ReThink Your Processes! A Review of Process Mining for Sustainability m

Abstract

The transition towards more sustainable practices requires companies to assess their impact on the social and ecological environment and establish new processes in complex inter-organisational systems. Process mining is a collection of data-driven techniques to visualise, analyse and improve business processes. Its potential to increase sustainable business processes has been acknowledged by academia and industry but not systematically reviewed. This work analyses process mining's application for sustainability by conducting two consecutive literature studies-the first on the broad domain of sustainability, and the second on the circular economy, a widely accepted approach for pursuing sustainability. Results show the potential of process mining for assessing and analysing sustainability in business processes, allowing for data-driven decision support and targeted improvement. They also show that process mining has yet to reach the sustainability domain. To enable collaboration between both communities, we present PM4S, a framework for applying process mining for sustainability.

Full text

ReThink Your Processes! A Review of Process
Mining for Sustainability
Nina Graves, Istv´
an Koren, Wil M.P. van der Aalst
Chair of Process and Data Science (PADS)
RWTH Aachen University
Aachen, Germany
{graves, koren, wvdaalst}@pads.rwth-aachen.de
Abstract—The transition towards more sustainable practices
requires companies to assess their impact on the social and
ecological environment and establish new processes in complex
inter-organisational systems. Process mining is a collection of
data-driven techniques to visualise, analyse and improve business
processes. Its potential to increase sustainable business processes
has been acknowledged by academia and industry but not
systematically reviewed. This work analyses process mining’s
application for sustainability by conducting two consecutive
literature studies – the first on the broad domain of sustainability,
and the second on the circular economy, a widely accepted
approach for pursuing sustainability. Results show the potential
of process mining for assessing and analysing sustainability in
business processes, allowing for data-driven decision support and
targeted improvement. They also show that process mining has
yet to reach the sustainability domain. To enable collaboration
between both communities, we present PM4S, a framework for
applying process mining for sustainability.
Index Terms—process mining, sustainability, systematic liter-
ature review
I. INT ROD UC TI ON
Every year, tonnes of waste are created, an increasing
amount of emissions are released into the atmosphere, and the
global distribution of wealth and health becomes increasingly
unbalanced [25]. These practices impact future generations’
quality of life [47] – for example, by exploiting scarce re-
sources and creating a climate crisis. With the United Nations’
2030 Agenda, the global society agreed to strive for sus-
tainable development (SD) [69]. In 2016, industrial processes
and transportation were responsible for 45.6% of global emis-
sions [61] and, therefore, strongly affect the success or failure
of SD [44]. To eliminate this tremendous impact, organisations
must realign their business models [6], supply chains (SC) [23]
and organisational practices [25].
Business Processes (BP) form the “arterial systems” of organ-
isations [12], thereby being the critical element of its contribu-
tion to SD [66]. Business process management (BPM) must,
therefore, incorporate sustainable objectives. As BPM sub-
stantially impacts industrial operations, an increasing amount
of research is put into “Green BPM” [30]. However, BPM’s
contribution so far is mainly of conceptual and qualitative
nature [20]. To provide more concrete and scalable support for
more sustainable BP, this work evaluates how process mining
©2023 IEEE, Author’s Preprint, Accepted for ICT4S Conference
(PM), the quantitative branch of BPM, can contribute to SD.
PM is a relatively new discipline considered to X-ray business
processes [70]. PM uses the data logged by IT systems when
used in a business’s day-to-day life to identify and visualise the
relationship between the different activities [71]. The resulting
process model is then used to analyse and identify the potential
for improvement in the BP. PM techniques enable the auto-
mated identification of deviations from a benchmark process
and extensive process analysis under the consideration of
additional business data [74]. As PM offers a holistic view of
as-is end-to-end processes, it has received increasing attention
in academia [22] and industry [21] to support processes in
a variety of domains, such as healthcare, manufacturing and
financial auditing.
In related work, scientists and practitioners have already noted
PM’s potential to drive sustainability. PM’s potential as an
auxiliary technique to support carbon accounting was assessed
in [7], and the application of PM has been recognised as a
technique supporting the assessment of a product’s impact
on its environment [20], [50]. Additionally, [59] provides an
overview of the status and future of PM, naming sustainability
as an essential domain PM can contribute to, by supporting the
efficient utilisation of scarce resources. PM software vendors
have started considering the impact of PM on sustainability
and begun integrating SD perspectives into their tools [5], [56].
Celonis, one of the leading PM tools according to Gartner
[34], has cooperated with a provider of emission data to
integrate the consideration of emissions into their tools and
applications [78].
In general, PM’s potential contribution to SD is considered
broad – from supporting the assessment of the sustainability
of business practices to actually contributing by automating
activities. Although the potential of PM for SD has been
recognised, its potential has only been analysed for selected
elements, such as impact assessment and emissions. Neither
academia nor industry considers the full potential of state-
of-the-art PM techniques, limiting consideration to process
discovery and conformance checking and only focusing on
PM for individual processes instead of process networks.
In the past, PM techniques have proven highly beneficial to
objectives such as the reduction of cost, an improved under-
standing of end-to-end processes and an increase in overall
performance [59]. In this work, we discuss the application of

Fig. 1: Principle of the circular economy (based on [2]), linear
economy (left) and CE (right).
PM techniques to SD and investigate the following research
questions: (RQ1) How can PM contribute to driving SD?
(RQ2) How must PM be applied to be most beneficial for
SD? (RQ3) What further research is required to make PM
techniques more effective in driving SD?
The remainder of this work is structured as follows: We
introduce SD, the circular economy (CE) and PM in Section II.
In Section IV, PM’s connection to the broad domain of
SD is assessed and discussed with a systematic literature
review (SLR-1) [37]. Based on the findings of SLR-1, we
conduct a second review (SLR-2) to analyse the application
of PM to the targeted and process-related field of the CE in
Section V. Finally, we create a framework for applying PM
for sustainability (PM4S) in Section VI. Section VII concludes
this work.
II. BACKGROU ND
To ensure a common understanding for all that follows, we
provide some background information on SD, the CE and PM.
A. Sustainable Development
The term sustainable development was coined in 1987 as
economic growth through “development that meets the needs
of the present without compromising the ability of the future
generations to meet their own needs” [47]. It is commonly
agreed [55] that SD encompasses three pillars: economic
prosperity, social equity, and environmental quality [14]. Eco-
nomic prosperity focuses on the increase of performance and
the strengthening of markets, and social equity deals with
the increase of human well-being and the development and
strengthening of societies [35]. The regeneration of nature,
the circulation of products and materials, and the elimination
of waste and pollution increase environmental quality [15].
B. Circular Economy
The circular economy (CE) is a widely accepted approach
to pursue mainly environmental SD [27]. The underlying idea
is to move away from the “end-of-life” concept, in which
material is sourced, and value is added to create products
which are then used and eventually disposed of. In the CE,
processes support material and energy regeneration by creating
material and energy loops [35]. The CE describes a “system
in which resource input, waste, emission, and energy leakage
Fig. 2: Frequent types of process mining [74].
are minimised by slowing, closing and narrowing material and
energy loops” [24]. The principles of CE can be described
by ten so-called R-imperatives [57], which address different
phases of a product’s life cycle [2], see Figure 1.
Before beginning a product’s life cycle, there is the option
to refuse (R0) using hazardous or virgin material. After
raw materials are sourced, smart manufacturing and targeted
material use [54] support the reduction (R1) of input resources,
waste, emissions and energy leakage [25]. Material loops can
be slowed by following strategies to extend product life [2]:
While a product’s value is still high, it can be reused (R2),
maintained or repaired (R3). When functions remain intact, but
minor components must be replaced, they can be refurbished
(R4). Remanufacturing (R5) describes a process in which all
defective parts are replaced. If the original added value cannot
be restored, products and materials can be repurposed (R6) –
used for something other than originally intended [57]. After
all added value is depleted, material loops can be closed by
retrieving end-of-life materials [54]. This involves recycling
(R7), where the product completely loses its structure, so
merely the material itself and none of the added value is
reused [27] and the recovery (R8) of energy trapped in the
product or material. Materials can also be re-mined (R9) from
a landfill when new options for material usage are found [57].
Apart from designing products for long life and life exten-
sion [6], the transition to the circular economy requires the
introduction and management of processes supporting these
practices [77]. Additionally, circularity [10] and the environ-
mental impact of existing processes [29] must be assessed. Life
cycle assessment (LCA) is a standardised method of quantify-
ing a product’s or service’s emissions, material consumption,
impacts on health and the environment, and natural resource
depletion issues throughout its life cycle [16]. Not only does
the transition to the CE mean that products must be used
as efficiently as possible [25], they must also be designed
accordingly [6].
C. Process Mining
IT systems typically support business operations. These
systems usually create an event log documenting every event
that occurs. An event describes the execution of a specific
activity at a certain point in time in connection to a certain

case. For example, the activity “book incoming goods” exe-
cuted at 2023-01-31 12:42 for the case “order 123-456” is an
event. PM is a set of techniques that use these event logs to
generate business intelligence and support the understanding,
analysis and management of processes [71]. Figure 2 provides
a conceptual overview of common types of PM as described
in [74]. Process discovery comprises all PM techniques that are
used to establish a connection between activities to generate an
overview of BPs – a process model. These models are used as
a basis for process analysis and process improvement. Using
the process model and (enriched) event logs as input, processes
are diagnosed in two ways. Conformance checking techniques
identify and quantify the differences between an event log
and a benchmark process model. They aim to show where
the log and model disagree. Besides considering deviation
between as-is and to-be, processes are diagnosed by applying
performance analysis techniques. Performance analysis aims at
uncovering problems within business processes by calculating
performance indicators. These indicators help, for example,
to uncover bottlenecks within a process. Process diagnostics
support the assessment of processes and thereby enable data-
driven decisions. Comparative PM uses event data and diag-
nostics from different procedures and detects differences and
commonalities in their application in consideration of their
outcomes. It helps uncover root causes and is a good tool
for inter- and intra-organisational benchmarking. Predictive
PM combines PM with machine learning (ML) techniques
to answer forward-looking questions. It uses process data,
diagnostics and additional (business) data to detect and analyse
reoccurring anomalies in processes and enables the prediction
and analysis of possible future scenarios and their impact.
Action-oriented PM aims at automating actions based on
process diagnostics of running processes. Apart from trig-
gering activities according to business rules, Robotic Process
Automation (RPA) is a technique in which repetitive, manual
tasks are performed by a software that mimics a human
clicking and entering data.
A relatively new area of PM is object-centric (OC) PM.
Instead of considering a single case, OCPM allows the as-
sociation of multiple objects to each event. Each object has
a certain object type. Therefore, OCPM not only allows the
analysis between activities but also considers the relationship
between activities and different object types as well as between
individual object types [73]. It uses an OC event log (OCEL)
as input. The current OCEL standard, the common basis for
OCPM techniques, assigns a single object type and attributes
to each object. Every event is associated with an activity, event
attributes and the objects related to its execution [17].
III. MET HO DO LO GY LI TE RATU RE RE VI EW S
The application of PM techniques to SD is discussed based
on the results of two SLRs performed after the framework
of Kitchenham [37]. This section describes the methodology
of these reviews, denoting the first review of PM in the
broad context of SD as SLR-1, and the second specifically
on PM and the CE as SLR-2. Attempting an exhaustive
Fig. 3: Methodology of the two systematic literature reviews.
search of relevant work, the publication databases Scopus,
Web of Science, ACM Digital Library, IEEE Xplore, Science
Direct and Google Scholar were used. The search strings were
applied to the title, and, where possible, to the abstract and
keywords. Forewords and conference reviews were excluded,
as well as any works without accessible full texts in English
or German. The terms searched for can be found in Figure 3,
where the methodology is depicted.
The search results were refined by applying the inclusion
criteria to all 34 full texts in SLR-1 and the full texts of
the publications remaining after abstract screening of the 120
search results of SLR-2. To be included, the publication was
required to refer to the application of process mining or the
analysis of event data in a manner described in Section II-C.
So, any publications referring to the (qualitative) analysis of
processes in the mining industry or using the term ”process
mining” in a different meaning were excluded. Additionally,
addressing an increase in SD (SLR-1) or a contribution to
the circular economy based on the definition presented in
Section II-B (SLR-2) was required. Publications in which
sustainability is merely mentioned in the metadata are also
excluded. The details can be taken from Figure 3. Due to
the few relevant search results found (14 each), no additional
criteria that might reduce the number of findings were applied.
Attempting to add unidentified publications, the inclusion
and exclusion criteria were applied to the references mentioned
in the related work sections. All publications associated with
environmental quality identified in SLR-1 were added to the
analysis of SLR-2. Unfortunately, this did not increase the
number of publications for SLR-1 and resulted in a final count
of 22 publications for SLR-2.
Based on the understanding of SD and the CE presented in
Section II, the addressed pillar and topic of sustainability are
determined for each identified publication. We also distinguish
whether SD is referred to directly or indirectly, e.g. by being
mentioned as a potential benefit or motivation. Additionally,
we categorise the publications as a publication describing a
methodology (M), a case study (CS), or a survey/position
paper (S/P). For the analysis of the papers, we also con-
sidered their year of publication, the publication’s objective,
the involved process and industry, and the purpose the PM

TABLE I: Relevant Publications of SLR-1 and SLR-2
References, Pillar (general SD, economic, social, environmental), Year, Type of Publication (Methodology, Case Study, Survey/Position), Relationship to SD (Direct, Indirect), Industry
the publication refers to (“Industry” if not specified but private sector approved), Type of process addressed, Topic, PM’s contribution, SLR-1 top, SLR-2 all general and environmental
Ref SLR Pillar Year Type D/I Industry Process Objective PM Contribution SD Aspect
[53] 1 soc 2021 CS D Industry Occupational Safety PM to improve health and safety Identify and analyse process, organisational connections,
and deviations
Health and Safety
[79] 1 soc 2014 S/P I IT Security Data Flow Improve control and transparency for information
exchange
Identify data flow Welfare and Security
[52] 1 soc 2015 S/P D Educational
Management
Core Processes Enable sustainable universities Identify and analyse process, organisational Connections
and deviations from regulations
Education
[36] 1 eco 2010 CS I Manufacturing Core Processes Increase organisational flexibility Identify process, analyse different process outcomes Performance
[81] 1 eco 2019 CS I Maritime
Transport
Information Processes Efficient freight export Monitor process, detect deviations, identify weaknesses Performance
[28] 1 eco 2016 M D Industry Core Processes Strengthen competitive sustainability Monitor and analyse process, identify inefficiencies Markets,
Performance
[38] 1,2 eco, env 2022 M, CS I Manufacturing Material Flow Identify constrained resources Monitor process, detect deviations Performance,
Resources
[62] 1,2 eco, env 2012 M I Industry Core Processes Support business network management Identify organisational connections (social network anal-
ysis)
Performance, Sourc-
ing
[49] 1,2 gen 2015 M D Industry Maintenance Sustainable maintenance through semantic PM Monitor and assess maintenance performance, identify
abnormalities
Availability,
Efficiency
[48] 1,2 gen 2018 M, CS D Industry Core Processes Holistic corporate sustainability Identify directly follows relation of activities Waste, Emissions,
Material Input
[65] 1,2 gen 2020 M D not specified not specified Automated compliance checking to SD targets Monitor processes, detect deviations SD Compliance
[60] 1,2 gen 2021 S/P I Manufacturing Material Flow Relationship process mining and process automa-
tion
Monitor and analyse production, Support Automation Efficiency
[31] 1,2 env 2021 M D Manufacturing Material Flow PM for sustainable value stream analysis Identify process, assses indicators LCA
[1] 1,2 env 2022 M, CS D Manufacturing Material Flow Identify energy efficient disassembly sequence Identify process Energy, Remanufac-
turing
[32] 2 env 2014 M, CS I Manufacturing Core Processes Agile operations management for green factory Monitor processes Energy, Efficiency
[51] 2 env 2018 M I Manufacturing Material Flow Flexible CPPS with optimised parameters Identify sequences of control methods Efficiency, Energy
[41] 2 env 2020 M D Industry Maintenance Feasibility of PM for machine health Identify healthy machine behaviour, monitor confor-
mance, support root cause analysis
Energy, Availability
[64] 2 env 2020 D Manufacturing Maintenance Optimal maintenance windows Identify directly follows relation of activities Efficiency, Availabil-
ity
[50] 2 env 2021 S/P D Industry Material Flow Framework for LCA using PM Monitor processes, assess indicators LCA
[11] 2 env 2021 CS D Energy
Systems
Maintenance PM for maintanance of wind turbines Identify and analyse process, bottleneck detection, con-
formance to benchmarks
Sourcing, Efficiency,
Availability
[46] 2 env 2021 D Energy
Systems
Application Process Assessment of suggested process changes Identify process, analyse performance Sourcing, Efficiency
[7] 2 env 2022 S/P D Industry Core processes PM for carbon accounting Identify process, assess CO2 emissions, check SD com-
pliance, simulate changes, identify reduction potential
Carbon Emissions
[13] 2 env 2022 M, CS D Aggriculture Crop Rotation Identify sustainable crop rotation strategy Identify directly follows relation of crops GHG Emissions,
Sourcing
[20] 2 env 2022 S/P D not specified not specified Sustainable BPM for LCA Monitor processes, assess indicators, detect deviations,
root cause analysis
LCA
[39] 1,2 env 2022 I Manufacturing Core Processes Framework for intelligent process automation Monitor process, identify bottlenecks, detect deviations,
root cause analysis, RPA
Waste, Resources,
Efficiency
[19]/ [18] 2 env 2022/21 M,CS I Manufacturing Material Flow Data-driven simulation for digital twins Monitor process, check conformance Energy, Efficiency
[63] 2 env 2017 M D Manufacturing Maintenance Optimal maintenance windows Identify directly follows relation of activities Availability,
Efficiency

application serves. Table I enlists the identified publications.
IV. PROCESS MINING AND SUSTAI NABLE DE VE LO PM EN T
This section considers the application of PM in the broad
context of SD. We first present the connections between PM
and SD established in literature and subsequently discuss how
PM has been beneficial in driving economic prosperity, social
equity, and environmental quality.
The fourteen publications related to SLR-1 can be found
at the top of Table I. Eight of them address SD directly.
A case study and two positions address the social pillar by
aiming to improve health and safety [53], education [52], and
welfare and security [79]. Two publications address SD at the
macro level (social and educational infrastructure). All three
publications show applications of PM to identify and analyse
processes in specific domains (industry or process). Two of
them involve the organisational perspective.
Five publications mention applications of PM for economic
prosperity by increasing performance or strengthening markets
through competitive advantage. Only one directly addresses
this connection to the economic pillar. Three include a case
study to improve performance and name sustainability as
potential or motivation. The fourth describes a framework for
sustainable competitiveness [28], and the last uses organisa-
tional process mining to increase performance by analysing
business connections [62]. PM techniques are applied to the
main processes and are used to detect, investigate and monitor
processes, thereby using continuous analysis.
Two publications address SD as a general concept, and
two explicitly mention all three pillars of SD [48], [49]. All
four were, therefore, not linked to a specific pillar of SD
but assigned to a general category. Three of these general
publication’s topics directly address SD - one by creating
a tool evaluating the compliance of processes to SD goals
and regulations [65], one by calculating the sustainability of
processes and detecting gaps to a pre-defined benchmark [48]
another by enabling decisions concerning all pillars of SD in
maintenance processes [49]. The potential for SD through an
increase of process automation achieved by PM is seen in [60].
In all these publications, PM is applied to assess the current
status and support business decisions for more sustainability.
Four publications belong to the environmental pillar, all
describing methodologies. In all of the introduced approaches,
PM is combined with other methods such as optimisation
techniques [1], the BPM-cycle [62], simulation models [38],
or value stream analysis [31].
A. Discussion SLR-1
Despite PM and SD being topics of increasing interest in
academia and industry, only fourteen publications refer to both
PM and SD directly, with four relating to a general concept
of SD. This can have various reasons, such as the absence
of a conceptualisation of SD scholars agree upon [20], the
benefits of PM not directly considered in the context of SD
or the lack of support PM techniques can provide for SD.
All of the presented case studies name the analysis of data-
driven performance and support in identifying improvement
potentials due to the application of PM techniques. This
indicates that PM, at least, supports economic prosperity.
The understanding that increasing performance contributes
to economic prosperity is supported by [9]. The underlying
sample of social sustainability indicates that the benefit of PM
to the social pillar lies in the application of PM to a specific,
human-centred domain and the consideration of the organisa-
tional perspective. Following this, we confirm the increasing
application of PM fields such as healthcare [43] and smart
cities [8] – two topics addressing SD on a macro level. PM
techniques are also known for analysing social networks [75].
Besides a targeted choice in the domain, the consideration
of social aspects is a dedicated topic in recent PM research.
Responsible PM is a new area of a PM that addresses fairness,
accuracy, confidentiality and transparency in the context of
PM [72]. These examples and considerations show that there
is research on the intersection of PM and social sustainability.
Whether this implies the application of PM as a possibility to
improve social equity has to be considered further. We see that
all publications mentioning the environmental pillar propose
methodologies. Additionally, the increase in publications in
recent years could be interpreted as an identified deficit. Apart
from three of the publications relating to the manufacturing
industry, the publications do not indicate any other relationship
in the application of PM, the SD-related topics they address
or the purpose that is served (optimising disassembly [1],
improving resource utilisation [38], assessing environmental
indicators [31], and enabling the selection of responsible
suppliers [62]).
We conclude that searching for an explicit connection
between PM techniques and SD yields few results and offers
only a partial overview of the potential of PM in SD. Over the
last two decades, academia and industry focused on applying
PM to improve the efficiency of processes [59]. Additionally,
companies, which are the primary beneficiaries of PM, have
an intrinsic interest in achieving economic prosperity. The
application of PM to targeted domains and including the
organisational perspective indicates benefits to social sustain-
ability. The relatively new area of responsible PM may also
contribute to social sustainability. However, many indicators
for social equity are established using qualitative data [3]. PM,
in contrast, is a set of quantitative techniques. According to
the social life cycle assessment guideline, processes only cover
about 10% of the scope related to social equity – compared to
the roughly 60% of environmental sustainability [68]. Based
on these insights and the indications of the ecological pillar
and PM being the least developed pillar, we decided to narrow
further analysis to contributions concerning PM and the CE.
V. PROC ES S MIN IN G AN D TH E CIR CU LA R ECO NO MY
As argued above, SD is a vast field, the analysis of which
yields few results, suggesting a varying impact of PM’s contri-
bution to the different pillars of SD. We, therefore, narrow our
field of interest from the broad area of SD to a more concrete

Fig. 4: Processes relevant to the CE.
field: The circular economy. In the remainder of this work, we
want to identify (1) what PM can contribute to the CE and
(2) how PM has to be applied to support the transition to and
maintenance of the CE.
When considering the possible contributions PM can make
to the CE, the obvious first step is identifying the processes of
main interest to the CE. As the CE addresses the circulation
of resources and energy, processes related to the traditional
supply chain (SC), so sourcing, making, and distributing core
products are relevant [4]. However, waste flows, product utili-
sation, and reverse logistics (RL) must also be considered [77].
RL includes all processes required to retrieve a product from
a consumer to restore product or material value [33]. Based
on these considerations, the R-imperatives and the relevant
processes identified in [77] we summarised an overview of
the relevant processes for the CE in Figure 4. The processes
involved in sourcing, making and using products and energy
form the heart of the CE. Working with circulated products
or energy may offer different challenges (e.g. more variability
in the control flow), but it involves the same SC and can,
therefore, be integrated into the same supply system used to
create products of virgin resources [42]. PM can be applied
to a process system part of this SC or only affect individual
processes, e.g. the disassembly of products.
In the following SLR, we look at the relevant publications
from two different points of view: The associated processes of
the CE and their contribution to the CE goals. By categorising
the publications by processes, we want to assess the status and
completeness of PM regarding the CE, whereas the specific
application of PM techniques to support the goals of the CE
is the latter’s focus. We treat [19] and [18] as one publication,
since [19] adds a case study to [18].
Sourcing of material and energy is addressed in four
publications: Two present case studies in the renewable energy
sector [11], [46], and [13] describes how PM can be applied to
reduce greenhouse gases in agriculture. A fourth presents the
application of PM to identify business networks, mentioning
this as an opportunity for a more targeted selection of suppliers
in terms of SD targets [62].
Production processes, so making products, is addressed by
most publications. Six describe PM’s part in some form of
smart manufacturing, such as digital twins [19], [38], [39],
production automation [32], [60] or cyber-physical production
systems [51]. Two address the assessment of environmental
impact [31], [50]. Two publications from the same authors
optimise maintenance windows to increase production effi-
ciency [63], [64].
Maintaining and repairing is the main concern of three
other publications: One presents a case study demonstrating an
improvement of managing maintenance jobs through PM [11],
another proposes a methodology to support knowledge man-
agement for sustainable behaviour [49], and [41] applies PM
to machine logs to support the detection of machine issues and
root-cause detection. Product distribution or utilising con-
sumer goods are not addressed to in any of the publications.
Reverse Logistics Processes are only directly referred to
in one work – in [1] a method to identify the most energy-
efficient disassembly sequence to support remanufacturing is
presented. Additionally, integrating remanufacturing processes
into production systems was mentioned as a requirement for
the digital shadow presented by [19], and [48] describes
a method to assess discrepancies to a benchmark of the
utilisation of recycled material but does not detail on this.
Regarding the objectives of the CE, we see that most
publications refer to smart manufacturing, leading to increased
digitalisation. More digitalisation, in turn, increases process
efficiency, which reduces material input, waste, emissions and
energy consumption [40]. More concretely, efficiency is im-
proved by automating parts of the process through RPA [39],
[51], [60] or the creation of simulation models to compare
alternatives and support decision-making [19]. The method
in [38] monitors processes to identify constrained resources
to support decision-making. A simulation model enriched
with information on processed quantities and locations to
support the material flow, which automatically creates and
evaluates alternatives, is presented in [32]. Machine avail-
ability is optimised by generating a probabilistic prediction
model to assign maintenance windows [63], [64]. Naturally,

the previously introduced publications on maintenance and
repair also contribute to the increase in resource availability.
Reducing resource input directly is achieved by remanu-
facturing [1] and the consideration of recycled material in [48].
Waste is mentioned in three of the publications: A frame-
work for intelligent process automation is presented in [39].
PM is first used to identify the current practices, acting as
a basis for the combination of RPA and an AI to operate a
machine. This optimises material utilisation and a reduction
in waste. In [48], waste is mentioned as an indicator, but no
further information is provided. Publication [19] mentions the
potential of integrating information on waste into a proposed
digital shadow.
Emissions can be reduced by appropriate crop rotation
strategies, according to [13], yet no integration of emission-
related data is mentioned in the approach. The application of
PM for carbon accounting is analysed in detail by [7] using in-
terviews with professionals working on the topic. It concludes
that it supports recognising, measuring, and monitoring CO2
emissions and checking compliance with targets. PM’s poten-
tial for the simulation of process changes and their impact
on carbon emissions and process performance and identifying
CO2 reduction potentials is pointed out. Additionally, the
utilisation of PM for the comparison of alternative practices is
addressed. Unfortunately, the author does not detail any of this
— we can only deduce that it requires information on carbon
emissions per activity. Getting this data is, so the author, a
problem to which the connection to emission databases via
APIs might be an answer.
Energy consumption is addressed in several publications.
In [1], the energy consumption per machine is analysed over
time. The event log is enriched with information on the used
machines. PM generates a process model, which is used to gain
the precedence relationships among the activities. Combined
with the energy consumption of these combinations are sub-
sequently optimised to detect the energy-optimised sequence
of activities. In other publications, energy consumption is
added as an event attribute [50], [51]. A method in which the
process model is enriched with information on machine setting
parameters to determine the optimal energy-efficient process
sequence and machine settings is presented in [51]. The
framework of [50] describes accumulating a product’s energy
consumption throughout the process. Other publications only
mention energy reduction as a consequence.
General contributions to the CE are addressed. One argues
for using PM for LCA in production processes [31] – only
a high-level framework is provided. A literature study on
BPM the CE concludes that BPM has the potential to support
LCA and considers PM a good option for operationalis-
ing LCA, detecting deviations from the planned path and
supporting deviations via root cause analysis [20]. A third
publication describes a high-level framework for integrating
PM to LCA [50]. It discusses the ability of different dis-
covery techniques for this task and mentions that discovery
techniques have to be adjusted to provide information on
cumulative energy consumption and display more details in
process models. It mentions the requirement for a stronger
focus on the material flow, the necessity of data the machines
use and produce, and energy consumption. The framework
presented in [49] should support decisions toward more
sustainable practices by annotating the discovered process
model with semantic information used for more detailed
knowledge extraction and process analysis. Additionally, two
works asses business processes’ compliance to SD targets,
policies and standards [65] or benchmarks [48]. For example,
[65] proposes a translation of these sustainability constraints to
formal logic, to which a compliance engine compares business
practices based on event logs and additional business data. In
general, all publications use PM to detect current practices.
Nineteen publications describe leveraging PM’s automatability
to operationalise the described approaches – none explains
how this is done in detail; conformance-checking of real-time
data is assumed. Four only use PM to identify directly follows
relationships [13], [48], [63], [64], and one performs social
network detection [62]. Conformance-checking techniques are
addressed for the detection of irregularities [11], [38], [41],
monitoring processes [60], validating models and assessing
model quality [19] and comparing practices to sustainability
performance rules [7], [49] and sustainability targets [65].
Performance analysis techniques are mentioned for calculating
performance indicators [11], [19], [39] and the detection of
bottlenecks [38], [39]. Of the comparative PM techniques, only
root-cause analysis is mentioned [20], [39], [41]. Only [19]
uses PM techniques to create the simulation model, and RPA
is mentioned in [60], and [39].
A. Discussion SLR-2
Considering the capabilities of PM and the processes mainly
affected by the CE, it is not surprising that most publications
describe the application of PM to increase efficiency and
resource availability in a manufacturing context. Increasing
efficiency and availability lowers the demand for additional
resources and reduces waste, emissions, and energy con-
sumption [25]. Many identified publications acknowledge the
application of PM techniques for the holistic management
of production systems due to their capability of end-to-end
process analysis. As most publications relate to production
management, we quickly see that the considered resources
are those the manufacturing industry works with, not the
products that are being worked on. This means that the main
contributions to the CE the publications of SLR-2 describe
are lowering demand for industrial goods and reducing excess
emissions, waste and energy consumption of practices as they
are. This undoubtedly contributes to the CE, yet its impact
is limited as industrial goods have significantly less influence
on environmental quality than consumer goods [80]. However,
PM’s capabilities to support rethinking BPs regarding energy
efficiency, material input, emissions and waste production are
also highlighted. Although most do not provide much detail
on how this can be done, the overall tenor of all publications
is that PM can support this transition. Based on this result,
we can answer the first research question concerning PM’s

contribution to SD: PM supports sustainability assessment
and the management of relevant processes. It also offers
methods to detect the potential for improvement in sustain-
ability, monitor processes about sustainability targets, compare
the sustainability of different procedures, and identify drivers
for high emissions, waste and energy consumption.
Regarding the processes mentioned in current literature
related to PM and the CE, we see a heap around production
management. The low coverage of consumer processes is
expected, as PM relies on the analysis of event data of IT
systems, and consumer behaviour regarding physical products
is rarely logged and even rarer evaluated systematically. Only
a single publication explicitly refers to a process not part
of the traditional SC. The additional challenges related to
reverse logistics, i.e., managing a material flow with high
variance in required operations [26], are not addressed, despite
PM’s strength in supporting the management of less struc-
tured processes. Furthermore, the application of PM regarding
intra-logistical or SC-related topics is not mentioned. This is
unexpected, as logistical topics are highly relevant [33] and
increased digitalisation and automation benefit the CE [40].
We also notice that only intra-organisational processes are
mentioned, despite the CE and, with it, new laws and reg-
ulations emphasise the consideration of the entire SC and
long product life. We see two likely reasons for the strong
emphasis on production systems yet the near neglect of all
forms of circularity, logistics and extended processes: (1) The
CE concept is not well-established yet [45], so the processes
PM could support may not be widespread enough. (2) Inter-
organisational PM as a discipline is still in its infancy [76] and
PM has mainly been adopted to increase the performance of a
single organisation’s [67] processes. Therefore, PM for inter-
organisational and logistical process analysis form research
gaps for PM in sustainability. Also, challenges arising from
the new types of processes have to be analysed to support the
development of targeted PM techniques.
Interestingly, the mentioned PM techniques and applications
show that the full potential of PM is not used. For exam-
ple, [38] neither applies nor discusses PM’s techniques for
identifying constrained resources, despite bottleneck detection
being the objective of the presented work. The proposals for
PM’s application for carbon assessment [7] and LCA [50] also
indicate that state-of-the-art PM has not been considered. Both
enlist challenges that the application of OCPM could solve,
such as PM’s limitation to a single type of case, processes
not representing the material flow but merely the control flow
and the required addition of resource data to each event. This,
as well as the few search results, leads us to conclude that
knowledge about state-of-the-art PM has yet to reach the
domain of SD.
VI. PRO CE SS MI NI NG F OR SU STAINA BI LI TY
Exploring the current use of PM for sustainability helped
us identify the processes relevant to the CE that must be
supported and the contributions PM can provide. However,
discrepancies in the identified potential of PM on the one
Fig. 5: Summary of the basic PM4S framework as basis for
additional process analysis to increase SD.
hand, and its described application on the other have become
apparent. We, therefore, map the contributions to SD men-
tioned in the results of SLR-2 to the different areas of PM to
provide a structured answer to RQ1 before going into more
detail on how this could be realised. Table II summarises PM’s
contributions PM to SD. PM techniques support detecting and
monitoring current practices using business data and establish-
ing a direct connection to SD assessment. To do so, different
elements of the process have to be considered and depicted
in the discovered models. This allows companies to use PM
techniques to keep track of SD regulations, goals, benchmarks
and standards. While driving sustainability, the toolset for
performance analysis can be used to establish a lever between
process performance and sustainability. Additionally, PM can
be used for comparisons of alternatives and to identify the
causes of high environmental impact. These results can support
business decisions, model the impacts of consumer behaviour
or product life extension strategies and increase digitalisation
and automation.
To guide existing research in applying PM for sustainability
and target gaps more effectively, we introduce the Process
Mining for Sustainability (PM4S) framework as an answer
to RQ2. The main principles of the CE are the efficient use
of resources, the extension of product life and the valuable
application of materials in end-of-life products [54] as well
as the reduction of emissions and energy consumption [24]
throughout the entire product life cycle [16]. We first consider
the processes and perspectives PM techniques must capture
and suggest a framework for assessing the environmental
sustainability of processes before stringing it together in VI-B.

TABLE II: PM’s contribution to sustainability based on the
results of SLR-2.
Discovery
Raw Material and Components
Intermediate and final Products
Process Ressources
Wastes
Conformance Checking Impact of Abnormal Behaviour
Checking Compliance to SD Goals
Process Analysis
Impact and Circularity Assessment
Performance Analysis
Bottleneck Detection
Comparative PM
Comparison of Alternatives
Trade offs between SD aspects
Root Cause Analysis
Predictive PM
Predict the Impact of Behaviour
Predicting Goal Achievement
Simulation
Action-oriented PM Automation
A. Perspectives
The flow of all raw materials, intermediate products, final
products, and waste lies at the core of the CE. The processes
relevant to the CE were introduced in the previous section and
are displayed in Figure 4. Currently, PM techniques are mainly
applied to analyse and support value-adding processes, such
as order management [58]. Additionally, the two SLRs have
revealed PM’s potential for maintenance processes. To assess
sustainability PM techniques must be applied to processes
including all relevant activities, including waste management
and logistical operations such intra- and intra-organisational
transportation and inventory management [77].
This means, the log must include events and objects asso-
ciated with the raw materials, intermediate, final, and waste
products. As the material flow of these products is inter-
dependent due to logical dependencies, shared resources or
batching PM4S requires OCPM, as only considering a single
case does not allow for a realistic analysis of the involved
system’s complexity. The current OCEL standard requires each
object to have a unique identifier, and PM techniques focus
on “tracking” individual objects through the process. For PM
techniques to be more helpful to logistical processes, they
have to allow the consideration of object quantities instead
of object identities and collections of interchangeable objects.
To consider inter-organisational processes, the data silos from
all involved parties must be broken down to create and analyse
a joint process model.
As process resources strongly impact the environment, e.g.
through produced emissions and their own life cycle, their util-
isation, as well as their own lifecycle, has to be captured. We
see two possibilities for including process resources in OCPM:
They can be added to the event log as an event attribute or con-
sidered objects. Adding them as event attributes neither allows
a direct connection to additional information, such as their
settings, nor enables the consideration of maintenance and
repair processes. Modelling resources as objects overcomes
these obstacles, yet the current OCEL standard does not allow
object attributes to change over time, which would support the
energy efficiency analysis of machine configurations similar to
the one described in [51]. Additionally, the resulting models
could lack conciseness and potentially include many loops
if an activity requiring the resource is performed several
times. Process resources could be considered as a second,
separate class of objects with a process relating to the resource
status (similar to [41]) associated with the control flow of
objects. As the options in line with the current OCEL standard
seem unsatisfactory, yet the consideration of process resources
is essential to sustainability, we consider the integration of
process resources into PM4S as a potential for further research.
B. Sustainability Assessment
The aim of PM4S is not to merely increase process effi-
ciency, which consequently leads to greater sustainability, but
to increase the sustainability of BPs themselves. Therefore,
assessing SD indicators through PM is not only an anticipated
benefit of applying PM (see SLR-2) but forms the main
contribution of PM4S. The environmental impact of a process
can only be monitored and considered in process changes if
incorporated into process management. Below, we consider
how the assessment of material input (i.e. circulation of a
product), emissions, energy consumption and waste can be
included in PM. There are several ways of measuring the
circularity of a product [10] – all of which require additional
information on the involved components, for example, toxicity,
scarcity, whether it is recycled material or the members of
already replaced. This information has to be added to the
event log, for example, as attributes to each object entering
the considered system. In the publications of SLR-2, emission
data is considered per event. The produced emissions per event
must be extracted from business data or collected via an API to
emission databases. The data can also be collected per activity
and added as an event attribute in the preprocessing of the
event log. If all data for the API calls are available in the
OCEL, adding the emission data to the individual activities or
events can also be integrated to process discovery techniques.
The same holds for energy data. The waste assessment is
only possible if each waste object is indicated as such by an
object attribute and its quantity (potentially as a fraction of the
input) is given in the object attributes. Using this information,
slightly adapted discovery techniques, as suggested in [50],
or an additional pre-processing step for other event attributes
allows the consideration of the progression of SD indicators
throughout the process. Their assessment requires the con-
sideration of inter-attribute dependencies. Figure 6 depicts
an example illustrating these dependencies. Event 1 occurs
before Event 2, each describing the produced emissions and
consumed energy as attributes. Leaning on LCA, we introduce
counters for each object, describing the energy consumed
and emissions created so far. Event 1 works on object a-b-c,
subsequently used in event 2. As event 1 produces emissions,
the amount of created emissions associated with object a-b-c
when consumed for event 1 is higher than before event 1 was
performed. Apart from object a-b-c, event 2 consumes object

Fig. 6: Example for the assessment of environmental sustainability factors per object.
d-e-f, producing a waste object of the same type as object a-b-
c and a product called object g-h-i. The emissions associated
with product g-h-i depend on the emissions produced during
event 2 and the counter of its components. The fact that object
a-b-c was not fully may also be taken into account. Similarly,
counters for circularity indices and energy consumption could
be computed.
This consideration requires some additional concepts for the
resulting PM techniques. For example, the described approach
requires distinguishing whether an object associated with an
event is merely used, (partially) consumed or created in the
event. As the individual objects are only assigned to events
without qualifying their relationship, the distinction could be
made by considering previous and future events. Also, rules for
calculating the sustainability indicators considering conversion
and division of products and waste creation. As we can see,
PM offers the possibility of creating these assessments, yet
the specific details on how these counters must be calculated
and integrated into PM techniques require further (interdisci-
plinary) research.
C. PM4S Framework
After deriving the contributions PM can make, the per-
spectives that must be captured and the assessment of the
fundamental sustainability factors, we now combine them into
a framework as depicted in Figure 5. As pointed out above,
object-centricity is vital to PM4S. The event log has to capture
all involved material flows (incl. waste), the involved process
resources, value-adding as well as logistical activities. Addi-
tionally, each event requires attributes on created emissions
and energy consumption. Apart from the event data, the objects
have to be enriched with further details, such as informa-
tion on materials, origins, information qualifying objects as
waste/resources and the settings of process resources. Using
this data, an object-centric process model can be created,
including basic information on the system’s and every object’s
emissions, energy consumption, and degree of circularity. The
resulting model and the sustainability assessment can then
be used as a basis for standard PM techniques focussing
on performance and improvement as well as new techniques
checking a process’ compliance to benchmarks, targets etc.
VII. CON CL US IO N
Business processes are the backbone of any organisation
and impact environmental, social, and economic sustainability.
When considering the application of PM for SD, we see that
the support for social equity is limited, yet the application
to human-centred processes is beneficial. For environmental
quality, we identify a strong focus on the manufacturing
industry and increasing efficiency. The identified publications
mainly provide high-level descriptions. We recognise that PM
techniques are not used to their full potential and conclude a
lack of knowledge of the application and benefits of state-
of-the-art PM techniques. Based on these SLRs, we have
derived a PM4S framework for the application to increase
SD in business processes. In this framework, the processes
and process elements essential to environmental sustainability,
as well as additional relevant information is used to build
an object-centric process model enriched with environmental
impact assessment metrics. This model is used as the basis
for additional process analysis and improvement, to ensure
that sustainability lies at the heart of all process changes.
PM4S’ main contributions lie in its ability of shedding light
on increasingly complex business processes, enabling end-to-
end process management and assessing process sustainability.
It supports decision-makers in comparing alternative practices,
enables simulations and digital shadows and highlights devia-
tions to given benchmarks.
OCPM is a novel branch of the fast-growing and fast-
developing discipline of PM [74], and PM4S poses an im-
portant use case to this new branch underlining its necessity.
However, as shown in the previous section that the ignorance
between the PM and SD domains is mutual – the require-
ments for assessing and improving sustainability highlight
certain research gaps. On the one hand, we identify further
research in the support of logistical process management (i.e.,
regarding federated PM and requiring unique, non-changing
object identifiers) and the consideration of process ressource
management. On the other hand, the capabilities of integrating
and using additional information on events, qualifying objects
and considering changing attributes need to be refined. The
intersection between the two domains is still a scientific
void offering much potential for additional, interdisciplinary
research.
ACK NOW LE DG EM EN TS
Funded by the Deutsche Forschungsgemeinschaft (DFG,
German Research Foundation) under Germany’s Excellence
Strategy - EXC-2023 Internet of Production - 390621612. We
also thank the Alexander von Humboldt (AvH) Stiftung for
supporting our research.

REF ER EN CE S
[1] Federica Acerbi, Adalberto Polenghi, Walter Quadrini, Marco Macchi,
and Marco Taisch. Fostering Circular Manufacturing Through the
Integration of Genetic Algorithm and Process Mining. In Advances
in Information and Communication Technology (IFIP), volume 664.
Springer, 2022.
[2] Elisa Achterberg, Jeroen Hinfelaar, and Nancy Bocken. Master Circular
Business with the Value Hill. White Paper, Circle Economy, 2016.
[3] Aser Alaa Ahmed, Mohammad A. Nazzal, Basil M. Darras, and I.M.
Deiab. A comprehensive multi-level circular economy assessment
framework. Sustainable Production and Consumption, 32:700–717,
2022.
[4] Luciano Batista, Michael Bourlakis, Palie Smart, and Roger Maull. In
search of a circular supply chain archetype – a content-analysis-based
literature review. Production Planning & Control, 29(6):438–451, 2018.
[5] Jos`
ephe Blondaut. Driving sustainability through process
optimization. Blog post, PEX – Process Excellence Network,
2022. https://www.processexcellencenetwork.com/business-
transformation/articles/driving-sustainability-through-a-focus-on-
people-planet-and-profit.
[6] Nancy M. P. Bocken, Ingrid de Pauw, Conny Bakker, and Bram van der
Grinten. Product design and business model strategies for a circular
economy. Journal of Industrial and Production Engineering, 33(5):308–
320, 2016.
[7] Lars Brehm, Jessica Slamka, and Andreas Nickmann. Process Mining
for Carbon Accounting: An Analysis of Requirements and Potentials. In
Digitalization Across Organizational Levels, pages 209–244. Springer,
Cham, 2022.
[8] Amelia B˘
adic˘
a, Costin B˘
adic˘
a, Ion Buligiu, and Liviu-Ion Ciora. Explor-
ing the Usability of Process Mining in Smart City. IFAC-PapersOnLine,
55(11):42–47, 2022.
[9] Craig R. Carter and Dale S. Rogers. A framework of sustainable supply
chain management: moving toward new theory. International Journal of
Physical Distribution & Logistics Management, 38(5):360–387, 2008.
[10] Blanca Corona, Li Shen, Denise Reike, Jes´
us Rosales Carre´
on, and Ernst
Worrell. Towards sustainable development through the circular econ-
omy—A review and critical assessment on current circularity metrics.
Resources, Conservation and Recycling, 151:104498, 2019.
[11] Li Du, Long Cheng, and Cong Liu. Process Mining for Wind Turbine
Maintenance Process Analysis: A Case Study. In IEEE Conference
on Energy Internet and Energy System Integration: Energy Internet for
Carbon Neutrality, pages 3274–3278. IEEE Inc., 2021.
[12] Marlon Dumas, Marcello La Rosa, Jan Mendling, and Hajo A. Reijers.
Fundamentals of Business Process Management. Springer, 2nd edition,
2018.
[13] Ambre Dupuis, Cam´
elia Dadouchi, and Bruno Agard. Predicting
crop rotations using process mining techniques and Markov principals.
Computers and Electronics in Agriculture, 194, 2022.
[14] John Elkington. Cannibals with forks: the triple bottom line of 21st
century business. Conscientious commerce. New Society Publishers,
1998.
[15] Ellen MacArthur Foundation. Circular Economy Introduction. Technical
report, Ellen MacArthur Foundation, 2021.
[16] European Commission, Joint Research Centre, and Institute for Envi-
ronment and Sustainability. International Reference Life Cycle Data
System (ILCD) Handbook - General guide for Life Cycle Assessment
- Provisions and Action Steps. Technical Report EUR 24378 EN,
Publications Office, Luxembourg, 2010.
[17] Anahita Farhang Ghahfarokhi, Gyunam Park, Alessandro Berti, and Wil
M. P. van der Aalst. OCEL Standard, 2022.
[18] Deena P. Francis, Sanja Lazarova-Molnar, and Nader Mohamed. To-
wards Data-Driven Digital Twins for Smart Manufacturing. In nterna-
tional Conference on Systems Engineering, ICSEng, volume 182, pages
445–454. Springer Science and Business Media Deutschland GmbH,
2021.
[19] Jonas Friederich, Deena P. Francis, Sanja Lazarova-Molnar, and Nader
Mohamed. A framework for data-driven digital twins for smart manu-
facturing. Computers in Industry, 136, 2022.
[20] Andreas Fritsch, Johanna Hammerstein, Clemens Schreiber, Stefanie
Betz, and Andreas Oberweis. Pathways to Greener Pastures: Research
Opportunities to Integrate Life Cycle Assessment and Sustainable Busi-
ness Process Management Based on a Systematic Tertiary Literature
Review. Sustainability (Switzerland), 14:11164, 2022.
[21] Gabriela Galic and Marcel Wolf. Global Process Mining Survey 2021 -
Delivering Value with Process Analytics – Adoption and Success Factors
of Process Mining. Technical report, Deloitte Global, 2021.
[22] Cleiton dos Santos Garcia, Alex Meincheim, Elio Ribeiro Faria Junior,
Marcelo Rosano Dallagassa, Denise Maria Vecino Sato, Deborah Ribeiro
Carvalho, Eduardo Alves Portela Santos, and Edson Emilio Scalabrin.
Process mining techniques and applications – A systematic mapping
study. Expert Systems with Applications, 133:260–295, 2019.
[23] Martin Geissdoerfer, Sandra Naomi Morioka, Marly Monteiro de Car-
valho, and Steve Evans. Business models and supply chains for the
circular economy. Journal of Cleaner Production, 190:712–721, 2018.
[24] Martin Geissdoerfer, Paulo Savaget, Nancy M.P. Bocken, and Erik Jan
Hultink. The Circular Economy – A new sustainability paradigm?
Journal of Cleaner Production, 143:757–768, 2017.
[25] Patrizia Ghisellini, Catia Cialani, and Sergio Ulgiati. A review on
circular economy: the expected transition to a balanced interplay of
environmental and economic systems. Journal of Cleaner Production,
114:11–32, 2016.
[26] Paul Goodall, Richard Sharpe, and Andrew West. A data-driven sim-
ulation to support remanufacturing operations. Computers in Industry,
105:48–60, 2019.
[27] Thomas E. Graedel, Julian Allwood, Jean-Pierre Birat, Matthias Buchert,
Christian Hagel¨
uken, Barbara K. Reck, Scott F. Sibley, and Guido
Sonnemann. What Do We Know About Metal Recycling Rates? Journal
of Industrial Ecology, 15(3):355–366, 2011.
[28] Teresa Guarda, Filipe Mota Pinto, Juan Pablo Cordova, Fernando Mato,
Geovanni Ninahualpa Quina, and Maria Fernanda Augusto. Pervasive
business intelligence as a competitive advantage. In Iberian Confer-
ence on Information Systems and Technologies, CISTI. IEEE Computer
Society, 2016.
[29] Eva Heiskanen. The institutional logic of life cycle thinking. Journal
of Cleaner Production, 10(5):427–437, 2002.
[30] Anaisa Hern´
andez Gonz´
alez, Coral Calero, Dianelys P´
erez Parra, and
Javier Mancebo. Approaching Green BPM characterisation. Journal of
Software: Evolution and Process, 31(2):e2145, 2019.
[31] Julia Horsthofer-Rauch, Susanne. Vernim, and Gunther Reinhart.
Sustainability-focused digital value stream mapping: Concept for em-
ploying process mining for sustainability-focused value stream mapping.
ZWF Zeitschrift f¨
ur Wirtschaftlichen Fabrikbetrieb, 116(9):590–593,
2021.
[32] Hyunjei Jo, Sang Do Noh, and Yongju Cho. An agile operations
management system for green factory. International Journal of Precision
Engineering and Manufacturing - Green Technology, 1(2):131–143,
2014.
[33] Vivianne Julianelli, Rodrigo Goyannes Gusm˜
ao Caiado, Luiz Felipe
Scavarda, and Stephanie Pinto de Mesquita Ferreira Cruz. Interplay
between reverse logistics and circular economy: Critical success factors-
based taxonomy and framework. Resources, Conservation and Recy-
cling, 158:104784, 2020.
[34] Marc Kerremans, Samantha Searle, and Kimihiko Tushar, Srivas-
tava Iijima. Market Guide for Process Mining. Market Guide
G00737056, Gartner Inc., 2021.
[35] Julian Kirchherr, Denise Reike, and Marko Hekkert. Conceptualizing
the circular economy: An analysis of 114 definitions. Resources,
Conservation and Recycling, 127:221–232, 2017.
[36] Mathias Kirchmer, Francisco Guti´
errez, and Sigifredo Laengle. Process
mining for organizational agility. Industrial Management (Norcross,
Georgia), 52(1):19–24, 2010.
[37] Barbara Kitchenham and Stuart Charters. Guidelines for performing
Systematic Literature Reviews in Software Engineering. Technical
report, Technical report, EBSE Technical Report EBSE-2007-01, 2007.
[38] Mahesh Kumbhar, Amos H.C. Ng, and Sunith Bandaru. Bottle-
neck Detection Through Data Integration, Process Mining and Factory
Physics-Based Analytics. In Advances in Transdisciplinary Engineering,
volume 21 of Swedish Production Symposium, SPS, pages 737–748. IOS
Press BV, 2022.
[39] Federico A. Lievano-Mart´
ınez, Javier D Fern´
andez-Ledesma, Daniel
Burgos, John W. Branch-Bedoya, and Jovani A. Jimenez-Builes. Intel-
ligent Process Automation: An Application in Manufacturing Industry.
Sustainability (Switzerland), 14(14), 2022.
[40] Ana Beatriz Lopes de Sousa Jabbour, Charbel Jose Chiappetta Jabbour,
Moacir Godinho Filho, and David Roubaud. Industry 4.0 and the
circular economy: a proposed research agenda and original roadmap for

sustainable operations. Annals of Operations Research, 270(1-2):273–
286, 2018.
[41] Jonas Maeyens, Annemie Vorstermans, and Mathias Verbeke. Process
mining on machine event logs for profiling abnormal behaviour and root
cause analysis. Annals of Telecommunications, 75(9-10):563–572, 2020.
[42] Pravin Kumar Mallick, Kim Bang Salling, Daniela C.A. Pigosso, and
Tim C. McAloone. Closing the loop: Establishing reverse logistics
for a circular economy, a systematic review. Journal of Environmental
Management, 328:117017, 2023.
[43] Niels Martin, Nils Wittig, and Jorge Munoz-Gama. Using Process
Mining in Healthcare. In Process Mining Handbook, Lecture Notes
in Business Information Processing, pages 416–444. Springer, Cham,
2022.
[44] Nigel P. Melville. Information Systems Innovation for Environmental
Sustainability. MIS Quarterly, 34(1):1–21, 2010.
[45] Roberto Merli, Michele Preziosi, and Alessia Acampora. How do
scholars approach the circular economy? A systematic literature review.
Journal of Cleaner Production, 178:703–722, 2018.
[46] Gerry Murphy. Process “Prospecting” to Improve Renewable Energy
Interconnection Queues: A Case Study. In International Conference
on Process Mining, ICPM, volume 406 LNBIP, pages 30–42. Springer,
2021.
[47] United Nations. Report of the World Commission on Environment
and Development: ”Our common future”. Technical report, United
Nations, Secretary-General and World Commission on Environment and
Development, New York, USA, 1987.
[48] Lia Ninda Safitri, Riyanarto Sarno, and Gita Intani Budiawati. Improving
Business Process by Evaluating Enterprise Sustainability Indicators
Using Fuzzy Rule Based Classification. In International Seminar
on Application for Technology of Information and Communication,
iSemantic, pages 55–60. IEEE Inc., 2018.
[49] S´
en´
echal Olivier, Sandro P. Pires, Eduardo R.F. Loures, Eduardo A.P.
Santos, and Jos´
e M.P.A. Cestari. Knowledge management for sustainable
performance in industrial maintenance. In IIE Annual Conference and
Expo, pages 2191–2200. Institute of Industrial Engineers, 2015.
[50] Christian Ortmeier, Nadja Henningsen, Adrian Langer, Alexander
Reiswich, Alexander Karl, and Christoph Herrmann. Framework for
the integration of Process Mining into Life Cycle Assessment. In The
28th CIRP Conference on Life Cycle Engineering, March 10 – 12, 2021,
Jaipur, India, volume 98, pages 163–168, 2021.
[51] Jens Otto, Birgit Vogel-Heuser, and Oliver Niggemann. Online pa-
rameter estimation for cyber-physical production systems based on
mixed integer nonlinear programming, process mining and black-box
optimization techniques. At-Automatisierungstechnik, 66(4):331–343,
2018.
[52] Radu V. Pascu. Particularities of implementing a sustainable man-
agement system in universities. Academic Journal of Manufacturing
Engineering, 13(2):104–111, 2015.
[53] Anastasiia Pika, Arthur H.M. ter Hofstede, Robert K. Perrons, Georg
Grossmann, Markus Stumptner, and Jim Cooley. Using Big Data to
Improve Safety Performance: An Application of Process Mining to
Enhance Data Visualisation. Big Data Research, 25, 2021.
[54] Jos´
e Potting, Marko Hekkert, Ernst Worrell, and Aldert Hanemaaijer.
Circular Economy: Measuring innovation in the product chain. Technical
Report 2544, PBL Netherlands Environmental Assessment Agency,
2017.
[55] Ben Purvis, Yong Mao, and Darren Robinson. Three pillars of sus-
tainability: in search of conceptual origins. Sustainability Science,
14(3):681–695, 2019.
[56] Christian Rahnsch. How to get from Sustainability Insights to Process
Action. SAP Company Website, SAP - Business Trends, 2022.
[57] Denise Reike, Walter J.V. Vermeulen, and Sjors Witjes. The circular
economy: New or Refurbished as CE 3.0? — Exploring Controversies
in the Conceptualization of the Circular Economy through a Focus on
History and Resource Value Retention Options. Resources, Conservation
and Recycling, 135:246–264, 2017.
[58] Lars Reinkemeyer. Process Mining in Action: Principles, Use Cases
and Outlook. Springer, Cham, 2020.
[59] Lars Reinkemeyer. How Process Mining can turn Sustainability
Targets into Action. LinkedIn Article, Celonis, 2022.
https://www.linkedin.com/pulse/how-process-mining-can-turn-
sustainability-targets-lars-reinkemeyer/.
[60] Stefanie Rinderle-Ma and Juergen Mangler. Process Automation and
Process Mining in Manufacturing. In International Conference on
Business Process Management, volume 12875 LNCS. Springer, 2021.
[61] Hannah Ritchie, Max Roser, and Pablo Rosado. CO2 and greenhouse
gas emissions. Our World in Data, 2020. https://ourworldindata.org/co2-
and-greenhouse-gas-emissions.
[62] Daniel Ritter. From network mining to large scale business networks.
In International Conference on World Wide Web, pages 989–996. ACM,
2012.
[63] Edson Ruschel, Eduardo Alves Portela Santos, and Eduardo de Fre-
itas Rocha Loures. Mining Shop-Floor Data for Preventive Maintenance
Management: Integrating Probabilistic and Predictive Models. In Pon-
tificia Universidade Catolica do Parana, volume 11, pages 1127–1134,
2017.
[64] Edson Ruschel, Eduardo Alves Portela Santos, and Eduardo de Fre-
itas Rocha Loures. Establishment of maintenance inspection intervals:
an application of process mining techniques in manufacturing. Journal
of Intelligent Manufacturing, 31(1):53–72, 2020.
[65] Clemens Schreiber. Automated Sustainability Compliance Checking
Using Process Mining and Formal Logic. In International Conference
on ICT for Sustainability, pages 181–184. ACM, 2020.
[66] Stefan Seidel, Jan Recker, and Jan vom Brocke. Green Business Process
Management. In Green Business Process Management, pages 3–13.
Springer, 2012.
[67] Malte Thiede, Daniel Fuerstenau, and Ana Paula Bezerra Barquet. How
is process mining technology used by organizations? A systematic
literature review of empirical studies. Business Process Management
Journal, 24(4):900–922, 2018.
[68] UNEP. Guidelines for social life cycle assessment of products and
organisations 2020. United Nations Environment Programme, 2020.
[69] General Assembly United Nations. Transforming our World: The 2030
Agenda for Sustainable Development. Resolution A/RES/70/1, United
Nations, New York, USA, 2015.
[70] Wil M. P. van der Aalst. Process Mining: X-Ray Your Business
Processes. CACM 99, page 11, 2011.
[71] Wil M. P. van der Aalst. Process Mining. Springer, 2016.
[72] Wil M. P. van der Aalst. Responsible Data Science: Using Event Data in
a “People Friendly” Manner. In Enterprise Information Systems, volume
291, pages 3–28, Cham, 2017. Springer.
[73] Wil M. P. van der Aalst. Object-Centric Process Mining: Dealing with
Divergence and Convergence in Event Data. In International Conference
on Software Engineering and Formal Methods, volume 11724, pages 3–
25, Cham, 2019. Springer.
[74] Wil M. P. van der Aalst. Process Mining: A 360 Degree Overview. In
Process Mining Handbook, volume 448, pages 3–34. Springer, Cham,
2022.
[75] Wil M. P. van der Aalst, Hajo A. Reijers, and Minseok Song. Discover-
ing Social Networks from Event Logs. Computer Supported Cooperative
Work (CSCW), 14(6):549–593, 2005.
[76] Wil M.P. van der Aalst. Federated Process Mining: Exploiting Event
Data Across Organizational Boundaries. In IEEE International Confer-
ence on Smart Data Services (SMDS), pages 1–7. IEEE, 2021.
[77] Dennis Vegter, Jos van Hillegersberg, and Matthias Olthaar. Supply
chains in circular business models: processes and performance objec-
tives. Resources, Conservation and Recycling, 162:105046, 2020.
[78] Philipp von Bieberstein. Climatiq partners with Celonis to drive
embedded enterprise carbon intelligence. climatiq Company Website,
blog post, Climatiq Technologies GmbH, November 2022.
[79] Sven Wohlgemuth. Is Privacy Supportive for Adaptive ICT Systems?
In Proceedings of the 16th International Conference on Information
Integration and Web-based Applications & Services, iiWAS ’14, pages
559–570, New York, USA, 2014. ACM.
[80] Agata Wolk-Lewanowicz. INSIGHT: Carbon foot-
print labelling – a growing trend among consumer
goods companies. Technical report, ICIS News, 2020.
https://www.icis.com/explore/resources/news/2020/07/17/10531480/carbon-
footprint-labelling-a-growing-trend-among-consumer-goods-
companies/.
[81] Pierluigi Zerbino, Davide Aloini, Riccardo Dulmin, and Valeria
Mininno. Towards analytics-enabled efficiency improvements in mar-
itime transportation: A case study in a mediterranean port. Sustainability
(Switzerland), 11(16), 2019.