we have sketched an approach for automated contextual process
configuration. The latter not only enables the automatic processing of context
data and the automatic composition of executable processes, it also puts focus
on the users interacting with our system. Thus, we have created a lightweight
way of modeling all necessary concepts and integrated a comprehensive set of
correctness checks to support them. We have further applied this approach for
complex data collection processes to supply chain communication in two do-
mains. As part of future work we plan to extend our approach with runtime
variability features and applying it in a comprehensive industrial evaluation.
Acknowledgement
The project SustainHub (Project No.283130) is sponsored by the EU in the 7th
Framework Programme of the European Commission (Topic ENV.2011.3.1.9-1,
Eco-innovation).
References
1. Grambow, G., Mundbrod, N., Steller, V., Reichert, M.: Challenges of Applying
Adaptive Processes to Enable Variability in Sustainability Data Collection. In:
SIMPDA 2013. (2013) 74–88
2. Grambow, G., Mundbrod, N., Steller, V., Reichert, M.: Towards Configurable Data
Collection for Sustainable Supply Chain Communication. In: CAiSE’14 Forum.
CEUR Workshop Proceedings, CEUR-WS.org (June 2014)
3. Ayora, C., Torres, V., Weber, B., Reichert, M., Pelechano, V.: Vivace: A Framework
for the Systematic Evaluation of Variability Support in Process-Aware Information
Systems. Information and Software Technology (May 2014)
4. Saidani, O., Nurcan, S.: Business Process Modeling: A Multi-perspective Approach
Integrating Variability. In: BPMDS 2014. (2014) 169–183
5. Eberle, H., Leymann, F., Schleicher, D., Schumm, D., Unger, T.: Process Fragment
Composition Operations. In: Proceedings of APSCC 2010, IEEE Xplore (December
2010) 1–7
6. Murguzur, A., De Carlos, X., Trujillo, S., Sagardui, G.: Dynamic Composition of
Pervasive Process Fragments. In: CAiSE 2014. (2014) 241–255
