See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/315508301

Industrial Cyberphysical Systems: A Backbone of the Fourth Industrial

Revolution

Article  in  IEEE Industrial Electronics Magazine · March 2017

DOI: 10.1109/MIE.2017.2648857

CITATIONS READS
25 1,064

5 authors, including:

Armando Walter Colombo Stamatis Karnouskos

Schneider Electric, University of Applied Sciences Emden/Leer SAP Research
174 PUBLICATIONS   2,735 CITATIONS    275 PUBLICATIONS   4,666 CITATIONS   

SEE PROFILE SEE PROFILE

Okyay Kaynak
Bogazici University
359 PUBLICATIONS   9,087 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

ProSEco View project

industrial processes View project

All content following this page was uploaded by Stamatis Karnouskos on 07 November 2017.

The user has requested enhancement of the downloaded file.

IEEE IEM, VOL. 11, NO. 1, 2017 1

Industrial Cyberphysical Systems:

A Backbone of the Fourth Industrial Revolution
Armando W. Colombo, Stamatis Karnouskos, Okyay Kaynak, Yang Shi, Shen Yin

Abstract—Cyberphysical systems (CPSs) are perceived as the pace of change, mainly because of adoption of emerging
pivotal enabler for a new era of real-time Internet-based com- Internet concepts, technologies, tools, and methodologies. The
munication and collaboration among value-chain participants, rapid advances in computational power, communication, and
e.g., devices, systems, organizations, and humans. The CPS
utilization in industrial settings is expected to revolutionize the storage, coupled with the benefits of the cloud and services,
way enterprises conduct their business from a holistic viewpoint, have the potential to give rise to a new generation of service-
i.e., from shop-floor to business interactions, from suppliers oriented architecture (SOA)-based industrial systems whose
to customers, and from design to support across the whole functionalities reside on-device and in-cloud and interact
product and service lifecycle. Industrial CPS (ICPSs) blur the seamlessly [1,4]. Their realization brings new opportunities
fabric of cyber (including business) and physical worlds and
kickstart an era of system-wide collaboration and information- as well as additional challenges that need to be researched,
driven interactions among all stakeholders of the value chain. analyzed, and efficiently tackled.
Therefore, ICPSs are expected to empower the transformation Similar to the societal changes, the manufacturing domain
of industry and business at large to a digital, adaptive, networked, is also undergoing a significant transition [5,6]. On the current
and knowledge-based industry with significant long-term impact industrial shop floor, behavior and (machine) intelligence pro-
on the economy, society, environment, and citizens.
gramming are concentrated on a handful of large monolithic
computing resources, accompanied by large numbers of dumb
I. BACKGROUND devices, that are tailored and individually programmed for
each process step. With the increasing penetration of CPSs,
The increasing penetration of Information Communication however, significantly more sophisticated scenarios are real-
Technologies (ICT) in industry is transforming the industrial ized that can enable production efficiency and collaboration
environment into a multifaceted system featuring a tight with internal and external stakeholders while also adhering
combination and coordination between the computational and to the requirements for flexibility, energy efficiency, and
physical elements, including their digital (virtual) representa- operational excellence imposed by business competition. The
tion e.g., in the cloud, resulting in the formation of the so- same trend is evident also in other domains such as energy,
called Industrial Cyber-Physical Systems (ICPS) [1]. Digital- health care, manufacturing, military, transportation, consumer,
ization and interconnection of products, services, enterprises, enterprise, robotics, and smart cities. It is also apparent in
and people are expected to generate significant opportunities complex solutions where CPSs forge the backbone and are
and benefits [2], assuming the risks and challenges are prop- the enabler of connectivity and interaction between those
erly addressed. seemingly disparate sectors, like transport and energy or health
Society is inexorably shifting in form and nature, influenced and economic growth [7,8].
by many powerful factors such as the aging population, climate The aim of this article is to briefly present aspects of the
changes, uncertain economic situation, demographic transition, emerging era of ICPSs, shed some light on the high-level
globalization, digitalization, and so on. Nowhere is this more supporting programs and countrywide activities carried out,
apparent than when digital technologies help tackle many as well as present key challenges that they pose. In the effort
of the challenges that derive from those societal changes to do so and to set the context of this diverse domain, it is
such as better and more easily accessible healthcare, energy unavoidable to utilize many modern labels that understandably
savings, smart transportation, and efficient production and are not fully distinguishable in the community and do not have
services [3]. Up to now, industrial systems had long-lasting clearly defined and widely accepted boundaries. Nevertheless,
lifecycles (spanning several decades in some cases); however, the views presented here may kick-start discussions on the
in the last few years, we have witnessed an increasingly rapid different characteristics that ICPSs bring to the table and, we
hope, the ambiguity of the terminology will not hinder the
A. W. Colombo is with the Institute for Industrial Informatics, Automation
and Robotics (I2AR), University of Applied Sciences Emden/Leer, Germany effort.
and also as a consultant for Schneider Electric, Industry Business Unit. Email:
[email protected] II. F ROM CPS TO ICPS
S. Karnouskos is with SAP, Germany. E-mail: stama-
[email protected] A decade ago, around 2006, the term CPS was coined
O. Kaynak is with Boğaziçi University, Turkey, and Harbin Institute of to “refer to the integration of computation with physical
Technology, China. E-mail: [email protected] processes” [9]. CPSs can be described as smart systems that
Y. Shi is with University of Victoria, Canada. E-mail: [email protected]
S. Yin is with Harbin Institute of Technology, China. E-mail: encompass hardware, software, and computational and physi-
[email protected] cal components, seamlessly integrated and closely interacting
IEEE IEM, Vol. 11, No. 1, 2017

to sense and control in real time the changing state of the real mechatronic components to complex monitoring and control
world. These systems involve a high degree of complexity at systems. The latter perform functions related to Supervisory
numerous spatial and temporal scales and highly networked Control And Data Acquisition (SCADA), Distributed Control
communications integrating their computational and physical Systems, and Manufacturing Execution Systems (MESs) and
components. As such, CPSs refer to ICT systems (sensing, operate in sync with other enterprise-wide systems, such as
actuating, computing, communicating, and so on) embedded Enterprise Resource Planning (ERP) and business objectives
in physical objects, interconnected through several networks, in real time [1,16]–[18].
including the Internet, and providing citizens and businesses The applicability of such ICPS-based infrastructures covers
with a wide range of innovative applications based on digi- a wide spectrum of domains, such as manufacturing, intelligent
talized data, information, and services. Therefore, CPSs are transportation, real-time health care, smart grids, smart cities,
also ubiquitous embedded cyberphysical applications that are cyber defense, aerospace, and enterprise systems, just to
surfacing (emerging) and are now bridging the physical and name a few. The resulting ICPS infrastructure is continuously
virtual worlds and share all kinds of collaborative networks evolving and depicts emergent behavior empowered by its
[1,10,11]. underlying complex capabilities of the ICPS. Hence, it can
Ontologically, the term CPSs means “hardware–software address problems that the individual components operating
systems that tightly couple the physical world and the digital- alone would not be able to realize, no matter how computa-
ized (virtual) world.” However, CPSs are not merely networked tionally powerful or intelligent they are (in a world of limited
embedded systems but software-intensive, intelligent systems resources [19]). In addition, it yields sophisticated control,
with the capability to collaborate, adapt, and evolve. In a automation, and management functionalities because of inter-
CPS ecosystem, on one hand, every real physical object has action, cooperation, composition of individual capabilities, and
one or more cyber representations, and, on the other hand, orchestration of existing and emergent features.
a cyber component or system can be linked to a physical From an economic viewpoint, the disruptive technologies
representation, i.e., an object in the three-dimensional human- emerging from combining the cyber and physical worlds are
tangible world. For both, the physical and cyber parts will already providing an innovative ecosystem for a broad range
form views from the mechatronics, information, communica- of industries, creating new markets and platforms for growth
tions, and control perspectives. Moreover, these objects are [2]. New products and services, mainly based on the use
increasingly interconnected, networked either permanently or and application of existing big data (now digitalized, easy to
in an asynchronous manner from time to time [1,11,12]. In assess, and tailored for specific contexts), are facilitating the
this context, ICPSs address the penetration and proliferation creation of new functionalities based on the collaboration of
of such ecosystems into the industrial environments. heterogeneous systems in cyberspace as well as the creation of
ICPSs forge the core of real-world networked industrial new and retention of existing high-value jobs and supporting
infrastructures having a cyber representation through digital- the continuous quality of life improvements for the citizens of
ization of data and information across the enterprise, along the the digitalized society.
product and process engineering lifecycle and from suppliers Since the Internet of Things (IoT) combines the power of
to customers along the supply chain. As such, the competitive ubiquitous systems, such as sensors, actuators, networking,
performance of ICPSs mainly depends on the ability to ef- and in-network (collaborative) processing, with modern CPS
fectively collect, analyze, and use large-scale digitalized data technologies, all (heterogeneous) things exhibit the capability
and information from many different and often heterogeneous to interact with each other and actively participate in global
sources to sustainably and efficiently manage, supervise (con- business processes [13]. These interactions build on an array
trol, monitor, diagnose, provide maintenance), and operate in of capabilities, such as information exchange concerning the
the industrial environments. This effective information-driven identity, location, states, and functionalities of physical objects
interaction of ICPSs with other CPSs and enterprise systems, that are made available (á la carte) by their cyber part over
extending to all business processes, is viewed as vital to the Internet and cloud [20], anytime and everywhere for those
modern industries and the rapid paces at which they operate stakeholders (e.g., other things) that need them. However,
[13]. while the ICPS acts as an enabler for building sophisticated
ICPS operate at multiple levels in an open and collaborative information-driven interactions, the fact is that the complexity
manner [10,14], forging the next generation of industrial of the involved stakeholders increases (e.g., of production
systems that are highly sophisticated and strongly coupled with and supplier networks), and, therefore, automated intelligent
the technical and business objectives pursued by the enterprise approaches are needed to effectively deal with it. Higher-level
[13,15]. As such, the engineering and industrial communities concepts of autonomous and self-X systems fit well with the
witness the emergence of a new generation of CPS-based evolution taking place in the infrastructure and the services
industrial systems that are multidisciplinary in nature, with they offer.
functions that encompass several layers of an enterprise, with ICPSs supervise, monitor, control, and manage real-world
wide applicability in several domains and capable of forming physical infrastructures and, therefore, have a real-world im-
large and complex ecosystems. For example, in industrial pact, especially in industrial infrastructures, as demonstrated
automation, the future shop floor is being transformed to a by the analysis and lessons learned from the Stuxnet virus
multifaceted ICPS ecosystem featuring, in various degrees, a [21]. This also implies a tremendous paradigm shift in the
wide variety of components, e.g., from individual sensors and behavior of the society that is in a symbiotic relation with

2
Preprint version of http://dx.doi.org/10.1109/MIE.2017.2648857
IEEE IEM, Vol. 11, No. 1, 2017

ICPS ecosystems: has been devoted to funding CPS activities [26]. Multiple
• The workforce is not only interacting with CPSs but also other agencies of the federal government have been promot-
becoming an integral part; i.e., transforming into another ing and accelerating the research and development of CPSs.
CPS that, in turn, interacts over Internet technologies with In the United States, the Smart Manufacturing Leadership
the ICPS ecosystem. Consortium (SMLC) [27] aims to design a smart factory
• Subject matter experts transform into knowledge workers featuring adaptability, resource efficiency, and ergonomics as
that analyze complex information at the right time in the well as the integration of customers and partners in business
right place and make decisions. and value processes. The Industrial Internet Consortium (IIC)
• Although the subject matter expert continues to be au- features several heavyweight companies active in the ICPS
tonomous, via the CPS interaction, his capabilities and domain and aims to create common interoperable and open
effectiveness are increased. architectural frameworks encompassing initiatives that connect
• The workforce now collaborates and offers its services, and integrate objects with people and processes [28].
which can be requested by any other CPS (including other The European research, development, and innovation pro-
machines). gram HORIZON 2020 (with a total budget of approx. e80
Effectively, the personnel are in a symbiotic relationship billion [29]), features several CPS-related actions, which are
and work in tandem with the ICPS, as both are part of a viewed as key enablers for several domains [30]–[32]. This is
larger collaborative CPS ecosystem. As such, design [22,23], also done in conjunction with the Private-Public-Partnership
implementation, and operation of ICPS and management of (PPP) program of the ARTEMIS Industry Association [33],
the resulting infrastructure are aspects of key importance where industry-specific aspects are addressed and, as discussed
for educating and forming the human–CPS [24] within this in the Strategic Research Agenda (SRA) [25], ICPS are
industrial environment. considered as a key multi-disciplinary enabler of the industrial
As analyzed, a new generation of smart systems is increas- digital revolution.
ingly embedded into the industrial environment, transforming In 2015, the CPS Public Working Group [34] completed
them into sophisticated infrastructures. The technological, and released the Framework for CPSs, and, at the same time
economic, and social impacts of these developments are so research and developments institutions from several European
enormous that the whole process is labeled as the Fourth countries are carrying out Road2CPS [12], creating a joint
Industrial Revolution, often referred to as Industry 4.0 in action plan, for the future development of CPS through
Germany [8]. Networks and processes have so far been limited roadmaps, impact multiplications and constituency building.
to one factory, but in an Industry 4.0-compliant scenario, The German government has recently initiated the imple-
these boundaries of individual factories are no longer a con- mentation of the Industry 4.0 Platform [35], which mainly
straint, and they are lifted to allow interconnection of multiple promotes the ICPS as the Fourth Industrial Revolution [7,8].
stakeholders, e.g., factories, suppliers, and customers, even Similar activities can be found in several other European
in different geographical regions or operated via (sometimes countries e.g., in Spain with the initiative Industria Conectada
competing) stakeholders according to their business needs. [36] and the Netherlands’ program on Smart Industry [37].
This evolution shows that changes in the emerging Industry 4.0 Corporate research and development programs, such as the
economy are likely to come more from the introduction of new Industry 4.0 from Bosch [38], the EcostruXure program from
business models, new organizing principles, and best practices Schneider Electric [39], the Mitsubishi Electric IoT / Industry
around which business is built, mainly capitalizing on 1) the 4.0 initiative [40], the Siemens Industrie 4.0-related programs
knowledge generated during the development and utilization [41], the SAP Industry 4.0 [42], or the Rockwell Automation
of services, exposed and/or consumed by the networked things Connected Enterprise initiative [43], are all addressing the
and 2) increasing ratio of technology time-to-market to tech- digital interconnection of humans, machines, products and
nology time-on-market [25]. systems as cyberphysical objects forming a holistic solution
ecosystem.
The stakes are high, and efforts are not expected to be
III. G LOBAL I NTEREST IN ICPS isolated to the ICPS domain, as their impact relies on building
CPSs and their penetration into the industrial environment up collaboration and interoperability. While we are still at the
have been widely investigated and are the focus of an enor- dawn of the era, after an initial introduction of initiatives at
mous set of research and innovation activities around the the country level, global cross-nation collaboration is needed,
world. Several labels have been used in recent years such as as ICPSs have the potential to transform the lives of billions
ICPS, (industrial) IOT, Industry 4.0, and the like, and, although of people. For instance, U.S. and European universities are
differences exist, at their core, similar aspects prevail. In this collaborating in the Trans-Atlantic Modelling and Simula-
article, the focus is on the area they cover at large and not on tion for CPS (TAMS4CPS) project [44], with the aim to
their differences. develop a strategic research and collaboration agenda to foster
Several programs with significant research budgets have trans-Atlantic research in modeling and simulation for CPSs.
been devoted in the last decade to CPSs in the United States, Another example is the Industrial Internet Consortium and
Europe, and all over the world. In the United States, the the Industrie 4.0 platform collaboration, which are working
National Science Foundation (NSF) has identified CPSs as together toward assessing their respective architectures, i.e.,
a key area of research, and a multi-million dollar budget the Reference Architecture Model for Industrie 4.0 (RAMI4.0

3
Preprint version of http://dx.doi.org/10.1109/MIE.2017.2648857
IEEE IEM, Vol. 11, No. 1, 2017

[45]) and the Industrial Internet Reference Architecture (IIRA) as a collection of CPS services. The latter exists in the CPSs
in order to align them and make them interoperable [46]. or traditional systems, as well as the cloud, giving rise to
Based on expert analysis [47], as well as input from advisory a highly heterogeneous, dynamic, and adequately performing
and consulting companies [48]–[50] and international industry ecosystem of services [4,14,20]. With such services, applica-
consortia [28], it is evident that new smart ICPS and related tions can cherry-pick the functionalities they need to rapidly
technologies [6,51] and infrastructures are currently driving and efficiently fulfill their goals.
research, innovation, competition and disruptive business op- In addition, the role of service-oriented interactions as well
portunities in a broad set of sectors such as agriculture, energy as the cloud are changing the way CPSs are designed, de-
and smart grid, transportation, collaborative and networked or- ployed, and managed. As shown in Figure 2, the cloud can host
ganizations, smart city, building design, industrial automation, a variety of auxiliary services and components that can interact
healthcare, and manufacturing. with the CPS and enhance its capabilities. By utilizing the
cloud-intrinsic capabilities, such as virtualization, scalability,
IV. R ESEARCH AND I NNOVATION C HALLENGES multi-tenancy, performance, and lifecycle management, better
CPSs can be realized that may include more lightweight
A. Major Challenges devices since demanding parts (e.g., computation intensive)
Business continuity and agility form the core modus can now be hosted in the cloud [1,4,20]. Cloud-assisted CPS
operandi of modern global enterprises [13], and efforts that is viewed as a key enabler for a multitude of scenarios both
yield results of more efficient automation systems are well- in Peer-to-Peer (P2P) scenarios as well as in cross-layer ones
justified. ICPSs, as a convergence of several complementary (e.g., between SCADA and ERP).
technologies, including the larger scope of CPS [1,6], IoT Future industrial infrastructures are expected to be complex
[51] and Internet of Services, can play a pivotal role towards System of Systems (SoS) [14,32,67] that will empower a
enabling enterprises to achieve their goals. As such, they face new generation of applications and services that are hardly
several challenges pertinent to their disaggregated technolo- realizable today or too costly to achieve. New sophisticated
gies, architectures and domains. Considering the key trends enterprise-wide monitoring and control approaches will be
identified [9,11,22,52]–[63], including Information Driven In- possible due to the prevalence of ICPSs and, especially, the
teraction, Distributed Business Processes, Cloud Computing formation of systems of ICPSs (SoICPSs). In ICPSs (and
and visualization, Cooperation, Multi-core systems and GPU SoICPSs), components can be dynamically added or removed,
computing, SOA-ready devices, the key question that arises is and dynamic discovery enables on-demand information com-
how to take advantage of their benefits in order to provide the bination and collaboration [54].
multi-faceted ICPS envisioned that fully cover the industrial Engineering industrial solutions based on SoICPS are offer-
requirements. ing new challenges and innovation opportunities [32] as the
The prevalence of digitalization across all layers of an following:
enterprise needs solutions that will support ICPS engineering • SoICPSs are continuously evolving, which softens, or
at device, system, infrastructure, and application levels. This even completely removes, the traditional separation be-
includes the whole lifecycle, from cradle to grave, of all kinds tween the engineering/design phases and the operational
of ICPS components that can be considered as a set of major stages.
challenges that need to be tackled. Examples of challenging as- • The high degree of heterogeneity, uncertainty, and par-
pects include multidisciplinary [58] engineering methods and tial autonomy of SoICPSs require new, fully integrated
tools for evolvable ICPS-architectures [64]; intelligent moni- approaches for their design, validation, and operation.
toring [65]; understanding and managing emergent behaviors • SoICPSs are highly flexible and thus subject to frequent,
in networked ICPSs; designing and implementing human-to- dynamic reconfiguration, which must be supported by
machine, business-to-machine, and business-to-human inter- design support tools to enable efficient engineering.
actions; improving existing or creating new communication To this end, supporting engineering tools [23,64] will also
and information technologies guaranteeing connectivity and need to be networked and integrated with the different phases,
interoperability among cyber and physical components; and from design through development, commissioning, deploy-
so on. ment, operation, and maintenance of the digitalized industrial
To achieve the pursued agility and continuity, business environment.
processes performed in highly distributed production systems ICPSs enable monitoring and control of (industrial) physical
need to be efficiently integrated with a sophisticated shop-floor processes and bridge the cyber and virtual worlds. Their
infrastructure that is capable of responding to dynamic adap- impact across the value chain is increasingly evident [68],
tations in a timely manner [15]. As an example, considering especially as they are the key towards digitalization of the
envisioned architecture transitions such as the one shown in industrial environment. Hence, future and emerging technolo-
Figure 1, the high-level changes imposed on engineering of gies [1,15,69] and paradigms for implementing ICPSs, such
future automation systems are becoming easier to recognize. as SOA, cloud computing, the IoT, big data, and the industrial
Figure 1 advocates that, in parallel to traditional hierarchical Internet, need to be deeply investigated, especially in real-
architectures in industrial infrastructures, selected function- world operations. This is by no means an easy undertaking,
alities at different levels [as defined by the International considering also that nowadays ICPSs are built with highly
Society of Automation (ISA)-95 paradigm] can be exposed heterogeneous hardware and software components and depend

4
Preprint version of http://dx.doi.org/10.1109/MIE.2017.2648857
IEEE IEM, Vol. 11, No. 1, 2017

Figure 1. The transitioning towards a SOA-based information-driven architecture [1,66]

on systems and services that are operated by third-party stake- entities are capable of working in a proactive manner, initiating
holders. Any research and innovation activity should integrate collaborative actions and dynamically interacting with each
well with existing prevalent architectures such as ISA-95/ISA- other to achieve both local and global objectives along three
88 [1,70,71]. New platforms for industrial innovation, from basic collaboration axes, i.e., enterprise, supply-chain, and
physical product to product-as-a-service in cyberspace, will lifecycle [14]. From the physical device control level up to the
have to be introduced in several industrial domains. However, higher levels of the business process management system, as
legacy systems must also be migrated into service-based col- defined in ISA-95 and ISA-88 [70] i.e., from suppliers through
laborative ICPS ecosystems, which is not to be taken lightly. the enterprise to the customer [14,72], and from design [58]
Business becomes digital business, and, as such, ICPSs are through operation to recycling phases of an engineering system
playing a pivotal role across the value chain. Empowered lifecycle, collaboration will be enabled if, on the one hand, the
by the cross-ICPS collaboration, the supply chain can utilize involved systems act and react on their environment, sharing
ICPSs to enhance its operations across all the areas of sourc- some principal commonalities and, on the other hand, have
ing, procurement, conversion, logistics, partner coordination, some different aspects that complement each other to form a
and collaboration within an organization and across organiza- coherent group of objects that cooperate with each other to
tions. ICPSs can empower the development and application interact with their environment [10,14].
of new business models, where new forms of interactions
and relationships between supplier and customer are realized:
interaction between products and interaction between services, Overall, creating synergies among ICPS stakeholders, es-
physical but also cyberinteractions repositioning a partner pecially in a cross-domain manner, is a major challenge.
within the value chain. New innovations toward supply-chain CPSs are viewed as a key part of critical infrastructures, e.g.,
resilience, traceability, accountability, compliance, and so on the energy domain [60,61]. Future smart cities will integrate
can be realized across its spectrum. Building new ICPS- multiple such systems in a harmonized way to enable new
dependent cross-stakeholder applications as well as operating innovative services for their citizens. Hence, the factory of the
them within the necessary service-level agreements is per- future may well be situated within the smart city and alongside
ceived as a challenging task, considering the dynamics and smart buildings and smart houses. The latter will take full
emerging complexity of the ICPS infrastructure. advantage of the energy available in the grid, and all forms of
Collaboration is a major challenge. This has to be done energy side-products, such as heat, will not be wasted but fully
in a cross-layer fashion across the enterprise, as well as integrated, e.g., for heating houses, public buildings, and so on.
with all external stakeholders [1]. The umbrella paradigm As such, a major challenge is to be able to identify synergies
underpinning novel collaborative systems is to consider the among different domains, create interoperable solutions that
set of intelligent system units as a conglomerate of distributed, take advantage of the sophisticated infrastructure, and optimize
autonomous, intelligent, proactive, fault-tolerant, and reusable them according to business, environmental, health, social, and
units, which operate as a set of cooperating entities [5]. These other high-level objectives.

5
Preprint version of http://dx.doi.org/10.1109/MIE.2017.2648857
IEEE IEM, Vol. 11, No. 1, 2017

Figure 2. A Cloud-based CPS Infrastructure [1,54]. M2M: machine-to-machine; GW: gateway; CEP: complex event processing; DCS: distributed control
system; P2P: peer-to-peer.

B. Cross-Cutting Issues hence they are perceived as key challenges because they act as
enablers and building blocks for further ICPS developments.
Several key challenges in multiple domains have been
identified for CPSs [9,11,22,52]–[63]. Taking these into con- On ICPS management areas, methods and tools addressing
sideration along with discussions with CPS experts in industry, the management aspects are perceived as a key enabler for
as well as empirical experiences building up such systems in industry. Management and coordination not only of standalone
a multitude of industry-led research projects, some key issues ICPSs but also of multilayer large-scale ICPSs are considered
that arise are shown in Table I. The listed cross-cutting chal- as a must to tackle increasing complexity and effectively
lenges are especially relevant for ICPSs and their system-of- integrate ICPSs in organizational processes. In addition the
system constellations and can be roughly clustered to six major security, safety, and trust aspects [18,21,74], not only at the
areas [56]: 1) ICPS capabilities, 2) ICPS management, 3) ICPS ICPS level but also large-scale systems, are seen as challeng-
engineering, 4) ICPS ecosystems, 5) ICPS infrastructures, and ing, especially due to the nature of interactions among ICPSs
6) ICPS information systems. It has to be pointed out that that are no longer a priori known and centrally managed. As
Table I provides a rule of thumb with some subjective views on an example, while safe and secure standalone ICPSs can be
key challenges, their difficulty, industry priority, and estimated constructed, when put in an ecosystem and via its interactions
time frame to reach maturity, i.e., a technology readiness level with other ICPSs, the safety aspects of the system and potential
(TRL) of at least TRL-7 that assumes the realization of a cascading effects can no longer be guaranteed nor be fully
system prototype demonstration in operational environment tested in lab [75]. Although industrially mature solutions are
(hence, it should be taken with a grain of salt). expected in the mid to long term, tackling challenges they
In the ICPS capabilities area, several issues related to real- pose, especially when considering very large systems, is also
time monitoring, control, and management on ICPSs as well as expected to be instrumental to ICPS acceptance.
SoICPSs and their design and optimization [59,73], which are ICPS and SoICPS engineering is viewed as a high-priority
perceived as highly challenging, are presented. The priorities endeavor, especially by industry practitioners who will design,
range from medium to high, and the control aspects [73] are deploy, and operate future ICPS-enabled landscapes. As such,
crucial for the acceptance of CPSs in industrial facilities. Some methods, tools, models, and practices related to lifecycle
of the aspects feature varying degrees of difficulty, and one management are needed for industrial settings [64,76]. Model-
could claim that they are, in part, even possible today, e.g., based engineering ICPS solutions, achievement of resilience
the servification of CPSs, energy efficiency, on-ICPS analytics, [77] and graceful degradation, safe programming and vali-
and so on. However, such efforts represent mostly a transfer dation, as well as simulation [76,78] of complex ICPS in-
of selected functionalities from other higher levels (e.g., from frastructures are high on the agenda. With the high degree
MES or ERP) and do not take in full consideration the intrinsic of hardware heterogeneity, new developments in operating
capabilities of ICPSs as well as their operational context, systems and programming languages tailored to CPSs may

6
Preprint version of http://dx.doi.org/10.1109/MIE.2017.2648857
IEEE IEM, Vol. 11, No. 1, 2017

Table I
K EY CROSS - CUTTING CHALLENGES FOR ICPS AND S YSTEMS OF ICPS (S O ICPS)

Area Key Challenges Difficulty Priority Maturity in

Real-time monitoring, control & management of ICPS / SoICPS high high 4-7 years
Seamless (real-time) service-based interaction of ICPS with human-actors/users high high 4-6 years

ICPS Optimization in ICPSs and their task-specific application high medium 4-7 years
Capabilities On-ICPS advanced analytics and decision making medium high 3-5 years
Sentient SoICPS: Autonomous collaboration among ICPSs high high 8-10+ years

Energy efficient ICPSs medium medium 3-5 years

Lifecycle management of ICPSs medium medium 5-8 years
ICPS
Management/coordination of multi-domain large scale ICPSs and SoICPSs high high 5-10 years
Management
Security and trust management for ICPSs and SoICPSs high high 5-8 years
Model-based engineering methods covering the full life cycle of ICPSs and SoICPSs high high 4-7 years
Safe programming, validation, resilient, risk-mitigating ICPSs and SoICPSs high high 5-10+ years
ICPS
Engineering tools and practices for ICPSs’ lifecycle support high high 3-7 years
Engineering
New operating systems and programming languages for ICPSs and SoICPSs medium low 3-6 years

Simulation of ICPSs and large-scale SoICPSs medium high 3-6 years

Multi-domain ICPS interoperability, management, control, QoS etc. medium high 4-7 years
Migration solutions to full ICPS infrastructure and cohabitation with legacy systems medium high 3-6 years
ICPS
Integration, resilience, robustness and sustainability of ICPS critical infrastructures high high 5-8 years
Infrastructures
Provision of ubiquitous ICPS data & information services medium medium 3-5 years
Economic, social, environmental etc. impact of ICPS infrastructures high high 4-8 years
Autonomic and self-X featured ICPSs high medium 7-10+ years

ICPS Education/training to enable fast assimilation of ICPSs (humans in the loop) high high 3-7 years
Ecosystems Collaborative ICPSs (intelligent autonomous ICPS collaboration) medium medium 5-8 years

Cross-industry knowledge base, best practices & emergent behavior at SoICPS level high high 8-10+ years
Artificial Intelligence in ICPSs / SoICPSs high high 7-10+ years
Cross-Domain large-scale information management in ICPS infrastructures medium low 6-9 years
ICPS Transformation of ICPS data and information analytics to actionable knowledge high high 4-8 years
Information
Systems ICPS automated knowledge-driven decision making, management, and risk analysis high medium 6-10+ years

Autonomous smart ICPSs within an ambient-intelligent-ecosystem high medium 8-10+ years

arise, which, however, also may be tackled by significantly resolve them so that they can act as enablers for the ICPS.
extending existing approaches [79]. A typical example is the social, environmental, and economic
In ICPS infrastructures, the key issues identified are related impact of ICPSs, which have cascading effects and impact
to management, control, interoperability, migration, quality the wider adoption of ICPSs in real industrial environments
of service, and so on, which become increasingly important, [68]. Migration is also an aspect that must be considered at
especially from the operational point of view. Especially for infrastructure level, as the pace of technology associated with
ICPSs in critical infrastructures, resilience, robustness [81], ICPS solutions is increasing.
safety and sustainability are preeminent [82]. The difficulty ICPS ecosystems and the business benefits they bring will
level is perceived as mostly medium, as significant efforts are increasingly be the focus once sophisticated ICPSs and in-
already underway on how to tackle them. Nevertheless, even frastructures are in place. In this area, the key challenges
if some of them might be more trivial, it is mandatory to would be related to design [22], deployment, and operation

7
Preprint version of http://dx.doi.org/10.1109/MIE.2017.2648857
IEEE IEM, Vol. 11, No. 1, 2017

Emergent Behaviours

Migration Path 1

Migration Path i
Service
Cloud

Control / Interlocking
Sensor/Actuators

Process Migration Path n

Orchestrated/Composed Service
Current/Legacy Systems Service Next Generation SoA-based Systems

Figure 3. The migration to SOA-based Infrastructure [1,80]

of collaborative, autonomous and self-X [19] features of ICPS view of the steps that need to be undertaken is depicted in
ecosystems. Such ecosystems build upon intelligent ICPSs Figure 3. Migration could be evolutionary, with incremental
and enhance all possible interactions with the environment migration of features to the new infrastructure, i.e., capturing
and the involved stakeholders, including the ICPS-to-human of functionalities as these are described in each ISA-95 layer
interaction [24] that may lead to enhanced workforce perfor- and making them available as ICPS services. As there are
mance and satisfaction. For the latter to be achieved, education several interdependencies, the potential migration paths must
and training to enable fast assimilation of ICPSs need to be assessed, and a migration should be done stepwise. Such
be properly tackled. Although many of the aforementioned migration will also unleash at system level emergent behav-
aspects are expected to mature only in the long run, e.g., iors because of the dynamic interactions among the different
the capability to exploit cross-industry information derived devices and systems. Top-down and bottom-up approaches
from the big amounts of generated and associated data and will need to be analyzed in detail [1,71,80], and the resulting
transform it to knowledge at ICPSs/SoICPSs level, the path to migration strategies can be highly complex, depending on the
dealing with these challenges is viewed as highly challenging, preconditions, requirements, and goals. Figure 3 makes it clear
and the technologies to do so have not been adequately that the migration is not a one-time operation but, rather, a
investigated yet. continuous one that the industry will have to get accustomed
ICPS information systems are also viewed as integral parts to. As ICPSs, their services, and infrastructure evolve, it is
in realizing the ICPS vision and being able to capitalize on the expected that migration will be part of the daily development
data, information, and knowledge acquired within the organi- and operational aspects in such systems. In addition, any
zation, as well as from the (collaborative) cross-organizational migration strategies have a multitude of goals that go beyond
interactions. The revolutionary progress is coupled with the technology and include cost effectiveness, resource efficiency,
highly intelligent ICPSs, as well as the emergent behav- agility, deterministic behavior, operational easiness, business
iors stemming from their interactions at the SoICPSs level. continuity, and so on [13].
Understanding and transforming information to knowledge
in an automated fashion [17,23], so that operations, such V. C ONCLUSIONS
as management, risk analysis, and decision making, can be ICPSs are perceived as a promising approach that extends
realized by ICPSs are perceived as challenging, especially the CPS’s overall activities to the industrial domain and
when one considers that these are not predefined tasks but, considers predominantly their requirements such as safety, mi-
rather, emerge from machine intelligence and operation of gration, compliance, agility, business continuity, performance,
ICPSs in real-world dynamic environments (e.g., self-driving and collaboration. Although we are still at the dawn of an era
cars). described as the Fourth Industrial Revolution, the contributions
Migration is often not given adequate attention in CPS of ICPSs toward the vision are pivotal and have far-reaching
solutions, but for ICPSs, it is highly prioritized. Since most impact on industry, the economy, society, and the environment.
industrial infrastructures are brownfields, any ICPS has to This is attested by several ICPS efforts carried out worldwide,
be capable of operating, co-existing, and integrating legacy tackling design and operation of ultralarge-scale systems [83],
systems [71], while in parallel transition to the new infras- as well as assessment of their impacts.
tructure. Considering the migration to an information-flat and Although the promises of ICPSs are significant and the
service-based infrastructure as shown in Figure 1, a high level opportunities they bring are multifaceted, these can be realized

8
Preprint version of http://dx.doi.org/10.1109/MIE.2017.2648857
IEEE IEM, Vol. 11, No. 1, 2017

only if key challenges are addressed to ease their industrial [20] S. Karnouskos and V. Somlev, “Performance assessment of integration
adoption. The prevalent focus ought to be placed on the in the cloud of things via web services,” in 2013 IEEE International
Conference on Industrial Technology (ICIT), 2013.
integration and collaboration of ICPSs not only within an orga- [21] S. Karnouskos, “Stuxnet Worm Impact on Industrial Cyber-Physical
nization but at large scale and within a global ICPS ecosystem. System Security,” in IEEE IECON, Melbourne, Australia, Nov. 2011.
Added value can be generated based on collaboration among [22] S. K. Khaitan and J. D. McCalley, “Design techniques and applications
of cyberphysical systems: A survey,” IEEE Systems Journal, vol. 9, no. 2,
disparate ICPS systems and services, and innovation can be pp. 350–365, jun 2015.
accelerated. Therefore, collaborative ICPS are perceived as [23] E. Francalanza, J. Borg, and C. Constantinescu, “A knowledge-based
enablers towards future society vision realization and goal tool for designing cyber physical production systems,” Computers in
Industry, vol. 84, pp. 39–58, jan 2017.
achievement. [24] D. S. S. Nunes, P. Zhang, and J. S. Silva, “A survey on human-in-
the-loop applications towards an internet of all,” IEEE Communications
Surveys & Tutorials, vol. 17, no. 2, pp. 944–965, 2015.
R EFERENCES [25] ARTEMIS, “ARTEMIS Strategic Research Agenda 2016,” ARTEMIS
Industry Association, Tech. Rep., Mar. 2016. [Online]. Available:
[1] A. W. Colombo, T. Bangemann, S. Karnouskos, J. Delsing, P. Stluka, https://artemis-ia.eu/publication/download/sra2016.pdf
R. Harrison, F. Jammes, and J. L. Lastra, Eds., Industrial Cloud-based [26] Cyber-Physical Systems (CPS) Active Awards. National Science
Cyber-Physical Systems: The IMC-AESOP Approach. Springer, 2014. Foundation (NSF). [Online]. Available: http://goo.gl/MxqPkx
[2] WEC, “Industrial Internet of Things: Unleashing the Potential of [27] Smart Manufacturing Leadership Coalition. [Online]. Available: http:
Connected Products and Services,” World Economic Forum (WEC), //www.smartmanufacturingcoalition.org/
Tech. Rep., 2015. [Online]. Available: http://www3.weforum.org/docs/ [28] Industrial Internet Consortium. [Online]. Available: http://www.
WEFUSA IndustrialInternet Report2015.pdf iiconsortium.org
[3] European Commission, “ICT for societal challenges,” 2013. [Online]. [29] Horizon 2020 Budget. [Online]. Available: http://cerneu.web.cern.ch/
Available: http://goo.gl/CS13BV horizon2020/budget
[4] R. Chaâri, F. Ellouze, A. Koubâa, B. Qureshi, N. Pereira, H. Youssef, [30] Cyber-Physical Systems: Uplifting Europe’s innovation capacity.
and E. Tovar, “Cyber-physical systems clouds: A survey,” Computer European Commission. [Online]. Available: https://www.amiando.com/
Networks, vol. 108, pp. 260–278, oct 2016. cps-conference.html?page=973125
[5] A. W. Colombo and S. Karnouskos, “Towards the Factory of the [31] Cyber-Physical Systems. European Commission. [Online]. Available:
Future: A Service-oriented Cross-layer Infrastructure,” in ICT Shaping https://ec.europa.eu/digital-single-market/en/cyber-physical-systems
the World: A Scientific View. European Telecommunications Standards [32] S. Engell, R. Paulen, M. A. Reniers, C. Sonntag, and H. Thompson,
Institute (ETSI), John Wiley and Sons, 2009, pp. 65–81. “Core Research and Innovation Areas in Cyber-Physical Systems of
Systems,” in 5th International Workshop on Cyber-Physical Systems
[6] A. J. C. Trappey, C. V. Trappey, U. H. Govindarajan, J. J. Sun, and
(CyPhy), Amsterdam, The Netherlands. Springer, 2015, pp. 40–55.
A. C. Chuang, “A review of technology standards and patent portfolios
[33] ARTEMIS Industry Association. [Online]. Available: http://artemis-ia.
for enabling cyber-physical systems in advanced manufacturing,” IEEE
eu/
Access, vol. 4, pp. 7356–7382, 2016.
[34] CPS Public Working Group. National Institute of Standards (NIST).
[7] ACATECH, “Cyber-Physical Systems: Driving force for innovation
[Online]. Available: http://pages.nist.gov/cpspwg/
in mobility, health, energy and production,” ACATECH – German
[35] Plattform Industrie 4.0. Federal Ministry of Education and Research,
National Academy of Science and Engineering, Tech. Rep., Dec. 2011.
Federal Ministry for Economic Affairs and Energy, Germany. [Online].
[Online]. Available: https://goo.gl/Q6WFQN
Available: http://www.plattform-i40.de/
[8] H. Kagermann, W. Wahlster, and J. Helbig, “Securing the future of
[36] Industria conectada 4.0. [Online]. Available: http://www.
German manufacturing industry: Recommendations for implementing
industriaconectada40.gob.es
the strategic initiative INDUSTRIE 4.0,” ACATECH – German National
[37] “Smart Industry.” [Online]. Available: http://www.smartindustry.nl/
Academy of Science and Engineering, Tech. Rep., Apr. 2013. [Online].
[38] BOSCH. [Online]. Available: https://www.bosch-si.com/solutions/
Available: http://goo.gl/oc3X4n
manufacturing/industry-4-0/industry-4-0.html
[9] E. A. Lee and S. A. Seshia, Introduction to Embedded Systems – A [39] EcosStruxure: Active energy management architecture from power
Cyber-Physical Systems Approach, 2nd ed. online, 2015. [Online]. plant to plug. Schneider Electric. [Online]. Available: http://www2.
Available: http://leeseshia.org schneider-electric.com/medias/ecostruxure/ecostruxure-brochure.pdf
[10] S. Karnouskos, V. Vilaseñor, M. Handte, and P. J. Marrón, “Ubiqui- [40] A. Nakayama. (2015, Jun.) Japan launches forum to counter German ini-
tous integration of cooperating objects,” International Journal of Next- tiative. Nikkei Asian Review. [Online]. Available: http://asia.nikkei.com/
Generation Computing (IJNGC), vol. 2, no. 3, 2011. Business/Trends/Japan-launches-forum-to-counter-German-initiative
[11] P. Leitao, S. Karnouskos, L. Ribeiro, J. Lee, T. Strasser, and A. W. [41] Siemens. [Online]. Available: http://www.siemens.de/industrie-4.0/
Colombo, “Smart agents in industrial cyber–physical systems,” Proceed- [42] SAP. [Online]. Available: http://news.sap.com/tags/industry-4-0/
ings of the IEEE, vol. 104, no. 5, pp. 1086–1101, may 2016. [43] The Connected Enterprise. Rockwell Automation. [Online].
[12] Road2CPS project. [Online]. Available: http://www.road2cps.eu Available: http://www2.schneider-electric.com/medias/ecostruxure/
[13] S. Karnouskos, “Efficient Sensor Data Inclusion in Enterprise Services,” ecostruxure-brochure.pdf
Datenbank-Spektrum, vol. 9, no. 28, pp. 5–10, Feb. 2009. [44] Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
[14] A. W. Colombo, S. Karnouskos, and T. Bangemann, “A System of (TAMS4CPS) project. [Online]. Available: http://www.tams4cps.eu
Systems view on Collaborative Industrial Automation,” in 2013 IEEE [45] DIN, “Reference Architecture Model Industrie 4.0 (RAMI4.0),”
International Conference on Industrial Technology (ICIT), feb 2013. Deutsches Institut für Normung (DIN), Tech. Rep., Apr. 2016. [Online].
[15] S. Karnouskos, “Realising next-generation web service-driven industrial Available: http://www.din.de/en/wdc-beuth:din21:250940128
systems,” The International Journal of Advanced Manufacturing Tech- [46] IIC. (2016, Mar.) Plattform Industrie 4.0 and Industrial Internet
nology, vol. 60, no. 1-4, pp. 409–419, Sep. 2011. Consortium agree on cooperation. Press Release. [Online]. Available:
[16] S. Karnouskos and A. W. Colombo, “Architecting the next generation http://www.iiconsortium.org/press-room/03-02-16.htm
of service-based SCADA/DCS system of systems,” in IEEE IECON, [47] J. Lee, “Smart factory systems,” Informatik-Spektrum, vol. 38, no. 3, pp.
Melbourne, Australia, 7–10 Nov 2011. 230–235, May 2015.
[17] S. Iarovyi, W. M. Mohammed, A. Lobov, B. R. Ferrer, and J. L. M. [48] G. Gorbach, C. Polsonetti, and A. Chatha, “Planning for the
Lastra, “Cyber–physical systems for open-knowledge-driven manufac- Industrial Internet of Things,” ARC Advisory Group, Tech. Rep.,
turing execution systems,” Proceedings of the IEEE, vol. 104, no. 5, pp. Jan. 2014. [Online]. Available: http://www.arcweb.com/brochures/
1142–1154, may 2016. planning-for-the-industrial-internet-of-things.pdf
[18] A. Sajid, H. Abbas, and K. Saleem, “Cloud-assisted IoT-based SCADA [49] McKinsey, “Industry 4.0: How to navigate digitization of the
systems security: A review of the state of the art and future challenges,” manufacturing sector,” McKinsey Digital, Tech. Rep., 2015. [Online].
IEEE Access, vol. 4, pp. 1375–1384, 2016. Available: http://www.mckinsey.de/sites/mck files/files/mck industry
[19] J. S. Preden, K. Tammemae, A. Jantsch, M. Leier, A. Riid, and 40 report.pdf
E. Calis, “The benefits of self-awareness and attention in fog and mist [50] Industry 4.0 in China: Catching up with the west. [Online]. Available:
computing,” Computer, vol. 48, no. 7, pp. 37–45, jul 2015. http://daxueconsulting.com/industry-4-0-in-china/

9
Preprint version of http://dx.doi.org/10.1109/MIE.2017.2648857
IEEE IEM, Vol. 11, No. 1, 2017

[51] A. J. Trappey, C. V. Trappey, U. H. Govindarajan, A. C. Chuang, and [74] T.-W. Um, G. M. Lee, and J. K. Choi, “Strengthening trust in the
J. J. Sun, “A review of essential standards and patent landscapes for the future social-cyber-physical infrastructure: an ITU-t perspective,” IEEE
internet of things: A key enabler for industry 4.0,” Advanced Engineering Communications Magazine, vol. 54, no. 9, pp. 36–42, sep 2016.
Informatics, dec 2016. [75] S. Karnouskos, “Smart houses in the smart grid and the search for value-
[52] J. Sztipanovits, X. Koutsoukos, G. Karsai, N. Kottenstette, P. Antsaklis, added services in the cloud of things era,” in 2013 IEEE International
V. Gupta, B. Goodwine, J. Baras, and S. Wang, “Toward a science of Conference on Industrial Technology (ICIT), feb 2013.
Cyber-Physical System integration,” Proceedings of the IEEE, vol. 100, [76] P. Hehenberger, B. Vogel-Heuser, D. Bradley, B. Eynard, T. Tomiyama,
no. 1, pp. 29–44, Jan. 2012. and S. Achiche, “Design, modelling, simulation and integration of cyber
[53] I. Lee, O. Sokolsky, S. Chen, J. Hatcliff, E. Jee, B. Kim, A. King, physical systems: Methods and applications,” Computers in Industry,
M. Mullen-Fortino, S. Park, A. Roederer, and K. K. Venkatasubrama- vol. 82, pp. 273–289, oct 2016.
nian, “Challenges and research directions in medical Cyber-Physical [77] Z. Huang, C. Wang, A. Nayak, and I. Stojmenovic, “Small cluster
Systems,” Proceedings of the IEEE, vol. 100, no. 1, pp. 75–90, Jan. in cyber physical systems: Network topology, interdependence and
2012. cascading failures,” IEEE Transactions on Parallel and Distributed
[54] S. Karnouskos, A. W. Colombo, and T. Bangemann, “Trends and Chal- Systems, vol. 26, no. 8, pp. 2340–2351, aug 2015.
lenges for Cloud-based Industrial Cyber-Physical Systems,” in Indus- [78] P. Garraghan, D. McKee, X. Ouyang, D. Webster, and J. Xu, “SEED: A
trial Cloud-based Cyber-Physical Systems: The IMC-AESOP Approach. scalable approach for cyber-physical system simulation,” IEEE Trans-
Springer, May 2014. actions on Services Computing, vol. 9, no. 2, pp. 199–212, mar 2016.
[55] P. Palensky, E. Widl, and A. Elsheikh, “Simulating Cyber-Physical [79] P. Soulier, D. Li, and J. R. Williams, “A survey of language-based
Energy Systems: Challenges, tools and methods,” IEEE Transactions on approaches to cyber-physical and embedded system development,” Ts-
Systems, Man, and Cybernetics: Systems, vol. 44, no. 3, pp. 318–326, inghua Science and Technology, vol. 20, no. 2, pp. 130–141, apr 2015.
Mar. 2014. [80] J. Delsing, J. Eliasson, R. Kyusakov, A. W. Colombo, F. Jammes,
[56] P. Leitão, A. W. Colombo, and S. Karnouskos, “Industrial automation J. Nessaether, S. Karnouskos, and C. Diedrich, “A migration approach
based on Cyber-Physical Systems technologies: Prototype implementa- towards a SOA-based next generation process control and monitoring,”
tions and challenges,” Computers in Industry, vol. 81, pp. 11–25, Sep. in IECON 2011 - 37th Annual Conference of the IEEE Industrial
2016, in Press. Electronics Society, Nov. 2011.
[57] S. Han, M. Xie, H.-H. Chen, and Y. Ling, “Intrusion detection in Cyber- [81] M. Rungger and P. Tabuada, “A notion of robustness for cyber-physical
Physical Systems: Techniques and challenges,” IEEE Systems Journal, systems,” IEEE Transactions on Automatic Control, vol. 61, no. 8, pp.
vol. 8, no. 4, pp. 1052–1062, Dec. 2014. 2108–2123, aug 2016.
[58] S. Tripakis, “Compositionality in the science of system design,” Pro- [82] C. Vellaithurai, A. Srivastava, S. Zonouz, and R. Berthier, “CPIndex:
ceedings of the IEEE, vol. 104, no. 5, pp. 960–972, may 2016. Cyber-physical vulnerability assessment for power-grid infrastructures,”
[59] I. Akkaya, P. Derler, S. Emoto, and E. A. Lee, “Systems engineering IEEE Transactions on Smart Grid, vol. 6, no. 2, pp. 566–575, mar 2015.
for industrial cyber–physical systems using aspects,” Proceedings of the [83] L. Northrop, P. Feiler, R. P. Gabriel, J. Goodenough, R. Linger,
IEEE, vol. 104, no. 5, pp. 997–1012, may 2016. T. Longstaff, R. Kazman, M. Klein, D. Schmidt, K. Sullivan, and
[60] X. Yu and Y. Xue, “Smart grids: A cyber–physical systems perspective,” K. Wallnau, “Ultra-Large-Scale Systems - The Software Challenge of
Proceedings of the IEEE, vol. 104, no. 5, pp. 1058–1070, may 2016. the Future,” Software Engineering Institute, Carnegie Mellon, Tech.
[61] X. Shi, Y. Li, Y. Cao, and Y. Tan, “Cyber-physical electrical energy Rep., June 2006. [Online]. Available: https://resources.sei.cmu.edu/
systems: challenges and issues,” CSEE Journal of Power and Energy asset files/Book/2006 014 001 30542.pdf
Systems, vol. 1, no. 2, pp. 36–42, jun 2015.
[62] L. Monostori, B. Kádár, T. Bauernhansl, S. Kondoh, S. Kumara, G. Rein-
hart, O. Sauer, G. Schuh, W. Sihn, and K. Ueda, “Cyber-physical systems
in manufacturing,” CIRP Annals - Manufacturing Technology, vol. 65,
no. 2, pp. 621–641, 2016.
[63] M. Broy and A. Schmidt, “Challenges in Engineering Cyber-Physical
Systems,” Computer, vol. 47, no. 2, pp. 70–72, Feb. 2014.
[64] R. Harrison, D. Vera, and B. Ahmad, “Engineering methods and tools
for cyber-physical automation systems,” Proceedings of the IEEE, vol.
104, no. 5, pp. 973–985, may 2016.
[65] C. Chen, J. Yan, N. Lu, Y. Wang, X. Yang, and X. Guan, “Ubiquitous
monitoring for industrial cyber-physical systems over relay- assisted
wireless sensor networks,” IEEE Transactions on Emerging Topics in
Computing, vol. 3, no. 3, pp. 352–362, sep 2015.
[66] S. Karnouskos, A. W. Colombo, T. Bangemann, K. Manninen, R. Camp,
M. Tilly, M. Sikora, F. Jammes, J. Delsing, J. Eliasson, P. Nappey,
J. Hu, and M. Graf, “The IMC-AESOP Architecture for Cloud-Based
Industrial Cyber-Physical Systems,” in Industrial Cloud-Based Cyber-
Physical Systems. Springer, May 2014, pp. 49–88.
[67] M. Jamshidi, Ed., Systems of Systems Engineering: Principles and
Applications. CRC Press, Nov. 2008.
[68] M. M. Herterich, F. Uebernickel, and W. Brenner, “The impact of cyber-
physical systems on industrial services in manufacturing,” Procedia
CIRP, vol. 30, pp. 323–328, 2015.
[69] A. Boyd, D. Noller, P. Peters, D. Salkeld, T. Thomasma, C. Gifford,
S. Pike, and A. Smith, “SOA in manufacturing - guidebook,” IBM
Corporation, MESA International and Capgemini, Tech. Rep., 2008.
[Online]. Available: ftp://public.dhe.ibm.com/software/plm/pdif/MESA
SOAinManufacturingGuidebook.pdf
[70] ISA-95 and ISA-88, International Society of Automation Std. [Online].
Available: http://www.isa.org
[71] T. Bangemann, M. Riedl, M. Thron, and C. Diedrich, “Integration of
classical components into industrial cyber–physical systems,” Proceed-
ings of the IEEE, vol. 104, no. 5, pp. 947–959, may 2016.
[72] L. Camarinha-Matos and H. Afsarmanesh, Collaborative Net-
works:Reference Modeling: Reference Modeling. Springer, 2008.
[73] S. Xin, Q. Guo, H. Sun, B. Zhang, J. Wang, and C. Chen, “Cyber-
physical modeling and cyber-contingency assessment of hierarchical
control systems,” IEEE Transactions on Smart Grid, vol. 6, no. 5, pp.
2375–2385, sep 2015.

10
Preprint version of http://dx.doi.org/10.1109/MIE.2017.2648857
View publication stats