=Paper=
{{Paper
|id=Vol-2900/WS8Paper1
|storemode=property
|title=Rethinking Interoperable CPS as Interactive Behavior Designs
|pdfUrl=https://ceur-ws.org/Vol-2900/WS8Paper1.pdf
|volume=Vol-2900
|authors=Christian Stary
|dblpUrl=https://dblp.org/rec/conf/iesa/Stary20
}}
==Rethinking Interoperable CPS as Interactive Behavior Designs==
https://ceur-ws.org/Vol-2900/WS8Paper1.pdf
Rethinking Interoperable CPS as Interactive Behavior Designs
Christian Starya
a
Johannes Kepler University, Business Informatics-Communications Engineering, Altenbergerstraße 69, 4040
Linz, Austria
Abstract
This paper considers the design of interoperable CPS as socio-technical development task, as
CPS represent evolving interactive systems. Pathways from level I to 6 are provided suggesting
structuring the development of a design language for modeling and architecting interoperable
CPS as Systems-of-Systems. The pathways address various levels of complexity of
architectural designs as well as language requirements for stakeholder-centered behavior
modeling.
Keywords 1
CPS, Interoperability Engineering, socio-technical system design, System-of-Systems,
Complex Adaptive Systems, Interaction Science, behavior modeling
1. Introduction
Interoperability Engineering is increasingly getting a design task requiring interactive access to
design representation in a stakeholder-centered way. According to Gartner’s forecasts 2019, over the
next several years, the workforce’s ability to exploit business-relevant technologies such as Internet-of-
Things (IoT) systems, will decide whether many organizations can build up or keep competitive
advantage (cf. https://www.gartner.com/smarterwithgartner/top-10-technologies-driving-the-digital-
workplace). Individual activities will be bound to digital actions creating an “Internet of Behavior” in
public communities and business settings (https://www.gartner.com/smarterwithgartner/gartner-top-
strategic-predictions-for-2020-and-beyond/). Consequently, Cyber-Physical Systems (CPS) will be
affected by continuous interactive change, requiring the (dynamic) adjustment of their components
ensuring interoperability.
Since such a pragmatic view on Interoperability Engineering is of prime concern for seamless CPS
operation in volatile environments, behavior data direct activities of CPS in real time, and encourage or
discourage a particular (human) behavior. For instance, shop floor or home management systems could
adapt their features to the status of available sensor systems and the behavior of the currently present
people. The understanding of the situation and specific needs of stakeholders grounds the design of
CPS systems and handling their interoperability. Behavior specifications abstract from the origin of
actors (machines, humans) and allow subsuming components to more complex systems.
Behavior specifications result in modeling CPS and the interplay of components as an immanent
and pervasive task. Both, developers, and users adapting behavior dynamically require proper language
and tool support for design. It is expected by 2023, that ‘40% of professional workers will orchestrate
their business application experiences and capabilities like they do their music streaming services’
(https://www.gartner.com/smarterwithgartner/gartner-top-strategic-predictions-for-2020-and-
beyond/). It is not sufficient for organizations or communities of practice to work with static designs or
monolithic applications. They need flexible schemes when technology or model adoption and task
restructuring become part of their work (cf. [10]). Otherwise they will be fail to arrange, customize, and
exploit CPS technologies for interoperable service and production developments.
The paper develops pathways based on a refined system understanding to reflect the connectivity
and dynamic modularity of CPS. Recognizing the interactive and socio-technical nature of evolvement,
Proceedings of the Workshops of I-ESA 2020, 17-11-2020, Tarbes, France
EMAIL: christian.stary@jku.at (Christian Stary);
©️ 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)
the pathways differentiate architecture concerns from behavior representations, as the latter address tool
requirements for modeling interoperable CPS.
2. Evolving CPS as Interactive Systems-of-Systems
CPS-architectures are traditionally run decentralized, being linked to communication and modular
(e.g., agent-based) structures. Some parts need to co-operate through exchanging information and
adapting to environmental changes, others deliver or process data on demand. Overall, CPS allow large-
scale interconnected processes as reconfigurable networks of locally autonomous actors, including IoT-
enabled sensory system [7]. Thereby, CPS link “cyber” (virtual, computational) and “physical”
components. Their openness and socio-technical character stem from interconnecting physical, social,
and virtual worlds.
CPS are heterogeneous systems, including distributed physical devices and computational
components. As flexibility and re-configurability matter for operation due to changes in environment,
the granularity of the system, the access to components, and the way information is exchanged are
crucial design parameters. Any CPS architecture encompasses physical objects, sensors, actuators,
computing devices, controllers and communication network(s) that are represented as digital twin in a
model of the actual system. As such, it has to be capable to represent and support human-device
interaction and device-to-device communication. According to Interaction Science, the ‘interaction is
considered as the exchange of material or immaterial goods between acting parties (biological or
technical entities) embodied in a certain context. Overtime, these interactions establish behavioral and
cognitive schemas that transfer as patterns and expectations to further activities and in this way
influence the acting stakeholders. Stakeholders are identified as the persons that are involved in system-
and interaction-relevant processes, either operating, (re-)design, monitoring or controlling a system. As
they interact in a certain environment, they create specific patterns in the system representing the
environment.’ [1]
These system-wide patterns shape the behavior of each actor (humans, robots, applications, etc.,)
resulting in a Complex Adaptive Systems and thus, creating challenges of interoperability in systems
(cf. [8] for production). A System-of-System perspective on these kind of ecologies helps coping with
complexity, taking into account emergent behavior and transformations (cf. [2]). Thereby, components
(sub systems) are linked in a way their internal structure can handle their interaction to form a unified
whole (cf. [5]), albeit their often physical and functional heterogeneity. As System-of-System they are
organized hierarchically, with each component contributing to an overarching system function.
Allowing for autonomous behavior of each component constitutes a federated system architecture. In
this way, CPS can evolve from autonomous sub systems towards a network of interacting components
(cf. [9]). In the course of interoperability engineering the protocol of interacting with the network actors
is decisive for dynamic alignment to contribute to a common objective of a CPS. It needs to be
considered part of the CPS architecture and as design enabler.
3. Pathways to a Design and Architecture Language
Engineering interoperable CPS supported by a design and architecture language requires not only a
specific system perspective, as already indicated in the previous section, but development paths that
take into account CPS architecting and design as an interactive process. Although being intertwined,
they need to meet different requirements due to the interactive and socio-technical nature of evolvement
processes.
Figure 1: Pathways
Figure 1 details for each of the milestones (development level) the architecture and language
perspective. The pathways towards an interoperability design and architecture language reflect how
alignment of architecture modeling and the deployment of interoperable CPS can bring value taking
into account a system perspective on modeling that accounts for the context of CPS operation. Each
pathway enhances the design space for CPS technologies in a domain-independent way due to
encapsulating actor behavior and standardized interaction. As such the migration from existing
architectures and systems towards interoperable CPS can be facilitated.
Level 1 addresses the dynamic nature of a system when designing CPS. Stakeholders and CPS
components constitute a Complex Adaptive System. In level 2, the 2 components establishing CPS as
socio-technical system are in the focus of development. Level 3 captures the federated system aspect –
sub systems can operate on their own while being part of a larger system. Level 4 tackles the
technological CPS constituents and their behavior representation. On level 5 the business logic and
thus, user and usage-relevant behavior is addressed. Level 6 explicitly refers to interoperability issues,
namely in terms of aligning behavior designs of the sub systems of CPS, ensuring operational
connectivity.
SoS development should lead to architectures allowing dynamic changes. Situation-sensitive
behavior is a key issue in CPS engineering support. A situation is analyzed in terms of how the different
complex adaptive system parts influence and relate to each other rather than decomposing it into parts
that are studied in isolation. The resulting system behavior focuses on actors of different kinds, e.g.,
designers tackling interoperability issues on the model that needs to be propagated to operation. The
designer can be the users of the CPS at hand. As such they play a dual role and need to work together
on several levels even when following complex paths of behavior.
Designers need to be able to keeping the whole in mind when working on CPS designs [4]. Besides
the CPS components designers need to set links and interconnections which influence the behavior.
Subject-oriented design (cf. [3]) can help due to its simple interaction structure and behavior
centeredness. It allows encapsulation of behavior and addressing actor behavior through data-driven
message exchange. A language providing simple mechanisms for modeling enables stakeholders to be
engaged more effectively into component alignment as they are relieved from transformation tasks.
In case behavior models are represented in a semantically precise language they can be executed for
probing. Such a feature supports stakeholders testing pragmatic interoperability of CPS. In this way
fundamental System-of-Systems factors (cf. [6]) are addressed: (i) autonomy, where components within
the CPS can operate and function independently; (ii) belonging (integration), which implies that the
constituent CPS components can integrate for System-of-Systems capabilities; (iii)
connectivity between components and their environment; (iv) diversity addressing different
perspectives and functions; and (v) emergence (foreseen or unexpected). As the latter addresses all
types or levels of interoperability, it is key to CPS evolvement. Hence, an effective design language
needs features for mapping emerging structures to an existing model and architecture, as well as
alignment functions for interoperability engineering.
4. References
[1] Bahr, G.S., & Stary, C. (2016). What is interaction science? Revisiting the aims and scope of
JoIS. J Interact Sci 4, 2 doi:10.1186/s40166-016-0015-5
[2] Carlock, P. G., & Fenton, R. E. (2001). System of Systems (SoS) enterprise systems engineering
for information‐intensive organizations. Systems engineering, 4(4), 242-261.
[3] Fleischmann, A., Schmidt, W., & Stary, C. (Eds.). (2015). S-BPM in the Wild: Practical value
creation. Springer, Cham.
[4] Frank, M. (2012) Engineering systems thinking. Cognitive competencies of successful systems
engineers. Procedia Comput. Sci. 8, 273–278.
[5] Heitmann, F., Pahl-Wostl, C., & Engel, S. (2019). Requirements based design of environmental
System of Systems: Development and application of a nexus design
framework. Sustainability, 11(12), 3464.
[6] Jaradat, R.M.; Keating, C.B. (2014). Bradley, J.M. A histogram analysis for system of systems. Int.
J. Syst. Syst. Eng. 5, 193–227.
[7] Lee, J., Bagheri, B., Kao, H. (2105). A Cyber-Physical Systems architecture for Industry 4.0-based
manufacturing systems, Manufacturing Letters 3, pp. 18-23, Elsevier.
[8] Neubauer, M., & Stary, C. (2017). S-BPM in the Production Industry. A stakeholder approach.
Springer International Publishing.
[9] Stary, C., & Wachholder, D. (2016). System-of-Systems support—A bigraph approach to
interoperability and emergent behavior. Data & Knowledge Engineering, 105, 155-172.
[10] Wally, B., Rausch, T., Nickovic, D., Krenn, W., Kappel, G., Dustdar, S., & Grosu, R. (2019).
CPS/IoT Ecosystem: A Platform for Research and Education. In Cyber Physical Systems. Model-
Based Design: 8th International Workshop, CyPhy 2018, and 14th International Workshop, WESE
2018, p. 206. Springer