A Generic Model For Sustainability

A Generic Model for Sustainability
Birgit Penzenstadler, Henning Femmer
Technische Universität München, Germany

penzenst|femmer@in.tum.de
November 26, 2012

Abstract
Motivation: Software systems as we know them have either a primary eco-

nomic purpose or fulfil human or social needs of their users. The economic
purpose is analysed in depth by the economy itself; the latter goals are analysed
in software engineering by service orientation and usability analysis.
Yet, as software systems are involved in nearly all daily business and non-
business processes they also have an impact on resources and environment.
Hence, besides economic, human and social goals, environmental sustainability
should be a major goal for software development projects.
Problem: Without a tangible definition and applicable guidance, sustainabil-

ity remains an unreachable ideal. Therefore, we need a definition and a concrete
decomposition of sustainability to make it tangible for software systems devel-
opment.
In addition, due to second and third order effects, it is not sufficient to
analyse environmental sustainbility on its own, but its interplay with economic,
social, human and technical sustainability in order to define appropriate actions
and understand their effects.
Principal idea: We analyse the dimensions of sustainability, their values with

respective indicators, and activities to support them. These elements are com-
posed to a conceptual model that allows for analysis as well as constructing an
outline of actions both for a company or a product point of view.
Contribution: We propose a generic sustainability model with a meta model

and with illustrating instances for companies and projects that are taken from
various case studies. By building an tangible model we make these effects ex-
plicit and thus enable analysis, support and assessment of environmental sus-
tainability in software engineering in a broad view.
Keywords: sustainability, software system, goal, requirements, definition

Contents
1 Introduction & Motivation 3
2 Related Work 5
2.1 Sustainability Modeling Approaches . . . . . . . . . . . . . . . . 5
2.2 Economic Domain: Balanced Score Card . . . . . . . . . . . . . . 6
2.3 Definitions of Sustainability . . . . . . . . . . . . . . . . . . . . . 7
2.3.1 Frameworks for Defining Sustainability . . . . . . . . . . . 7
2.3.2 Defining Sustainability in the Context of Software Engi-
neering . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.4 Analyses: Relation between ICT and Sustainability . . . . . . . . 8
3 Dimensions of Sustainability in Software Engineering 10

3.1 Dimensions of Sustainability . . . . . . . . . . . . . . . . . . . . . 10
3.2 Relations between the Dimensions of Sustainability . . . . . . . . 11
3.3 Resulting Central Questions for Analyzing Sustainability in Soft-
ware Engineering . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
3.4 The Concept of Value . . . . . . . . . . . . . . . . . . . . . . . . 13
4 Stakeholders for Sustainability 14

4.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
4.2 Overview of General Stakeholder Classes . . . . . . . . . . . . . . 15
4.3 Stakeholders according to Sustainability Dimensions . . . . . . . 15
5 The Generic Sustainability Model 18

5.1 The Meta Model . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
5.2 The Generic Sustainability Model . . . . . . . . . . . . . . . . . . 20
5.2.1 Dimensions . . . . . . . . . . . . . . . . . . . . . . . . . . 21
5.2.2 Values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
5.2.3 Indicator . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
5.2.4 Regulation . . . . . . . . . . . . . . . . . . . . . . . . . . 21
5.2.5 Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
5.3 The Generic Sustainability Model Database . . . . . . . . . . . . 22
5.3.1 Environmental Sustainability . . . . . . . . . . . . . . . . 22
5.3.2 Human Sustainability . . . . . . . . . . . . . . . . . . . . 23
1
5.3.3 Social Sustainability . . . . . . . . . . . . . . . . . . . . . 24
5.3.4 Economic Sustainability . . . . . . . . . . . . . . . . . . . 24
5.3.5 Technical Sustainability . . . . . . . . . . . . . . . . . . . 26
5.4 First Validation via Common Positions on Sustainability . . . . . 27
6 Process: Instantiation of the Generic Model 28

6.1 Roles during Instantiation of the Generic Sustainabilty Model . . 28
6.2 Modelling Process . . . . . . . . . . . . . . . . . . . . . . . . . . 30
6.2.1 Analysis Phase . . . . . . . . . . . . . . . . . . . . . . . . 30
6.2.2 Application & Assessment Phase . . . . . . . . . . . . . . 32
6.3 Example: Scenario Smart Mobility . . . . . . . . . . . . . . . . . 32
7 Application: Instances of the Sustainability Model 36

7.1 Product-Specific Instance @BMW . . . . . . . . . . . . . . . . . 36
7.1.1 Car Sharing Platform: Stakeholders . . . . . . . . . . . . 36
7.1.2 Instantiation of the Sustainability Model . . . . . . . . . . 39
7.2 Company-Specific Instance @jambit . . . . . . . . . . . . . . . . 39
7.3 Requirements Engineering Conference 2013 . . . . . . . . . . . . 42
7.4 Assessment: Evaluation and Discussion . . . . . . . . . . . . . . 46
7.4.1 Benefits . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
7.4.2 Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
7.4.3 Threats to Validity . . . . . . . . . . . . . . . . . . . . . . 46
8 Conclusion 47
8.1 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
8.2 Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
2
Chapter 1
Introduction & Motivation
Sustainability has been recognized as a relevant topic in software engineering,

e.g., by the 2012 ICSE theme and a number of workshops. It refers to environ-
mental, social, and economic aspects of software development and the usage of
software systems.
When looking at software systems and these aspects of sustainability, we find
that most software systems usually are developed with an economic benefit in
mind and that this benefit shall also last over a longer period of time. In case of
software systems that are not intended to serve economic purposes, for example,
open source software, the main purpose of the software system is a social benefit
which can be making specific work tasks easier or just plain entertainment and
having fun.
Taking into account that the economic and social aspects of sustainability
are already present in most software systems, our work especially focusses on
supporting the environmental perspective. Our proposed approach shall help
integrating the objective of environmental sustainability in and via software
systems with other system development objectives.
Hilty et al. [1] provide an analysis of the relevance of information and commu-
nication technologies for environmental sustainability and conclude that there
is no such thing as a “general ICT policy for environmental sustainability” [1].
Especially for an adequate analysis of the latter, we need a tangible decom-
position of the concept of sustainability and supporting methods in requirements
engineering. These methods can then enable to include the concept of sustain-
ability during requirements engineering and help to develop such a general ICT
policy for environmental sustainability over time.
Problem Currently, we are still missing a definition of what software engi-

neering can contribute to improving the sustainability of the systems under
development (except for some aspects of Green IT). Thus, sustainability is not
tangible enough as a concept to actually be able to transform it into software
requirements.
3
There is no guidance available describing how to decompose sustainability for
a concrete project and we are missing procedures for incorporating sustainability
as an explicit goal into requirements engineering.
Apart from the decomposition, we are also lacking a reference catalogue
of sustainability-improving activities linked to indicators they affect. These
activities and indicators would allow for a sustainability assessment in order to
evaluate whether the activities affect the sustainability of a project or company.
Contribution In this report, we present a reference model for sustainability

that decomposes sustainability into five dimensions: environmental, human,
social, economic, and technical sustainability. The model provides a number of
activities and relates them to the values they support and the indicators they
can be assessed with.
It is intended to serve as a reference model for a process engineer who in-
stantiates the model for a software development company or for a requirements
engineer who instantiates it for a specific system under development. They in-
stantiate the values in the model with system-specific goals and derive specific
actions from the activities, thus accounting for the realisation of the system’s
(or company’s) sustainability goals. The instantiation is illustrated by two ex-
amples taken out of current case studies.
A specific aim is to show how supporting environmental sustainability can be
aligned with the other sustainability dimensions, as environmental sustainability
is the aspect that is least supported by our current ways of developing software
systems.
Outline The report is organized as follows: Chapter 2 discusses related work

and positions the contribution. Chapter 3 presents our understanding of sus-
tainability and how it builds on and relates to existing definitions. Chapter 4
describes the stakeholders for the different aspects of sustainability and their
potential roles. Chapter 5 presents the generic sustainability model and its
meta model. Chapter 6 describes the process of how to instantiate the models.
Chapter 7 presents the application of the instantiation in a number of examples
and case studies. Chapter 8 draws conclusions and points out future work.
4
Chapter 2
Related Work
Related work can be differentiated into sustainability modeling approaches,

modeling approaches from the economic domain, definitions of sustainability,
and analysis of sustainability in specific domains.
2.1 Sustainability Modeling Approaches

Sustainability is starting to be acknowledged in software engineering, with the
Journal on Sustainable Computing, the conferences on Sustainable Comput-
ing, the conference series EnviroInfo, the 2010 Workshop on Services, Energy
and Ecosystems, the 2009, 2010, and 2011 Workshops on Software Research
and Climate Change, the 2012 Workshop on Green and Sustainable Software
(GREENS), and the 2012 Workshop on Requirements Engineering for Sustain-
able Systems (RE4SuSy) at REFSQ’12.
Framework for sustainable software engineering Naumann et al. [2] pro-

vide a framework for sustainable software engineering. They investigate how web
pages can be developed with little environmental impact, i.e., energy-efficiently,
and offer a respective guideline for web developers. In such a framework, the
reference model proposed in the work at hand could be used as key element in
order to better promote the method.
Strategy models Gu et al. [3] propose a green strategy model that pro-
vides decision makers with the information needed to decide on whether to
take “green” strategies and eventually how to align them with their business
strategies. They consider such strategies as “green” which achieve lower energy
consumption and perform a case study with Dutch data centers. In contrast,
the paper at hand considers a broader definition of sustainability and thus gives
a broader view on sustainable software engineering.
5
i-star modeling Cabot et al. [4] report on a case study for sustainability as
a goal for the organization of the ICSE’09 conference with i*-models to sup-
port decision making for future conference chairs. Stefan et al. [5] extend that
work for managing environmental sustainability with quantitative goal modeling
techniques. Both works provide model instances for specific case studies while
the work at hand also provides a generic reference model.
Requirements engineering techniques Mahaux et al. [6] present a case

study on a business information system for an event management agency that
advertises environment-friendly events. They assessed how well some current
RE techniques support modeling of specific sustainability requirements in that
case study. In contrast, our aim is to provide modelling means explicitly for
integrating sustainability into the software development process as a major ob-
jective.
Goal Modeling Lamsweerde [7] decomposes business goals into system re-
quirements, but does not explicitly reflect on sustainability. His work relates
positive and negative influences between goals, but our work provides activities
for direct realisation.
2.2 Economic Domain: Balanced Score Card

One exemplary approach from the economic domain that can help to opera-
tionalize and monitor sustainability in a company is the Balanced Score Card
(BSC) by Kaplan and Norton’92 [8]. They combine financial and operational
measures. Their used perspectives are:
• Customer: How do customers see us?
• Internal: What must we excel at?

• Innovation and learning: Can we continue to improve and create value?
• Financial: How do we look to shareholders? [8, p. 4]
Matching of BSC perspectives to SRM dimensions as instantiation

One major goal of the instantiations of the BSC is the monitoring and improve-
ment of the financial prosperity of the implementing company. The major goal
of the instantiation of the GSM for a company is to monitor and improve the
sustainability of that company.
The balanced score card (BSC) can be seen as instantation of the generic
sustainability model (GSM). The goals identified in the BSC are thereby in-
stances for the values of the GSM. The goals do not necessarily refine the value
into a more specific goal but they may do so.
6
2.3 Definitions of Sustainability
The background for the presented definition approach is structured into frame-
works for defining sustainability, the relation between ICT and sustainability,
and earlier approaches to defining sustainability in the context of software en-
gineering.
2.3.1 Frameworks for Defining Sustainability

The following frameworks are designed to serve for defining sustainability in
general, without a specific reference to software systems. We present them
to emphasize the necessity of adequate definitions and use them as context to
discuss our proposed definitions.
Dobson [9] suggests a framework for the comparison of sustainability notions
that poses conceptual questions, which any conception of sustainability must
answer: What is to be sustained? Why is it to be sustained? Who/what is
concerned? In what respect is substitutability allowed? This is sufficient for a
topology for comparison but avoids formality.
Burger and Christen [10] propose a capability approach of sustainability. To
avoid “yet another sustainability conception”, their methodological approach
is to first formulate adequacy conditions for concepts of sustainability, then
illustrate a categorial framework with the required general concepts, and finally
propose a conception of sustainability based on the capability approach. They
consider the idea of sustainable development as a problem-solving strategy (for
the “developmental dilemma” [11]1 ) and their adequacy conditions are future-
orientation, normative power, justice, universality, limited natural conditions,
and high-level strategic actions [10, p. 788]. The developed approach is supposed
to serve as basis for empirical research in development studies.
Robèrt et al. [12] present the Framework for Strategic Sustainable Develop-
ment that is now mainly promoted by the Swedish NGO ‘The Natural Step’2 .
Their objective is to show how the increasing number of tools and approaches
to develop sustainability (e.g., the ISO14001 [13], Life Cycle Assessment [14],
and Ecological Footprinting [15]) relate to each other and build on each other
when used for planning for sustainability. The framework consists of five inter-
connected levels: the understanding of constitutional principles of the system,
principles of sustainability, principles for sustainable development, activities for
sustainable development, and tools for monitoring compliance and impacts of
the actions. Thereby, it reflects sustainability from the foundation to the prac-
tice. Following a constructive rather than an analytic approach, this framework
also avoids formality but is intended as guideline.
Following a stricly analytic approach in a theoretic contribution to the dis-
course in sustainable development, Christen and Schmidt [16] counter-propose
1 Developmental Dilemma: Millions of people are in need of societal and economic devel-
opment to surpass the poverty line, however this development could overstrain the natural
conditions necessary to guarantee further development. [11]
2 http://www.naturalstep.org/
7
(w.r.t. the approach by Robèrt et al. [12]) a Formal Framework for Concep-
tions of Sustainability that intends to solve the problem of arbitrariness. The
framework consists of five modules that serve to help to elaborate the elements
answering the ‘what’ and ‘how’ questions. The modules are the sustainability
problem, the normative principle of justice, the descriptive principle of integra-
tion, the criteria for sustainability, and the transformation into practice.
2.3.2 Defining Sustainability in the Context of Software

Engineering
Penzenstadler et al. [17] provide a systematic literature review that points to
two definitions of sustainability in the context of software engineering, i.e. by
Mahaux et al. [6] and by Naumann et al. [2].
Mahaux et al. base their research on the Brundtland definition [11] plus
the statement that IT changes behavior and therefore has considerable effect on
society and environment, which is supported by greenIT concepts and analysis
of the usage processes of a software system [6].
Naumann et al. define “green and sustainable software” such that “direct
and indirect negative impacts on economy, society, human beings, and envi-
ronment that result from development, deployment, and usage of the software
(...) ha(ve) a positive effect on sustainable development”, and “Green and Sus-
tainable Software Engineering” such that the “negative and positive impacts on
sustainable development (...) are continuously assessed, documented, and used
for a further optimization of the software product” [2, p. 296].
These two works already show that the term sustainability is strongly de-
pendent on the taken perspective. Even the second definition cannot yet be
operationalised, which is crucial if SE researchers want to contribute.
Consequently, scope and context have to be clearly defined to be able to make
any statements. A first approach is taken from systems thinking [18] by using
three general parameters: Which system shall be preserved in which function
over which time horizon?
2.4 Analyses: Relation between ICT and Sus-

tainability
Hilty, Lohmann, and Huang [19] provide an overview of the fields of ICT in the
service of sustainability: Environmental Informatics, Green IT, and Sustain-
able Human-Computer Interaction. As technological efficiency alone will not
produce sustainability (cf. Jevon’s paradoxon [20]), they state that sustainable
development requires a combination of efficiency and sufficiency strategies, inter
alia by decoupling economic growth from environmental impacts and from the
use of natural resources.
Furthermore, Hilty, Arnfalk, Erdmann et al. [1] analyse the relevance of infor-
mation and communication technologies for environmental sustainability. They
8
present the impacts of ICT on environmental sustainability on different levels:
first order effects like increasing electronic waste streams, second order effects
such as improved energy-efficiency of production, and third order effects like
a product-to-service shift or rebound effects in transport. In their prospective
study, they present an analysis of the anticipated future (positive or negative)
environmental impacts of different types of ICT applications: ICT applications
supporting a product-to-service shift, for heating management, for passenger
transport efficiency, for mobile work, and for freight transport efficiency.
With a wider scope, the Smart 2020 report [21] by the Climate Group
presents a way of enabling the low carbon economy in the information age.
The book Vision 2050 [22] by Hiroshi Komiyama and Steven Kraines presents
a roadmap for a sustainable earth.
9
Chapter 3
Dimensions of
Sustainability in Software
Engineering
This chapter presents the dimensions of sustainability that serve to identify the
stakeholder for sustainability in Chapter 4 and represent the dominant decom-
position means for the generic sustainability model in Chapter 5. Furthermore,
this chapter discusses their relations, and introduces the concept of value that
is also a building block for the generic sustainability model.
3.1 Dimensions of Sustainability

We consider five dimensions of sustainability as important for the analysis of
software systems. The basic dimensions for a general sustainability analysis
(without referring to software systems) are human, social, economic and envi-
ronmental as defined by Goodland [23]. The three latter ones are the dimensions
also known from the UN definition of sustainable development [24].
But these four dimensions do not offer a possibility to claim and support
long-term evolution of technical systems and adequacy for long-term use. Con-
sequently, when looking at (software) systems, we need technical sustainability
as an additional dimension.
Human sustainability Human sustainability refers to the maintenance of

the private good of individual human capital. The health, education, skills,
knowledge, leadership and access to services constitute human capital [23].
Social sustainability Social sustainability means maintaining social capital

and preserving the societal communities in their solidarity. Social capital is
investments and services that create the basic framework for society [23].
10
Economic sustainability Economic sustainability aims at maintaining as-
sets. Assets do not only include capital but also added value. This requires
to define income as the amount one can consume during a period and still be
as well off at the end of the period, as it devolves on consuming added value
(interest), rather than capital [23].
Environmental sustainability Environmental sustainability seeks to im-

prove human welfare by protecting natural resources. These are water, land,
air, minerals and ecosystem services; hence much is converted to manufactured
or economic capital. Environment includes the sources of raw materials used
for human needs, and ensuring that sink capacities recycling human wastes are
not exceeded [23].
Technical sustainability From a point of view of (software) systems engi-

neering, there is another dimension that has to be considered. Technical sustain-
ability has the central objective of long-time usage of systems and their adequate
evolution with changing surrounding conditions and respective requirements.
These five dimensions are used in the sustainability model to decompose sus-
tainability into more tangible units.
The dimensions Social, Economic, Environmental and Technical can be an-
alyzed on a mirco as well as on a macro level. The decision on which level they
are investigated depends on the scope of the system under analysis.
These five dimensions are not necessarily encompassing. One can argue,
for example, for considering politics & law as separate dimension instead of
including it in the social dimension. On the other hand, the government can
be considered as institutionalization of society and therefore can be treated as
subdimension of the social dimension.
However, for the analysis scope of software systems, we believe the five given
dimensions to be an adequate representation as they serve only as structuring
means for the to-be-derived model.
3.2 Relations between the Dimensions of Sus-

tainability
Naturally, the dimensions are strongly related to each other. Figure 3.1 depicts
the dimensions as layers for how the dimensions generally build upon the oth-
ers. This is one of various possible visualizations and is rather meant to be
interpreted as a general idea than as an apodictic explanation.
Other common visualizations depict the dimensions as concentric or inter-
secting circles, either of which serves to strengthen a particular aspect in discus-
sion. However, for our purpose of explaining which dimension of sustainability
is a precondition for another one, the layered visualization is the most suitable
one.
11
Economic Sustainability Technical Sustainability
Social Sustainability
Human Sustainability
Environmental Sustainability
Figure 3.1: Dimension Relations
Environmental S. — Human S. Environmental sustainability is the basis

for all dimensions. Without natural resources, life is impossible. A healthy
human is also part of environmental sustainability. Therefore, environmental
sustainability is the basis for human sustainability.
Human S. — Social S. In turn, humans are the individual elements that

society is based on. Therefore, human sustainability is the basis for social sus-
tainability. On the other hand, humans define themselves strongly by their
relations to other humans, so human sustainability can hardly be regarded in-
dependently of a social context.
Social S. — Economic S. Society is organized by individual humans, and

institutionalized by means of a government and its infrastructure. This infras-
tructure also enables economy. Therefore, social sustainability is a prerequisite
for economic sustainability. On the other hand, economy again interacts with
society by providing wages for its employees, which is fed back into community
by supporting family and infrastructure, and by establishing small communities
of colleagues and customers.
Economic S. — Technical S. Technical systems are developed within a

regulated economical context. At the same time, technical sustainability, the
longevity of systems and services, is a fundamental element of the economy
in terms of goods and services that are traded in that economy. Therefore,
economic and technical sustainability depend on each other to a certain extent.
Economy motivates research and development and vice versa.
3.3 Resulting Central Questions for Analyzing

Sustainability in Software Engineering
Splitting up the matter of sustainability into the five dimensions mentioned
above leads to five software related questions to be answered by research, society
and each individual project:
12
Environmental Sustainability How does software affect the environment dur-
ing development and maintenance [25]?
Human Sustainability How can software be created and maintained in a way
that enables developers to do their job happily over a long range of time?
Social Sustainability Which effects do software systems have on the society

(e.g. communication, interaction, government, . . . )?
Economic Sustainability How can software systems be created so that the
stakeholders’ long term investments are as safe as possible from economic
risks?
Technical Sustainability How can software be created so that it can easily

adapt to future change?
These questions are meant to be starters for further discussion and research.
They are questions that build upon the definitions of the sustainability domains
and are based on the values we consider important for the domain.
3.4 The Concept of Value

One central question laying the basis for the analysis of sustainability support is:
What is the value notion behind sustainability? Is the notion nature-centered
or human-centered?
Plato According to Plato, objects that are attributed values can be distin-
guished into objects with intrinsic value and objects with instrumental value.
An intrinsically valuable thing is worth having for itself. [26]
In the context of our research, we understand a value as intrinsic, indepen-
dent of a specific justification.
Kant A value is a reason by which an action or state is judged, which is only

justified in itself (atomicity). For example when looking at working rights, most
people agree that an employee’s physical and mental health should not be risked
through the work he or she is employed for. Various arguments support this
goal, for example, the fact that a sick person is prevented from working for the
company until health is reestablished. Yet, this argument can be abstracted up
to the company’s goal of long term prosperity. In contrary, another reason why
the individual’s health should not be risked, is a value that is perceived as being
right in most cultures of our time: the physical integrity of the individual. This
value can hardly be abstracted or based in a higher reason, which is why it can
be classified as a value. [27]
In contemporary software companies this issue relates, for example, to work-
ing overtime.
13
Chapter 4
Stakeholders for
Sustainability
This chapter details on the stakeholders for sustainability, their interests, and
their relation to the sustainability dimensions introduced in the preceeding chap-
ter.
4.1 Motivation
When discussing sustainability in the IT context very often people ask: Who
are the stakeholders1 for sustainability? This question is motivated from two
backgrounds:
1. Domain-Knowledge: The stakeholders of a certain domain usually has
widespread knowledge in this field. This leads to the point that the stake-
holder can judge whether or not a certain indicator is a good approx-
imation for a value, if suggested activies can be put in place, how the
activities relate to the indicators, etc. For example, when discussing the
impact of establishing use of green data centers, software developers, fi-
nance and advertising stakeholders should discuss the costs and benefits
of this avtivity.
2. Stakeholder-Support: Besides the knowledge of the stakeholders, it is a

well known fact in industry that many projects fail because of missing
support from management stakeholders. The same holds for all other
stakeholders in the company. For example, if the developer understands
the rational behind a the decision to use a green data center, he will
be more willing to support the idea, which will decrease the risk for the
project.
1 By a stakeholder we refer to anyone who has interest in or is affected by the system under
consideration.
14
4.2 Overview of General Stakeholder Classes
Values differ by the community under analysis. When facing Sustainability in
Software Engineering we usually refer to this community in terms of stakehold-
ers, thereby including customers, users, developers and anybody else who has
an interest in the system under analysis.
Exemplary classes of stakeholders in software engineering for the sustain-
ability dimensions introduced in Chapter 3 are:
• Management: Managers concerns are represented primarily the economic

dimension because their major task is to preserve and improve the value
of a company’s shareholders. As employees they are also subjects to the
human and social dimension.
• Employees / Developers: Concerns of employees or developers are primar-
ily represented in the social dimension in terms of the need for making a
living and desiring a pleasant environment, but they are also subjects to
the human dimension with basic needs like health and nutrition.
• Suppliers: Subcontractors are necessary for raw material, hardware and
software, therefore they influence primarily the economic and technical
dimensions.
• Legal Representatives: These stakeholders’ concerns influence all dimen-
sions of a system under analysis, as there are laws and regulations for
human rights, environment and society as well as economy and technol-
ogy.
• Customers: The clients influence primarily the economic and technical
dimensions as they want to get value for their investment and want that
value to be preserved.
• Users: The actual users of a system influence primarily the technical, but
also the social dimension as they want the systems to contribute to some
aspect of their life but generally will want them to last.
• Destructors: Companies occupied with recycling or disposing waste mainly
take influence on the environmental dimension, but also on the economic
one in case this affects costs for the customers or users.
These classes are not necessarily encompassing but intended to serve as

checklist when analyzing the stakeholders for a particular system.
4.3 Stakeholders according to Sustainability Di-

mensions
For the elicitation of stakeholders once again the generic sustainability model can
serve as a guideline. This is performed by going through the five sustainability
15
dimensions and understanding which stakeholders are possibly affected by this
dimension. In the following we will give a generic, very reusable list for possible
stakeholders. This list is just a rough overview and in no means complete.
Figure 4.1: Generic Stakeholders of the Sustainability Model.
• Economic Sustainability:
– CEO: The chief executive officer (CEO) is the highest-ranking execu-
tive of a company. The CEO has to integrate the sustainabilty goals
into the company’s vision.
– Project manager: Especially when instantiating a product-specific
sustainability model, it is important to have the project manager
agree in what ways the project should support sustainble aspects.
– Finance responsible: As sustainable software engineering affects many
parts of the company, among others, many financial decisions have
to be made to implement a sustainable software engineering model
in a company.
• Technical Sustainability
– Admin: The administrator of a software system has a very strong
motivation for long-running, low-maintenance systems as it makes
his work easier.
– Maintenance
– Customer/User: Users are often interested in persistency of the sys-
tems they are using []. This regards to user interface and required
16
soft- and hardware. Many users, however, are interested in gaining
features with updates.
• Human Sustainability
– Customer/User: The user must be able to use a certain system.
– Developer: The developer is heavily involved in creating the system.
This process needs to be careful about sustainable pace and growth
of the developer.
– Employee representative: The mental and physical safety of individ-
uals during the development of a product needs to be maintained.
Employee representatives watch the rights of the employees involved.
– Legislation (individual rights): Systems must respect the right of
their users. Hence, a legislation representative must respresent the
privacy and data protection laws that are in place.
• Social Sustainability
– Legislation (state authority): The state has a strong interest in un-
derstanding a systems influence on the society. In contrary to the
individual rights legislation representative, the state authority repre-
sentative speaks from the perspective of the state as a whole.
– Community representative: In addition to the state authority, other
communities such as the local government (e.g. the mayor) or non-
government clubs might be affected by a software system. A complete
analysis must take their views into account.
– Customer Relationship Manager (CRM): The CRM wants to estab-
lish long-term relationships with their customers and create a positive
image for the company.
– Corporate Social Responsibiliy (CSR) manager: Some companies
created a special position that initiates and implements a company-
specific vision of its social responsibility.
• Environmental Sustainability
– Legislation (state authority): Environment protection laws are in
place to ensure sustainability goals. These laws must be reflected in
the model.
– CSR manager: The CSR manager is often also responsible for envi-
ronmental aspects.
– Nature conservation activists and lobbyists (e.g., WWF, Greenpeace,
BUND)
Again, many stakeholders can be reused through various projects. However,

in order to be of greater use, this list needs to be instantiated according to
project or company needs.
17
Chapter 5
The Generic Sustainability

Model
This chapter describes the generic sustainability model built upon the concepts
introduced in the previous chapters.
Our proposed approach comprises the generic sustainability reference model
(M1 level1 ), the respective meta model behind it (M0 level), and the instances
(M2 level) that are derived from the reference model. These can be derived
either for the development process in a specific software company or for a specific
software system or product.
5.1 The Meta Model

The meta model is comprised by the types Dimension, Value, Indicator, Reg-
ulation, and Activity. The model including relations between the elements is
depicted in Figure 5.1.
Type <Dimension> A dimension is an aspect of or viewpoint on sustain-
ability, for example, “environmental sustainability”. A dimension is rep-
resented by a set of values that express the abstract objectives of the
dimension.
Type <Value> A value is a rationale that is rooted in itself, and is approxi-
mated by indicators.
Type <Indicator> An indicator is a qualitative or quantitative metric and is
related to a value.
Type <Regulation> A regulation is an optional element belonging to a value.
Type <Activity> An activity is related to values that it supports and poten-
tially improves.
1 similar to the OMG’s Meta Object Facility’s levels [28]
18
*
influence
1..*
1 is aspect of 1..*
Goal Dimension
1..*
represents
1..* *
subvalue
1..* supports 1..* 1

Regulation Value
1..*
1..*
1..*
affects approximates contributes to
1..*
1..* 1..* *
subactivity
1..* influences 1..* 1

Indicator Activity
Figure 5.1: Meta model of generic sustainability model (M0).
19
Hypotheses The following hypotheses have been made for the meta model:
1. A dimension can be represented by a set of values.
2. A value can always be approximated by indicators. They may be quanti-
tative or qualitative.
3. Qualitative and quantitative indicators cannot be aggregated to find an

absolute measure for a value in case it is approximated by a set of indica-
tors instead of one single value.
5.2 The Generic Sustainability Model

An excerpt of the generic sustainability model is provided in Fig. 5.2. The
model consists of three levels: the top level contains the dimensions; the middle
level contains values, indicators, and regulations; and the lower level contains
activities. Each element in the generic sustainability model is an element of a
type from the meta-model explained before. In the following we will explain
examples from the generic model, structured by their type.
<Goal>
Sustainability
<Dimension> <Dimension> <Dimension> <Dimension> <Dimension>

Economic Technical Environmental Human Social
Sustainability Sustainability Sustainability Sustainability Sustainability
<Value> ... ... <Value> ... <Value> ... <Value> ...

healthy <Value>
long-term <Value> human human
<Value> environment community Legend
profit maintain health capital
long-term <Value> Contributes
ability use conservation Influences
of biodiversity Supports
<Value>
education
<Regulation> <Value> Goal
<Indicator>
human rights trust
continuous
ROI Dimension
<Regulation>
millenium goals
Value
Regulation
<Activity>
<Activity> optimize <Activity>
resource Activity
long-term Restoration
strategy consumption
<Activity> <Activity> <Activity>

use sustainable <Activity> <Activity> <Activity> knowledge education <Activity>
resources reduce reuse recycle management programs mentoring
exemplary and incomplete
Figure 5.2: Exemplary excerpt of the generic sustainability model (M1).
20
5.2.1 Dimensions
A dimension is an aspect of or viewpoint on sustainability, for example, envi-
ronmental sustainability. As described in Sec. 3.1, there are five dimensions to
be considered when analyzing sustainability for software systems engineering.
A dimension is detailed in a set of values.
5.2.2 Values
A value is a moral or natural good that is perceived as an expression of a specific
dimension. Each of the five dimensions is represented by a set of values. Values
do not necessarily belong exclusively to one dimension but can be considered
for a number of dimensions, for example, healthy environment, which applies for
both the environmental as well as the human dimension.
5.2.3 Indicator
An indicator is a qualitative or quantitative metric that expresses a specific
degree or score with regard to a value, for example, satisfaction indices as
qualitative metric and carbon emissions or return on investment as quantitative
metrics. A set of indicators thereby approximates a value.
For the indicators, there are catalogues by, e.g., Bell and Morse [29] and the
ESI [30].
5.2.4 Regulation
A regulation is an optional element that supports and/or enforces a value, for
example, emission regulations. Regulations commonly set limits for a specific
indicator to be of legal use.
Many values belonging to the different dimensions are heavily regulated,
either supported or restricted in order to protect them. For example, freedom
of the individual is supported by the human rights, and healthy air is supported
by the European Union’s directive on carbon emissions.
5.2.5 Activity
An activity is a measure taken to contribute to a specific value or a set of values,
for example, use train for mid-distance traveling instead of aircraft. The impact
of these activities on a value is measured by the indicators it influences. For
the travel example, using a train instead of an aircraft improves the emissions
account of the traveler. For each value, there is a number of activities that
can be implemented to support a value. Thereby, the impact of an activity is
measured by the indicators it influences.
Whenever we say “good” in activities, the degree can be measured by the
indicators.
21
The generic sustainability model is intended to serve as reference and as a
basis for the instantiation of company- or system-specific instances.
5.3 The Generic Sustainability Model Database

An encompassing generic sustainability model requires a database. We have
decided for a simple database that is extended over time. Figures 5.3 - 5.7
illustrate excerpts of that database for each sustainability dimension.
5.3.1 Environmental Sustainability
<Dimension>
Environmental
Sustainability
<Value> <Value>
environmental preservation
health of diversity
<Value> <Value> <Value> <Value> <Value>

<Value> preserve clean balanced diversity diversity
clean air scarce water ecosystem of fauna of flora
resources <Indicator>
animal
population
<Regulation>
emission law <Indicator>
plant
population
<Activity>
save
resources <Activity> <Activity>
conservation restoration
<Activity> <Activity>
reduce reuse <Activity>
<Activity> recycle
reduce
<Activity>
emissions
Place&computer&centers&at&
places&where&they&can&be& <Activity> <Activity> <Activity> <Activity>
powered&by&renewable& reduce reduce energy reuse old recycle
energy&sources. waste consumption hardware packaging
Figure 5.3: Environmental dimension of the sustainability model (M1).
The model excerpt for environmental sustainability is depicted in Figure 5.3.

It is structured in the same way and with the same elements as Figure 5.2.
Below the dimension, there is a number of values that further decompose
the values of environmental health and preservation of diversity. One of the
sub-values, preserve scarce resources is realised by the activity save resources,
which is again stepwise refined into the activity reduce and reduce emissions.
22
One option for realising feewr emissions is the activity place computer centers
where they can be powered by renewable energy sources., for example, to place
data centers in islands that have a high geothermal activity, e.g., iceland. Pow-
ering data centers with geothermal energy on iceland reduces the production of
emissions.
5.3.2 Human Sustainability
<Dimension>
Human
Sustainability <Value>
<Value> curiosity
<Value>
human <Value>
<Value> freedom
health <Value> <Value> self-
human <Value> <Value>
acceptance development empower
capital happiness dignity
/ respect / growth ment
<Value>
mental <Value>
health education
<Regulation>
<Value> <Value> human rights
<Value>
physical creativity safety
health
<Value>
sustainable
pace
<Activity> <Activity> <Activity> <Activity>

health service educational create provide
recreation provide
& support programs slack time democratic literature
<Activity> <Activity> hierarchies
realistic time creating
management <Activity>
<Activity>
<Activity> provide
<Activity> mentoring
socializing feedback
<Activity> active
dialogue opportunities
workout / sports <Activity>
<Activity> protect safety guarantee free
work-life
good nutrition of individual expression of
balance
opinion
Figure 5.4: Human dimension of the sustainability model (M1).
The model excerpt for human sustainability is depicted in Figure 5.4. Here,
one value is happiness, with the sub-value creativity.
The implementing activity creating can, for example when instantiating the
model for a software company, be translated into the creativity in the developing
processes of the software engineers. If they experience a certain degree of cre-
ative freedom in their work, this is likely to increase their happiness. This also
has side effects on their productivity and thereby on the economic dimension,
but this influence is not depicted in the figure.
23
5.3.3 Social Sustainability
<Dimension>
Social
Sustainability
<Value> <Value> <Value> <Value> <Value> <Value>

community tolerance trust fairness peace culture
<Value> <Value> <Value> <Value> <Value>

<Value>
justice equality sustainable social progress of
openness
government peace knowledge
<Value> <Value>
political expression
<Value> <Value> of thoughts
<Regulation> peace
honesty transparency and emotion
constitution
<Value>
preservation
<Value>
of cultural
continuity
inheritage
<Activity> provide for equal science
spend time reveal wages <Activity> <Activity>
work on good chances
together in create long- arts,
communication <Activity>
intense <Activity> term theater,
execute legal preserve
situations perspectives music
actions traditions
communication <Activity> university
rules <Activity> perform collabor-
prosecute appraisal ations
rethoric team building
seminar events mobbing interviews
Figure 5.5: Social dimension of the sustainability model (M1).
Figure 5.5 depicts the excerpt for social sustainability. Some of the central
values are tolerance, trust, fairness, and culture. Trust can, for example, be
built up during team events where the members spend time together in intense
situations as the respective activity suggests.
Tolerance requires openness, which can be supported by work on good com-
munication. One action to improve communication in a software development
company can be to set up rules for positive, open, and efficient communication.
Social sustainability in a company might best be supported via a company
culture that builds on the above named values.
5.3.4 Economic Sustainability

Figure 5.6 depicts the excerpt for economic sustainability. Values are security,
fairness, safety, trust, and long-term profit. The latter is realised by the activites
good infrastructure and long-term strategy.
As depicted on the left of igure 5.6, such a long-term strategy can be sup-
ported by investing in academia, for example, in bilateral collaboration projects.
24
<Dimension>
Economic
Sustainability
<Value>
<Value> <Value> <Value> <Value>
long-term
security fairness safety trust
profit
<Indicator> <Value>
continuous <Value> <Value> <Value>
<Regulation> reliability
ROI financial job image
workers' rights
security security
<Value>
<Regulation> transparency
competition law
long-term good
strategy infrastructure <Activity> <Activity> <Activity>
long-term <Activity> communicate image
job strategy fair trade honestly creation
<Activity>
invest in <Activity> <Activity>
<Activity> establish
society technology <Activity> pay fair good
risk wages working <Activity>
<Activity> analysis
invest in <Activity> conditions invest in
academia location charity

<Activity> offer partial
<Activity> low risk transparent
invest in retirement
networking investment risk response
startups
Figure 5.6: Economic dimension of the sustainability model (M1).
25
Another possibility that brings future prospects, but might also have a cross-
influence on risk management, is investing in startups.
On the right hand side of Figure 5.6, the value of trust is decomposed into
the subvalues reliability, transparency, and image, where the latter is realised
via image creation, for example, be investments in charity.
5.3.5 Technical Sustainability
<Dimension>
Technical
Sustainability
<Value> <Value> <Value> <Value> <Value> <Value>

maintainability adaptability reliability security portability usability
<Regulation>
<Indicator> standards
quality of
<Value>
documentation
long-term
orientation
<Activity>
<Activity>
risk
"good"
analysis
software
engineering
<Activity> practice <Activity>
document design
properly <Activity>
establish good user
<Activity> interfaces
<Activity> good
safety
clean code architecture
analysis
design
Figure 5.7: Technical dimension of the sustainability model (M1).
The excerpt for the technical dimension of sustainability is depicted in Fig-

ure 5.7. The central values are traditional quality characteristics, for example,
maintainability, adaptability, reliability, security, portability, and usability.
Many of these values can be realised and supported by what might be called
‘good software engineering’, e.g., by writing clean code, performing safety anal-
yses, establishing good architecture design, and good documentation. Thereby,
good documentation is not the most extensive one, but a high quality documen-
tation that serves the task of being able to recapitulate why decisions have been
taken as they have.
Especially usability has become a crucial success factor in the domain of
life-style devices like the range of hand-held devices during the past decade.
26
Therefore, the activity of designing a good user interface will strongly influence
the technical sustainability of such a software system.
5.4 First Validation via Common Positions on

Sustainability
A first indicator of whether the proposed generic model might actually be ben-
eficial is by checking its ability to represent popoular or established positions or
perspectives for sustainability.
Environmental sustainability is the conservation of biodiversity. This

point of view is represented in the model by the <Value> of conservation of
biodiversity.
Technical sustainability is long time usage of systems. This pragmatic

definition is represented within the submodel in Figure 5.7 by the value long-
time orientation and the supporting activities.
Technical sustainability is cheap maintenance of systems. This def-

inition is a combination of the technical and the economic dimension and is
represented within the model by these two dimensions.
Happy Planet Index The happy planet index [31] is a global measure of
sustainable well-being that relates how much resources a country has to how
happy its inhabitants are. This relates the social dimension to the economic
one and can also be modeled in the proposed model.
27
Chapter 6
Process: Instantiation of
the Generic Model
The generic model gives an idea of what sustainability means in general. How-
ever, the concrete ideas of sustainability in a given company or project highly
depend on the values existent in this context. Therefore, it is inevitable to to
evaluate, instantiate and maybe also extend these values with the stakeholders
to create a common sustainability model for a company or a project. These
models may then be used to create discussion for newly forming teams, which
is especially important in cross-cultural projects (see for example [32] on issues
in cross-cultural projects).
As the sustainability model highly depends on team members as well as
project goals, it can be instantiated for development processes (companies) and
for software systems (products). For the former, a sustainability goal model
that represents the goals of the company is created; for the latter, this instan-
tiation happens during the requirements engineering phase of each product.
The method for developing the sustainability model, however, stays the same.
In the following, we describe this method for developing a sustainability goal
model first and give examples for both company-specific and product-specific
instantiations afterwards.
The process consists of two phases: the analysis and the application & assess-
ment phase. Various roles are partaking in developing a specific sustainability
model within these phases.
6.1 Roles during Instantiation of the Generic

Sustainabilty Model
During the analysis phase, the model is instantiated by the Sustainability Ar-
chitect. This refers to the person, who knows most about the company’s or
project’s ideas and goals. Usually this leads to the Business Analyst, the Re-
28
quirements Engineer or the Process Engineer, depending on whether we create
a project or company-specific instance.
For application we need two different roles: For one, the Sustainability Ar-
chitect overviews the set of actions in total. Additionally each action needs an
Sustainability Activity Responsible, who organizes and instruments one concrete
activity.
Lastly, the assessment is performed by a Quality Engineer, who organizes
evaluation and assessment of the indicators. He also reports to the Management
regarding the status quo of the project or company. This life cycle is depicted
in Fig. 6.1.
Analysis is aspe
influence
ct of
is asp
of
influence
is asp
ect is aspect ofect
influence
is aspec
of
t of
influence
Key
represents
ts
represents
ts
re
en
en
pr
res
es
es
en
pr
to rep
... ... ... ... ...
ts
re
ports
appro
Phase
supports
subvalue
alue
subv
contributes to
sup
contributes
xim
con
con
ate
tribute
tribute
supports
s
supports
ences
s to
s to
to
contr
contributes to
tes
influ
ibu
ibu
subactivity vity
tes
contr
subacti
to
Generic
Sustainability Model
Role
is aspe
ct of
pect
is
of is aspec aspe
t of ct of
is aspec
t of
In charge for
is as
influence influence influence influence
represents
ents
represents
ts
re
en
pr
res
es
es
en
repr
rep
ts
... ... ... ... ...
Application
app
ports
e
suppo
subvalue
subvalu
rox
to
sup
ima
contributes
contributes to
con
rts
con
tes
tribute
tribute
supports
supports
s to
s to
ences
con
to
contributes to
tes
& Assessment
influ
trib
ibu
ute
contr
subactivity vity
s to
subacti
Sustainability
Phase Specific Architect

Sustainability Model
Activity
Responsibles Activities Indicators
30"
20"
10"
0"
Quality Engineer 1" 2" 3" 4" 5" Management
Sustainability
Report
Figure 6.1: The life cycle of a specific goal model.
29
6.2 Modelling Process
Applying sustainability modeling within a company usually consists of two al-
ternating phases: the analysis and the application & assessment phase.
Whereas the analysis phase mainly concretizes the definition of sustainability
in the present context, the application & assessment phase starts the selected
activities and supervises whether the activities are working successfully. Even
though these are two separated phases we expect these phases to be alternating,
thus refining and tuning the specific instance of the goal model iteratively over
time.
During the whole process the model guides the tailoring of values, indicators
and activities to the context (company- or project-specific) by structuring the
goals and giving examples and suggestions.
6.2.1 Analysis Phase

In the analysis phase the sustainability architect tailors the generic goal model
to the context (a specific company or software system). This formalizes what
sustainability really means for a certain company or product. As depicted in
Fig. 6.2, the analysis phase includes three steps:
1. For each dimensions the architect needs to instantiate the generic values
that are provided by the generic goal model into context-specific goals.
These goals need to be prioritized to help solving conflicts between poten-
tially contradicting goals.
2. Afterwards the architect defines activities to implement the goals and
defines indicators that make it possible to assess the state of this goal at
the moment as well as in future situations. In this step he might find trade-
offs between indicators that are more exact with respect to this value and
indicators that are easier to apply. For example, the architect might have
to decide whether we want to measure our resource consumption by weight
or by item number (which can probably be automated). The generic goal
model serves as a reference for selecting activities and indicators.
3. When values and indicators for a certain dimension are selected, the ar-
chitect can now relate the activities that have a positive impact on the
indicators. For example, he could find reusable items in this project and
set up policies that enforce this reuse.
Indicators and activities are usually developed iteratively. The engineer can
choose whether it is more adequate for them to start with defining activities or
to list the most important indicators for a specific sustainability assessment.
This way we instantiated the generic model with goals, indicators and ac-
tivities specific to a company or a product.
30
How to instantiate a Sustainability Goal Model Legend
Contributes
<Goal>
Sustainability Influences
... Supports
Goal
for each <Dimension> <Dimension> <Dimension>
Environm. Economic Social ...
dimension
Sustainability Sustainability Sustainability Dimension
...
Value
<Value> <Value> <Value>
instantiate Regulation
company- company- company-
values
specific goal specific goal specific goal
Activity
<Indicator> <Activity>
define
energy agile
indicators
consumption practices
<Activity>
knowledge
select <Activity> management
activities optimize <Activity> - educational offers
resource long-term - best practices
consumption strategy - mentoring
- culture of trust
- building - business plan
- infrastructure - long-term
- equipment investors
- waste
Figure 6.2: How to instantiate the generic model.
31
6.2.2 Application & Assessment Phase
After formalizing the company’s or product’s specific sustainability goal model,
two parallel tasks are needed: First of all, each activity needs a Sustainability
Activity Responsible who is in charge for this concrete activity. Together, the
sustainability architect and the activity responsible’s support and enforce the
implementation of the previously selected activities. Second, the Quality En-
gineer needs to continuously monitor the company’s state with respect to its
sustainability model. He can do this by assessing the list of indicators that are
defined concretely in the model. The results of this assessment are reported to
the management and the sustainability architect, so that changes or adjustments
can be made accordingly. The iteration time between assessments should be as
short as possible to maximize the transparency of the current state. However, it
heavily depends on the automation potential of the company-specific indicators
and the investment the company or project is willing to make.
In summary, the main goal of the application & assessment phase is putting
activities into play and monitoring how the company or product is performing
with regards to its own definition of sustainability.
6.3 Example: Scenario Smart Mobility

Our first example for the instantiation of a goal model is one that was developed
within the publicly funded research project ARAMiS: Automotive, Railway, and
Avionics in Multicore Systems [33, 34], which build on the Agenda CPS, a study
by the German academy of technical sciences (acatech) on the perspectives in
CPS research, development, and application [35].
ARAMiS Project Background One class of systems of systems with the

additional challenge of integrating different system types are cyber-physical
systems (CPS). CPS are integrations of computation and physical processes.
Lee [36] defines CPS’ as embedded computers and networks that monitor and
control physical processes, usually with feedback loops where physical processes
affect computations and vice versa. This leads to complex functionality that
spans a variety of application domains.
The projects duration is 3 years, it has a total budget of 36,5 mio Euro, and
it is partially funded by the German federal ministry of education and research.
Partners are inter alia KIT, AUDI, BMW, Bosch, Continental, Daimler, Air-
bus, EADS, Diehl, Liebherr, Freescale, Infineon, Intel, SYSGO, Vector, Wind
River, FHG IESE and AISEC, OFFIS e. V., TU Braunschweig, TU Kaiser-
slautern, TU München, Universität Paderborn, Universität Stuttgart, and For-
TISS GmbH.
The main goal of ARAMIS is to provide for the technological basis for im-
proving safety, efficiency, and comfort in the mobility domains of automotive,
railway, and avionics by using multicore technology. The insights gained in
32
the project build the indispensable foundation for the successful integration of
embedded systems to cyber-physical systems.
The project is structured in 6 sub-projects: scenarios and requirements, con-
tinuous development method, system design, hardware, software, and demon-
strators. Results that are common to all application domains are captured in the
so-called Domain Common. Our research group has the academic sub-project
lead in “scenarios and requirements” together with AUDI as industrial lead.
Within that subproject, sustainability was identified as a cross-cutting con-
cern we needed to analyse.
Smart Mobility Scenario As there is a large conjoint of partners, we chose

to analyse and model for a product rather than for a company.
We first decided on which of the cyber physical system scenarios to analyse
sustainability for. The smart mobility scenario was selected. It is a contrived
future mobility scenario that combines the automotive, the railway and the
avionics domains. The scenario describes the journey of a family from Munich to
Oslo and chiefly highlights the enhanced comfort and support from the viewpoint
of the end users. With the help of this scenario, the reader becomes aware of
the reach and sophistication of the systems envisioned in the project ARAMiS.
Figure 6.3: Smart Mobility Scenario Overview
33
The scenario’s story is depicted in Figure 6.3. Ms Rosemarie Weber plans
to spend the next Christmas break with her two children at her mother’s, Ms
Pauline Mayer. The Weber family lives in Munich, Ms Mayer lives in Sandvika
near Oslo. Ms Weber’s intention is to pick up her children from school and
from there to travel directly to her mother. In the scenario Ms Weber plan is
worked out. Ms Weber enters departure time as well as from and to locations, a
maximum cost amount for the entire route as well as passengers’ names in her
smart device. The mobile device is connected to various providers and to Ms
Weber’s private cloud, and makes her suggestions for the trip from her home
to pick up her children at their school in the city centre and onward to her
mother. Ms Weber decides to use public transportation to the school, from
there to continue with her children to the airport with a hybrid car-sharing
vehicle (CSV) with autonomous driving capabilities, and finally to reach Oslo
by plane, as this is the most energy-efficient and least expensive alternative. The
necessary travel documents such as public transport ticket, CSV authorization
and flight ticket are transmitted to the mobile device of Ms Weber. Further
profile details are automatically taken into account; e.g., preferences concerning
meals on board are directly transmitted from Ms Weber’s private cloud to the
airline.
Analysis Phase In the analysis phase (see Sec. 6.2.1), we started with the
dimension of environmental sustainability, as depicted in Figure 6.4.
The goal instantiated for the values of preserve scarce resources and clean
air (cf. Fig. 5.3) is reduce resource consumption.
There are four major indicators that can be taken into account for assessing
the effectiveness of activities: energy footprint, carbon emissions, bought items
list, and physical waste.
Activities to achieve the goal of reducing resource consumption in the smart
mobility service are, for example, to make owning cars superfluous for the ser-
vice, to offer older vehicles for lower rates, to include car sharing into the busi-
ness model, and to minimize the energy consumption of the participating mobility
devices.
Application and Assessment Phase In the application and assessment

phase, the indicators are measured to capture the current state of the situa-
tion. Subsequently, the activities are implemented and then the indicators are
reassessed.
For the ARAMiS project, however, the smart mobility service is only one of
several vision scenarios that form the rationale for the development of demon-
strator systems while the actual overall service is still a future (and futuristic)
business model. Therefore, we cannot provide details on the application phase
of this instance.
34
Smart Mobility Service Example Instance Legend
Contributes
<Goal>
Sustainability Influences
Supports
chose <Dimension> Goal

dimension Environmental
Sustainability Dimension
Value
instantiate value <Value>
as goal Reduce Regulation
resource
consumption
Activity
define
indicators
<Indicator>
<Indicator> <Indicator> energy
physical carbon footprint
waste emissions
<Indicator>
bought
items list
select <Activity>
activities <Activity> <Activity> minimize
reduce include private energy
waste cars in fleet consumption
by service
<Activity>
reuse old <Activity> <Activity>
hardware make owning minimize
cars <Activity>
energy minimize
superfluous for consumption
service energy
of participating consumption
mobility in journey
devices
<Activity>
offer older
vehicles for
lower rate <Activity>
minimize minimize include car
energy energy sharing into
consumption consumption business
SW HW model
Figure 6.4: Smart Mobility Service Example Instance
35
Chapter 7
Application: Instances of
the Sustainability Model
This sections describe example instantiations of the generic goal model. We

will see examples for stakeholder analysis as described in Chapter 4 as well as
Product- and Company-specific instances as described in Chapter 5 and 6.
7.1 Product-Specific Instance @BMW

The first instantion took place at BMW Munich for a car sharing plattform.
With DriveNow1 , the german car manifacturer BMW created the following
business model: Instead of buying a car, drivers can get a membership for
the program and pick up a DriveNow car whenever they need it. The cars are
maintained and refilled by the manifacturer. In consequence, the driver has no
fixed costs, but pays per time using the car. The program has been established
in Munich, Berlin, Duesseldorf, Cologne and San Francisco. In this instantiation
we will focus on the car sharing plattform for Munich.
In the following, we will first conduct a stakeholder analysis and afterwards
go into the instantiation of the sustainability model.
7.1.1 Car Sharing Platform: Stakeholders

For the concrete instance of a car sharing plattform we can instantiate the
general stakeholder list from Chapter 4. Each stakeholder has a certain view on
the project and has his own goals. It is important for the success of the project
to be aware of these goals [7].
For DriveNow we identified the following stakeholders (see Fig. 7.1):
• CEO of BMW: The CEO makes sure that the sustainability model is in
accordance with the company’s general goals. As such his goals are the
1 http://tinyurl.com/9fc8yl3
36
Figure 7.1: Sustainability stakeholders of the DriveNow projects.
following:
– Survive, succeed, prosper of company ([8])
– Financial success
– Positive image for company
– Customer satisfaction
• DriveNow Project Manager: The project manager’s primary focus is the
project itself. His goals are consequently:
– Survive, succeed, prosper of project
– Visibility of project
– Big user community
• Finance of BMW: The BMW finanzing section is interested in the finanzial
benefits of the project. Their goals are:
– Financial success
– Correct accounting
• Admin of the Software Solution: The administrators needs to watch the
running software solution over the next few years. Hence, they care for a
long-lasting infrastructure.
• Software maintenance of the Developed Software: The software mainte-
nance wants to be able to change the car sharing software and fix bugs
that appear during its life time. Their goals are:
– Automated registration and management
– As little effort as possible
37
• Vehicle maintenance of BMW Drive Now Cars: The people who repair
the DriveNow cars are interested in having long-living vehicles with low
maintenance at the vehicle. This means:
– Little abrasion
– Little mending
• Employee representatives of BMW: The employee representatives are elected
to represent the workers rights. Consequently their goals include:
– Good working conditions
– Safety
• Driver (Customer): The customer is the person that pays for the DriveNow
service. In order to use the service, he expects:
– Easily understandable user interface
– Quick registration and simple billing
– Economic mobility services
• Software programmer at BMW CarIT (Developer): The programmer needs
to create the software. Inter alia his goals are:
– Good interface definitions
– Successful development
– Sleek code ;)
• Legislation of German Government: The german government is in charge
for the traffic in their cities. They represent the society and the interests
of the country as a whole. Among others, they focus on:
– Safety
– Minimal environmental impact
• CRM: The CRM wants to establish long-term relationships with their
customers. The CRM’s goals are:
– Positive image for company
– Customer satisfaction
– Long-term profitable customers
• Mayor of Munich (Community representative): The Mayor is a repre-
sentative of the inhabitants of Munich. He focusses on the goals of the
community in Munich.
– Beautiful city
– Traffic reduction
– Good reputation
– Increased percentage of public transportation
• CSR manager: The CSR manager
– Positive image with regard to environmental impact and conservation
efforts
38
• Greenpeace (Nature conservation activists): Because nature does not have
a representative to speak out for it, lobbyists and activists from Green-
peace and similar organizations act as their representatives. Hence, their
goals include:
– Conservation of nature
– Reduction of emissions and waste
This stakeholder analysis is an example of a helpful starting point for a

proper analysis of the sustainability dimensions. Together with the stakeholders
we can now build a project-specific instance of the sustainability model.
7.1.2 Instantiation of the Sustainability Model

We created a product-specific instance of the generic sustainability model for
DriveNow. An excerpt is depicted in Fig. 7.2. According to the method de-
scribed in Chapter 6, we first instantiated the generic model in a product-specific
goal model in the analysis phase. This works like the following: For environ-
mental sustainability the value reduce environmental impact is chosen. We can
select similar indicators from the generic goal model as in the company-specific
instantiation, such as the energy bill. Yet, we need to tailor our activities specifi-
cally to this product: For example, in this system the usage of green data centers
is advised. Furthermore, the business process can be varied by offering older
cars for a lower rental rate (instead of deposing them) and thereby reducing
waste, and by introducing cars with hydro or electric power.
Again, in the application & assessment phase, we need to make sure that
these activities are followed during the development of a product and measure
the impact of activities.
7.2 Company-Specific Instance @jambit

We also developed a company-specific instance the generic sustainability model,
of which an excerpt is depicted in Fig. 7.3. For environmental sustainability we
instantiate the value Reduce resource consumption into the goal Reduce resources
consumption by 30% within 12 month. We find three indicators for resource
consumption: physical waste, the company’s energy bill and the size of the
bought items list. We can immediately see differences in how easy one can
measure these indicators. From these indicators we can select activities that
have an influence on the indicators. For example to reduce the size of the
bought items list, we can place incentives for less resource consumption. All
values, indicators and activities can be refined within the model. For example,
the unspecific activity reduce waste is refined within the subactivity recycle
packaging for own shipping.
In the application & assessment phase we would now need to establish the
activities and measure our indicators. From this information we can deduce
39
Legend Product-specific Example Instance
Contributes
<Goal>
Influences Sustainability
Supports
chose
Goal <Dimension> dimension
Environment.
Dimension Sustainability
instantiate value
as goal
Value
define
Regulation <Value> indicators
reduce
Activity environmental <Activity>
impact use green
data center
<Activity> <Indicator> select

optimize energy bill activities
<Indicator>
resource
emissions
consumption
<Indicator>
<Activity> physical
reduce waste
<Activity> emissions
offer
incentives for
offer incentives offer cars instead of
returning car <Activity>
for car pooling older than 5 exchanging
to hot spots use cars years for them for
with hydro less new ones
or electro
power
Figure 7.2: Example instances of the generic sustainability model for the prod-
uct DriveNow (M2).
40
how well our activities are performing and whether or not we work towards the
company-specific sustainability goals.
The first industrial case study for a company-specific instance is currently
under development in a small software development company (jambit GmbH2 ,
50 developers) in a series of workshops that are accompanied by a Master’s
thesis.3
Company-specific Example Instance Legend

<Goal> Contributes
chose Sustainability Influences
dimension Supports
<Dimension> Goal
Environmental
Sustainability Dimension
instantiate value
as goal
Value
<Value>
Reduce resource
Regulation
define consumption by 30%
indicators within 12 months
Activity
<Indicator>
physical
waste <Indicator>
energy bill
select
activities
<Indicator>
bought
<Activity>
items list <Activity>
reduce
waste <Activity> choose
<Activity> design energy-
reuse old partially efficient
hardware offline patterns
<Activity>
<Activity>
establish
recycle
incentives for
packaging for
less resource
own shipping
consumption
Figure 7.3: Example instance of the generic sustainability model for a company
(M2).
2 http://www.jambit.com
3 http://tinyurl.com/9xdpqgc
41
7.3 Requirements Engineering Conference 2013
The model depicted in Figure 7.4 - 7.8 is used by the organizing committee of
the International Requirements Engineering Conference 2013.
Legend
<Goal>
Contributes Sustainability
Influences
Supports
<Dimension>
Goal Human
Sustainability
Dimension
Value
<Value> <Value>
<Value> <Value>
Regulation Human Development
Curiosity Dignity
Health / Growth
Activity
Indicator <Indicator> <Activity>

Vividness of Create
discussions workshops for <Activity>
newbies Discussion: How
can people
<Activity> make most of
<Activity> <Activity> the conference?
Create
Provide time and Enable access
sustainable
space for for people with
pace in time
discussions disabilities
table

Enable Enable people Enable people
participation who cannot travel who cannot see/
in discussions to follow the hear to follow the
lateron conference conference
<Activity>
digitalize
discussion
Figure 7.4: Example instance for the human dimension of the Requirements
Engineering Conference (M2).
42
Legend
<Goal>
Contributes Sustainability
Influences
Supports
<Dimension>
Goal Social
Sustainability
Dimension
Value
<Value>
<Value> <Value> <Value>
Regulation Community
Fairness Trust Tolerance
Building
Activity
Indicator
<Activity>
<Activity> Ensure the conf helps
<Indicator> <Indicator> people to discover the
Create open
Number of Questionnaire on Brazilian people and
data
returning subjective culture
plattforms
attendees Impression
regaring value <Activity>
Ensure Conf has a
positive impact on
people working for us
there (fair trade)
<Activity>
educate on
<Activity> discussion <Activity>
Make fairness Watch
attendees want discussion
to return fairness
Figure 7.5: Example instance for the social dimension of the Requirements
RE'13 Submodel Environmental Sustainability

<Goal>
Sustainability
<Dimension>
Environmental
Sustainability
<Value>
Environmental health
<Indicator> <Indicator> <Indicator>
water <Value> individual number of used
consumption clean water assessment
<Indicator> garbage bags
at venue
aggregated
emissions
<Value> <Value>
clean air less waste
<Value>
less emissions
<Activity>
paper folder
<Activity> (no bag) for
educate <Activity> <Activity> conference
people about walk to use materials at
saving water conference <Activity> sustainable registration
<Activity> dinner suggest to catering
put stickers buy emission (university?)
to sinks compensation <Activity> <Activity>
for flight use local food recycled
<Activity> production <Activity> paper for
get water bottles with register for program and
<Activity> printed
filters for tab water mass transport <Activity> <Activity> flyers
www.waterbobble.com conference
from airport Smart Venue that minimize materials
saves energy (smart waste by using
mgmt of light, real plates &
<Activity> ventilation...) cutlery
<Activity>
shared rides
follow
conference
via skype <Activity>
no 1-use
plastic cups
Figure 7.6: Example instance for the environmental dimension of the Require-
ments Engineering Conference (M2).
43
RE'13 Submodel Economic Sustainability
<Goal>
Sustainability
<Dimension>
Economic
Sustainability
<Indicator> <Value> <Value>

total sum <Indicator> <Value> <Value>
Save money for individual Economic long term
of costs save money
organisation expenses Fairness break even
for attendee
<Activity>
Create "fair"
choose use cheap <Activity>
conference
economic accomodation support
fees
venue (e.g. hostel) regular
participation
<Activity> of attendees
choose economic
option (yet
original/typical) <Activity>
for conference <Activity> Student
banquet Create stipends for "helping
people who need hands"
it, e.g. students,
"poor scientists"
Figure 7.7: Example instance for the economic dimension of the Requirements
44
Legend
Contributes <Goal>
Influences Sustainability
Supports
Goal <Dimension>
Technical
Sustainability
Dimension
Value
<Value>
Regulation Conservation of
Knowledge
Activity
Indicator
<Indicator>
Number of
useful
documents
found
<Indicator>
Organisation
Effort
<Activity>
<Activity> Create a Workshop
Create and for transferring
use a wiki knowledge
between chairs
Figure 7.8: Example instance for the technical dimension of the Requirements
45
7.4 Assessment: Evaluation and Discussion
The previous examples already show both benefits as well as problems and
future work for modeling with the generic sustainability model.
7.4.1 Benefits
First, we can see the purpose of the sustainability goal model: Many elements of
the model (such as the physical waste indicator) reappear in various instances.
Hence, we can imagine a process for modeling sustainability where much of the
modeling can be reused and only needs to be selected.
Second, we can directly see during the modeling how sustainability, a rather ab-
stract concept, is turned into a concrete and measurable property of a company
or a product. By making these implicit definitions explicit via the model, it is
possible to discuss and evaluate a company’s or product’s impact on sustain-
ability.
7.4.2 Issues
While creating the instances we were faced with the question of how to measure
the impact of the proposed improvement activities for a product that does not
yet exist. For example, when taking a product design decision such as using
green data centers, there is no “before” that could be compared to an “after”
situation as the system does not exist yet. Consequently, we can only compare
to “standard design”, but this is a less convincing alternative.
In that case, the model can still serve well in a constructive way for decomposing
sustainability goals and selecting activities to realise them, but the analytic
phase after implementation reveals less concrete results.
7.4.3 Threats to Validity

One threat to validity arises from our constructive research approach for the
model. The risk is limited by our experience in developing such models, by
reviews from and discussion with other researchers, and by feedback from in-
dustrial partners.
The other threat is that the evaluation in practice is still work in progress which
might lead to further insights and changes in the methodic aspects of the ap-
proach.
Both threats will be addressed by industry collaborations in future work.
46
Chapter 8
Conclusion
We propose a generic sustainability model with process- and product-specific in-

stances that can help requirements engineers to analyse their projects according
to the different dimensions of sustainability and choose actions for improvement.
Special emphasis was put on environmental sustainability in and via software
systems, as this dimension of sustainability is the one that is least supported by
our traditional ways of developing software systems.
8.1 Results
The approach we suggest consists of a meta-model, a generic sustainability
model and an approach on how to develop company- and project-specific sus-
tainability models. In oder to explain how to repeat this instantiation we give
examples for sustainability modelling for BMW Drive Now, for a Munich-based
company and for the upcomming RE2013 conference.
When instantiating the sustainability models we realized the high potential
for reuse. We understood that large parts can be reused, which argues for cre-
ating a reusable sustainability model. Nevertheless, various decisions towards
sustainable development of systems and sustainable companies have to respect
the specialities of these projects and companies. Hence, the sustainability model
must be tailored to solve the conflicts between the different goals of the stake-
holders. The resulting sustainability models are still very scarce but already
serve as a good basis for discussion of goals, actions and performance.
8.2 Future Work

One important issue for future work is how to treat conflicting goals. On first
sight, conflicts can be identified by contradicting influences on same indicators.
For example, the ever-present conflict between a tight budget and quality im-
provement measures is also found in the sustainability goals and activities (i.e.
47
in indicators for economical vs. indicators for technical sustainability). How-
ever, more challenging is the question whether there are conflicts that are not
as obvious and how to track them down if they do not relate to same indicators
in the first place.
The other major issue is to understand and quantify the reuse potential of
instances (how systems-specific are they really?) and the gradual extension of
the reference model by knowledge gained from instantiating the model for pro-
cesses and products in various companies. We expect continuous improvement
by evolving the model over time in the ongoing cooperation with our industrial
partners.
48
Acknowledgments
We would like to thank Susanne Klein, Oliver Feldmann, and Jonas Eckhardt
for helpful comments on drafts of this paper as well as Debra Richardson and
Bill Tomlinson for feedback on early ideas.
ToDo: include further reviewers
49
Bibliography
[1] Hilty, L.M., Arnfalk, P., Erdmann, L., Goodman, J., Lehmann, M., Wager,
P.A.: The relevance of information and communication technologies for
environmental sustainability — a prospective simulation study. Environ-
mental Modelling & Software 21 (2006) 1618 – 1629
[2] Naumann, S., Dick, M., Kern, E., Johann, T.: The greensoft model: A
reference model for green and sustainable software and its engineering.
Sustainable Computing: Informatics and Systems (2011) –
[3] Gu, Q., Lago, P., Potenza, S.: Aligning economic impact with environmen-
tal benefits: A green strategy model. In: First International Workshop on
Green and Sustainable Software. (2012)
[4] Cabot, J., Easterbrook, S., Horkoff, J., Lessard, L., Liaskos, S., Mazon,
J.N.: Integrating sustainability in decision-making processes: A modelling
strategy. In: Software Engineering - Companion Volume, 2009. ICSE-
Companion 2009. 31st International Conference on. (2009) 207 –210
[5] Stefan, D., Barrett, M., Letier, E., Stella-Sawicki, M.: Goal-oriented sys-
tem modelling for managing environmental sustainability. In: International
Workshop on Software Research and Climate Change (WSRCC). (2010)
[6] Mahaux, M., Heymans, P., Saval, G.: Discovering Sustainability Require-
ments: an Experience Report. In: 17th REFSQ. (2011)
[7] van Lamsweerde, A.: Goal-oriented requirements engineering: A guided
tour. In: 5th IEEE Intl. Symposium on Requirements Engineering. (2001)
249
[8] Kaplan, R.S., Norton, D.P.: The balanced scorecard: measures that drive
performance. Harvard Business Review 83 (2005) 172–80 7.
[9] Dobson, A.: Environment sustainabilities: An analysis and a typology.
Environmental Politics 5 (1996) 401–428
[10] Burger, P., Christen, M.: Towards a capability approach of sustainability.

Journal of Cleaner Production 19 (2001) 787–795
50
[11] United Nations World Commission on Environment and Development: Re-
port: Our Common Future. In: United Nations Conference on Environment
and Development. (1987)
[12] Robert, K.H., Schmidt-Bleek, B., de Larderel, J.A., Basile, G., Jansen,
J., Kuehr, R., Thomas, P.P., Suzuki, M., Hawken, P., Wackernagel, M.:
Strategic sustainable development — selection, design and synergies of ap-
plied tools. Journal of Cleaner Production 10 (2002) 197–214
[13] Organization, I.S.: ISO 14001 Environmental Management System Stan-
dard (2009)
[14] Bauman, H., Tillman, A.M.: The Hitch Hiker’s Guide to LCA: An Orienta-
tion in Life Cycle Assessment Methodology and Applications. Professional
Pub Serv (2004)
[15] Rees, W.E.: Ecological footprints and appropriated carrying capacity:
what urban economics leaves out. Environment and Urbanization 4 (1992)
121–130
[16] Christen, M., Schmidt, S.: A formal framework for conceptions of sustain-
ability — a theoretical contribution to the discourse in sustainable devel-
opment. Sustainable Development (2011) DOI: 10.1002/sd.518
[17] Penzenstadler, B., Bauer, V., Calero, C., Franch, X.: Sustainability in
Software Engineering: A Systematic Literature Review for Building up a
Knowledge Base. In: Proceedings of the 16th International Conference on
Evaluation & Assessment in Software Engineering. (2012)
[18] Meadows, D.H.: Thinking in Systems - A primer. Earthscan (2008)
[19] Hilty, L., Lohmann, W., Huang, E.: Sustainability and ICT — an overview
of the field. In: Proceedings of the EnviroInfo 2011. (2011)
[20] Jevons, W.S.: The coal question. 2nd edn. Macmillan and Co. London
(1866)
[21] Group, T.C.: Smart 2020: Enabling the low carbon economy in the infor-
mation age. Technical report, Global eSustainability Initiative (2008)
[22] Komiyama, H., Kraines, S.: Vision 2050 - Roadmap for a Sustainable
Earth. Springer (2008)
[23] Goodland, R.: Sustainability: Human, Social, Economic and Environmen-
tal. In: Encyclopedia of Global Environmental Change. John Wiley and
Sons (2002)
[24] UN World Comission on Environment and Development: Our Common
Future. In: UN Conference on Environment and Development. (1987)
51
[25] Lago, P., Kazman, R., Meyer, N., Morisio, M., Müller, H., Paulisch, F.,
Scanniello, G., Zimmermann, O.: First international workshop on green
and sustainable software at icse (2012)
[26] Zimmerman, M.J.: Plato’s intrinsic vs. extrinsic value. In
Zalta, E.N., ed.: The Stanford Encyclopedia of Philosophy. Winter
2010 edn. http://plato.stanford.edu/archives/win2010/entries/
value-intrinsic-extrinsic/ (2010)
[27] Johnson, R.: Kant’s moral philosophy. In Zalta, E.N., ed.: The Stanford
Encyclopedia of Philosophy. Summer 2012 edn. http://plato.stanford.
edu/archives/sum2012/entries/kant-moral/ (2012)
[28] Atkinson, C., Kühne, T.: Model-driven development: A metamodeling

foundation. IEEE Software 20 (2003) 36–41
[29] Bell, S., Morse, S.: Sustainability Indicators — Measuring the Immeasur-
able? Earthscan (2008, 2nd Ed.)
[30] Yale Center for Environmental Law and Policy Yale University, Center
for International Earth Science Information Network Columbia Univer-
sity: Environmental sustainability index report 2005. www.yale.edu/esi/
ESI2005_Main_Report.pdf (2005)
[31] Marks, N.: Happy planet index. ‘the new economics foundation’, http:
//www.happyplanetindex.org/ (2012)
[32] Boden, A., Avram, G., Bannon, L., Wulf, V.: Knowledge Management in
Distributed Software Development Teams - Does Culture Matter? In: 2009
Fourth IEEE International Conference on Global Software Engineering,
IEEE (2009) 18–27
[33] sponsored by the BMBF: ARAMiS — Automotive, Railway, and Avionics

in Multicore Systems. www.projekt-aramis.de (2012)
[34] Penzenstadler, B., Eckhardt, J.: A Requirements Engineering Content
Model for Cyber-physical Systems. (2012)
[35] Geisberger, E., Broy, M., Cengarle, M.V., Keil, P., Niehaus, J., Thiel, C.,
Thönnißen-Fries, H.J.: agendacps — integrierte forschungsagenda cyber-
physical systems. Technical report, acatech — Deutsche Akademie der
Technikwissenschaften (2012)
[36] Lee, E.A.: Cyber physical systems: Design challenges. Technical Report
UCB/EECS-2008-8, EECS Department, University of California, Berkeley
(2008)
52

A Generic Model For Sustainability

Uploaded by

Copyright:

Available Formats

A Generic Model For Sustainability

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

A Generic Model For Sustainability

Uploaded by

Copyright:

Available Formats

A Generic Model for Sustainability

Birgit Penzenstadler, Henning Femmer

Technische Universität München, Germany

November 26, 2012

Motivation: Software systems as we know them have either a primary eco-

Problem: Without a tangible definition and applicable guidance, sustainabil-

Principal idea: We analyse the dimensions of sustainability, their values with

Contribution: We propose a generic sustainability model with a meta model

Keywords: sustainability, software system, goal, requirements, definition

1 Introduction & Motivation 3

3 Dimensions of Sustainability in Software Engineering 10

4 Stakeholders for Sustainability 14

5 The Generic Sustainability Model 18

6 Process: Instantiation of the Generic Model 28

7 Application: Instances of the Sustainability Model 36

Introduction & Motivation

Sustainability has been recognized as a relevant topic in software engineering,

Problem Currently, we are still missing a definition of what software engi-

Contribution In this report, we present a reference model for sustainability

Outline The report is organized as follows: Chapter 2 discusses related work

Related work can be differentiated into sustainability modeling approaches,

2.1 Sustainability Modeling Approaches

Framework for sustainable software engineering Naumann et al. [2] pro-

Requirements engineering techniques Mahaux et al. [6] present a case

2.2 Economic Domain: Balanced Score Card

• Internal: What must we excel at?

Matching of BSC perspectives to SRM dimensions as instantiation

2.3.1 Frameworks for Defining Sustainability

2.3.2 Defining Sustainability in the Context of Software

2.4 Analyses: Relation between ICT and Sus-

3.1 Dimensions of Sustainability

Human sustainability Human sustainability refers to the maintenance of

Social sustainability Social sustainability means maintaining social capital

Environmental sustainability Environmental sustainability seeks to im-

Technical sustainability From a point of view of (software) systems engi-

3.2 Relations between the Dimensions of Sus-

Figure 3.1: Dimension Relations

Environmental S. — Human S. Environmental sustainability is the basis

Human S. — Social S. In turn, humans are the individual elements that

Social S. — Economic S. Society is organized by individual humans, and

Economic S. — Technical S. Technical systems are developed within a

3.3 Resulting Central Questions for Analyzing

Social Sustainability Which effects do software systems have on the society

Technical Sustainability How can software be created so that it can easily

3.4 The Concept of Value

Kant A value is a reason by which an action or state is judged, which is only

2. Stakeholder-Support: Besides the knowledge of the stakeholders, it is a

• Management: Managers concerns are represented primarily the economic

These classes are not necessarily encompassing but intended to serve as

4.3 Stakeholders according to Sustainability Di-

Figure 4.1: Generic Stakeholders of the Sustainability Model.

Again, many stakeholders can be reused through various projects. However,

The Generic Sustainability

5.1 The Meta Model

1..* supports 1..* 1

1..* influences 1..* 1

Figure 5.1: Meta model of generic sustainability model (M0).

3. Qualitative and quantitative indicators cannot be aggregated to find an