MOSAIC: A Structured Multidisciplinary Analysis for Managing the Integration of Inter-Organizational Changes

Abstract

Sociotechnical systems like railways are becoming increasingly complex as numerous changes are constantly being integrated to improve or maintain desired performances. A crucial aspect enabling this integration is to define the scope of the changes, referred to as the system of interest. This can be challenging for engineering managers due to mounting system complexity caused by: an inter-organizational environment, a multitude of diverse stakeholders, the multidomain environment, and an increasing number of interdependencies. By employing Design Science Research, this study proposes an analysis that supports engineering managers in managing sociotechnical and inter-organizational change integration (MOSAIC) by facilitating a structured, multidisciplinary assessment of system impacts and interdependencies. The design was evaluated by applying it to an inter-organizational project in the Dutch railway system. The findings show that with MOSAIC, a system of interest can be collectively defined using the proposed process and a multidisciplinary and inter-organizational group of experts, creating mutual understanding. MOSAIC can further aid engineering managers in front-end project planning.

Keywords:

• Interdependencies
• Inter-Organizational
• Multidisciplinary
• Project Management
• Sociotechnical System
• System Integration

EMJ Focus Areas:

• Operations Management
• Program & Project Management
• Systems Engineering
• Technology Management

Introduction

Rapid change, globalization, fierce competition, rising customer expectations, and rapid advancement of technology characterize current society (Rasmussen, 1997). Because of the advancement of technology, the connectivity of different systems increases, which supports engineering managers (EMs) in achieving a more optimized performance. In addition to the abovementioned trends, numerous operations now require cooperation between employees from different organizations, as their operations are increasingly inter-organizational (Milch & Laumann, 2016). These developments give rise to complex sociotechnical systems and add to system complexity (Perrow, 1999), as Geels (2002) illustrated in Exhibit 1.

Exhibit 1. Actors Involved in Sociotechnical Systems (Geels, 2002)

In complex sociotechnical systems, various domains, such as processes, personnel, capacity, technical systems, and rules and regulations, align with existing technology (Clemson & Lowe, 1993; Geels, 2002). However, because of this, new modifications often do not match existing sociotechnical frameworks and are often difficult to integrate and establish (Geels, 2002). Therefore, there is an increasing need for EMs to understand the complexity to identify how changes and the projects that strive to realize them (Potts et al., 2021) can be integrated into this context. As such, EMs must pay significant attention to inter-organizational and multidomain environments, interconnections, and interdependent elements, thus increasing the scope of integration.

The broadening scope of system integration (SI) poses many challenges for EMs:

1. While the SI phase, where all unexpected and unforeseen problems surface, is systematically underestimated in conventional projects (Muller, 2007), the number of unexpected and unforeseen problems increases in large-scope projects.

2. More organizations, suppliers, users, and processes are included when SI increases scope. More inter-organizational collaborations are required, making the SI phase significantly more complicated (Muller, 2007).

3. In practice, crosscutting functionality and quality suffer from decomposition. As the scope broadens, these challenges increase (Muller, 2007).

4. In large projects, there can be a lack of ownership and communication across organizational boundaries (Muller, 2007; Ramtahalsing et al., 2020).

5. The stakeholders can view complex systems differently (Hagan et al., 2011). As the project increases in scope, more stakeholders are to be included.

6. The number and type of interfaces also increase as the SI scope is broadened.

EMs need to pay continuous selective attention to integration in order to meet these challenges (Muller, 2007), especially regarding what specifically needs to be integrated. Therefore, determining the scope of the change is vital for setting bounds on different aspects considered to be of interest, which is defined as the system of interest (SoI). Railway systems are examples of such sociotechnical systems where changes are constantly implemented to upgrade system performance, and insight into SoIs is important for EMs to achieve this.

During the past few decades, there has been an increased focus on seamless railway mobility across borders, especially in Europe. Existing railway systems are adopting novel technologies to achieve mobility within and between countries, particularly in control and communications (Dumolo, 2007). These developments require integration to improve the functionality and efficiency of the railway system as a whole (Dumolo, 2007) and require collaboration between various multidisciplinary and (non)engineering stakeholders. Depending on the change to be integrated, these can include engineering, logistics, procurement, finance, operations, configuration management, safety, and risk management.

The European Railway Agency (2009) mentions the need to describe changes: ’‘When incorporating a new element into the railways, or modifying an existing one, the change must be clearly and completely described, as well as limits of the railway system where the change is integrated (whether technical, operational, or organizational).’’

According to the European Union Agency for Railways (2020), an exhaustive overview of the SoI is most pertinent to and critical for comprehensive risk management and safe integration. This provides insight into the change and should describe all interfaces/interdependencies between the change, human operators, or subsystems clearly and completely. As a result, considerable attention is paid to SoIs from the safety perspective within the railway industry.

However, the scope of integration in complex inter-organizational systems is not limited to safety. As systems are continuously changing, adequately sharing information between involved organizations is a prerequisite for adequate management of the railway system because a change in one part may affect other parts of the system (Guillerm et al., 2012). These systemic changes can vary widely in nature, and the impacts of these changes tend to affect multiple domains spanning different organizations. This not only broadens the scope, as mentioned but also makes the SoI more ambiguous for EMs (Potts et al., 2021), which can result in several challenges:

1. Create confusion and errors, for example, when different researchers or practitioners work on the same problem related to an assumed SoI (Cumming & Collier, 2005).

2. Cause challenges to effective communication and understanding of interfaces for example (Haskins & Striegel, 2006).

3. The people element of each system/subsystem can cross boundaries and blur distinctions (Wilson, 2014).

4. Discussions concerning scope (Dasher, 2003).

This makes managing changes increasingly difficult for EMs but also shows the necessity of the scope definition, the SoI, and describing what exactly is required to be integrated.

Research Aim

This study aims to support EMs in managing inter-organizational change integration in the sociotechnical railway system by designing an expert-based analysis that aids in jointly creating insight into high-level inter-organizational SoIs.

The remainder of this paper is organized as follows: The next section presents a comprehensive literature review. The methodology section describes the methodology that resulted in the design of the management of sociotechnical and inter-organizational change integration (MOSAIC) analysis. The ‘Design of MOSAIC’ section presents the MOSAIC analysis, and the ‘Demonstration of MOSAIC’ section demonstrates this through application to a case study. This is followed by the ‘Evaluation & discussion’ section, which discusses the implications of this paper for EMs, and finally, it concludes with the contributions.

Literature Review

This section provides an overview of the current state of research and describes where the research fits in the overall body of work in engineering management. After this, it focuses on several characteristics related to the present inter-organizational and sociotechnical context that require attention to support EMs in managing change integration. Based on these characteristics, several subgoals and related design principles are derived, which will be included in the design and development of MOSAIC at a later stage.

Change projects in complex systems can significantly impact other processes or departments across organizations; thus, it is vital to set bounds on different aspects to be of interest. Current literature describes this as project scope definition (Fageha & Aibinu, 2013) providing information for identifying the work which needs to be performed. Defining project scope using input from all stakeholders is a vital task that needs to be adequately carried out at an early stage (Dasher, 2003). While adequate front-end project planning with a clear project scope definition can avoid negative effects on project performance, inadequate project planning and poor scope definition can lead to expensive changes, delays, rework, cost overruns, schedule overruns, and project failure (Fageha & Aibinu, 2013). Therefore, a well-defined scope during the front-end planning stage is crucial for successful project execution and achieving a satisfactory project outcome (Fageha & Aibinu, 2013).

While a change may appear simple at first glance, this can become more complex when attention is paid to the context in which it must be integrated. There have been concerns about the appropriateness of traditional tools and techniques developed for simple projects for use in complex inter-organizational projects (San Cristóbal et al., 2018). The described context has several characteristics that make it challenging: a multitude of multidisciplinary stakeholders and their coordination, the existence of various independent models, a multidomain environment, the existence of interdependencies, diverse objectives, and a dynamic environment. These are elaborated upon in the following sections.

Characteristic: A Multitude of Multidisciplinary Stakeholders and Their Coordination

Milch & Laumann (2016) refer to inter-organizational complexity as a complex sociotechnical system that involves multiple companies and work processes, requiring the collaboration of employees from different organizations and coordination across organizational boundaries. In this context, integration is as important as difficult because knowledge and information are dispersed across different departments and organizations. Moreover, the diverse engineering and non-engineering stakeholders need to collaborate for effective change integration.

In such cases, each project stakeholder has a different mental model of the project, assumptions about it, interpretations of realities, and expectations (Danilovic & Browning, 2007). Thus, such intensive interactions often cause conflicts due to differences in experience, knowledge, organizational or professional loyalties, understanding of the purpose and goals, and contradictory purposes and goals (Proehl, 1996).

This leads to subgoal 1: Facilitate multidisciplinary stakeholder/expert involvement.

Multidisciplinary or cross-functional teams should be advocated to address this by providing collective expertise, information, and resources for effective model-building and problem-solving (Madni, 2007). Hence, experts from different technical and non-technical backgrounds should be brought together (Sousa-Poza & Kovacic, 2008).

This can be addressed through integral expert-based sessions. In these sessions, experts from different backgrounds contribute their knowledge and expertise, exchange information, and discuss the impact of change (Smith and Hinchcliffe, 2003). Thus, tacit knowledge is deployed (Abbas et al., 2020), and multiple multidisciplinary stakeholder perspectives can be synchronized. These experts can include actors, parties, or organizations with whom an agreement is needed to implement the change. Ruitenburg (2017) showed that expert sessions deliver insights and aid in identifying opportunities that might otherwise be overlooked. Furthermore, expert sessions improve stakeholder engagement, close the gap between experts and professionals, and combine scattered resources. Instead of having a single person provide information for the SoI, a multidisciplinary group of experts is employed to define it collectively. This process enables a broad, holistic perspective and means that trust and acceptance are built during the process itself (Haanstra et al., 2021).

Thus, subgoal 1: Facilitate multidisciplinary stakeholder/expert involvement can be achieved through integral expert-based sessions.

Characteristic: The Existence of Various Independent Models

Stakeholders from various disciplines and organizations have their own insights, models, and approaches to describing and understanding an SoI (Haveman, 2014; Potts et al., 2021). Thus, diverse stakeholders cannot easily discuss an SoI when they do not share a common language (Haveman, 2014; Madni & Sievers, 2010). In these instances, a shared model can test and align participants’ mental models through discussion, leading to a joint understanding of the reality of projects (Danilovic & Browning, 2007).

During the last decade, systems engineering has improved significantly with the advent of model-based systems engineering and systems engineering markup language. These advances have enabled collaborative engineering teams to communicate using a common language and share information in digital models (Madni et al., 2014a). However, a disadvantage of these methods is that they are not suitable for addressing the needs of all stakeholders, especially non-engineering stakeholders (Madni et al., 2014b).

Rouse (2007) explained that large-scale complex systems require a broad perspective. Systems thinking is often described as balancing multiple perspectives to understand and guide problem resolution (Sauser & Boardman, 2015). Additionally, Browning (2002) suggested that systems thinking would enhance the management of SI by synchronizing multiple perspectives into an overview of the SoI. Systems thinking as a practice is intended to aid in creating an overarching perspective, understanding how independent elements come together into a unified SoI, understanding the environment in which it should perform, identifying the synergy of combined systems, and describing the SoI from all relative perspectives (Boardman & Sauser, 2008; Potts et al., 2021). Moreover, Browning (2002) mentioned that a classic approach to addressing and understanding complex reality is through modeling. A model is an abstract representation of reality built, analyzed, and manipulated to increase the understanding of reality (Browning, 2002). Browning (2009) states that tools are needed to provide improved visibility, appropriately simplify complexity, and highlight important areas. In the context of multidisciplinary teams, three main approaches have been identified (Haveman, 2014): (1) approaches based on functional modeling, describing the SoI at an abstract level, (2) visualization applications (Madni & Sievers, 2014), and (3) condense architectural information into a single and accessible overview. This is performed, for example, in the A3 architecture overview method (Borches Juzgado, 2010).

This leads to subgoal 2: Create a shared SoI among various stakeholders.

Moreover, attempting to provide all information about the SoI can cause information overload. This is often more detrimental than not providing information, owing to the false assumption that effective communication has occurred (Browning, 2009). As such, models that attempt to contain everything about a project are cumbersome to build, maintain, understand, and use (Little, 1970). It has been noted that managers prefer simple models, which they understand and trust, over more realistic ones (Little, 1970).

Because the details of a project develop throughout its duration, it is necessary to update the SoI at various stages of the process (Rail Safety and Standards Board, 2014). Thus, it should be flexible. In addition, the models must be based on the latest and most accurate input information if they provide helpful output (Danilovic & Browning, 2007). Therefore, high modeling flexibility is appropriate for realizing different railway projects in a changing environment. As such, it can be stated that any approach to creating a common understanding of the SoI should be compatible with existing SoIs, simplify complexity, be on the abstract level, be visual, combine multiple perspectives, and be condensed. This means that subgoal 2: Create a shared SoI among various stakeholders can be achieved through integral expert-based sessions, high-level insights, and high modeling flexibility.

Characteristic: A Multidomain Environment

Infrastructures are complex sociotechnical systems. They are highly interconnected networks of interacting social and technical components that cannot be addressed separately (Kroes et al., 2006). A change’s impacts can include processes, personnel, capacity, technical systems, and rules and regulations in sociotechnical systems such as railways. As such, it is important to understand the technical elements of a change and the organizational and operational procedures and human actions required.

This leads to subgoal 3: Take a broad perspective that should include more than the technical perspective.

Describing multiple domains provides a sound basis for structured analysis during the later stages of a project (Rail Safety and Standards Board, 2014). Bartolomei et al. (2012) formalized the identification and definition of domains common to all sociotechnical systems and projects as follows: (1) environmental domain, including exogenous components that affect or are affected by the sociotechnical systems, such as laws, policies, and regulations; (2) social domain, including individual stakeholders, teams, and organizations; (3) functional domain, including goals and purposes of the sociotechnical systems, as well as its functional architecture; (4) technical domain, physical, and nonhuman components of the system, including hardware, infrastructure, software, and information; (5) process domain, processes, subprocesses, and tasks performed within or by the system. Furthermore, Qureshi et al. (2007) explain that sociotechnical systems should be viewed as encompassing at minimum human, technical, and organizational domains, with intrinsic complexity arising from interactions and interdependencies between components. This is illustrated by Bartolomei et al. (2012) in Exhibit 2.

Exhibit 2. Distinct Domains and Their Interactions in Sociotechnical Systems (Bartolomei et al., 2012)

Thus, subgoal 3: Take a broad perspective that should include more than the technical perspective, can be achieved by considering sociotechnical viewpoints.

Characteristic: The Existence of Interdependencies

Davies and Mackenzie (2014) suggested that organizations cope with complexity by decomposing a project into different levels of SI with interfaces between individual subsystems. The definition of the parts may depend on technology or geography for example. In addition, the EM can impose interfaces to help manage the project. These include interdependencies, organizational or contractual interfaces, relationships, and shared or separate responsibilities.

However, in complex inter-organizational contexts, coordination among diverse groups becomes more challenging as their number and interdependencies increase. Regardless of how tasks are divided, linking various parts is always complicated, and as such, there is a constant need for their management to facilitate the exchange of information across these interfaces (Long & Spurlock, 2008; Stretton, 2016).

Identifying interdependencies indicates where a collaborative approach is needed from actors from either side of the interface. These can exist between departments in the same organization or across different organizations (Rail Safety and Standards Board, 2014; Stretton, 2016), and knowing where relevant interdependencies exist (Dasher, 2003) reveals which topics require knowledge and information to be shared across organizations.

This leads to subgoal 4: Facilitate the structured exchange of inter-organizational knowledge and information.

The interdependencies can be classified as pooled (each part provides a discrete contribution to the project, irrespective of other parts), sequential (one organization’s output becomes an input for another part), or reciprocal (outputs of each unit become inputs for others and vice versa) (Geyer & Davies, 2000). Additionally, a dependency structure matrix (DSM), which displays the relationships between the components of a system in a compact, visual, and analytically advantageous format, can provide an excellent approach for mapping interfaces in SoIs (Browning, 2001). However, in complex inter-organizational contexts where different perspectives make up the SoI, additional preliminary steps are required first to identify the SoI.

Furthermore, in an inter-organizational project, distinct parties perform multiple tasks to achieve the main objective. However, not all tasks must be performed simultaneously. A method for distinguishing these is the circle technique, which focuses on progress-oriented instead of problem-oriented work (Visser, 2013). This method consists of three steps: (1) writing down everything that has already been achieved and operates well in the inner circle; (2) noting what still needs to be achieved in the outer circle; and finally, (3) discussing which of the items in the outer circle need to be worked on to be moved to the inner circle (Visser, 2013). This technique mainly revolves around inventorying, brainstorming, and the creative process behind these to gather as much useful information as possible.

Thus, subgoal 4: Facilitate the structured exchange of inter-organizational knowledge and information can be achieved by creating insight into inter-organizational interdependencies/interfaces and creating insight into task sequencing.

Characteristic: Diverse Objectives

In temporary organizational systems such as projects, independent and interdependent entities cooperate for a limited period to achieve specific objectives (Pezzillo et al., 2012). The objective is typically a short statement outlining the purpose and function of the change and whether it is technical, operational, or organizational (Rail Safety and Standards Board, 2014). Depending on the type of change, it is helpful to explain the reason for the change, for example, whether it is for an improvement in capacity, safety, or reduction in cost.

According to Foster-Fishman et al. (2001), reducing organizational barriers and creating mutual goals and objectives is important to enhance inter-organizational collaboration and facilitate the exchange of information and resources across organizational boundaries. In such cases, all project participants must be clear about their goals and objectives, including owners, managers, contractors, and consultants (San Cristóbal et al., 2018). Thus, the objective provides a useful context and clarifies the main reasons for the change and the requirements of a successful assessment. Furthermore, discussing the objective aids in jointly producing a mutually valued outcome.

Thus, subgoal 4: Facilitate the structured exchange of inter-organizational knowledge and information, can be achieved by determining a clear objective.

Characteristic: A Dynamic Environment

As previously stated, the current society is characterized by rapid changes and advancements in technology. Moreover, system changes occur at various time scales (Sterman, 2001). This means that while jointly creating insight into high-level inter-organizational SoIs to support EMs, additional information might become available, be subject to changes, or even the intended change itself might be susceptible to adjustments (Keating et al., 2008). Establishing processes to maintain stability while dynamically responding to uncertain and changing conditions is one of the most challenging aspects of SI (Davies & Mackenzie, 2014).

A more flexible and incremental approach is required to solve this problem.

This leads to subgoal 5: Cope with significant modeling changes, which can be achieved by providing a high degree of modeling flexibility.

While it can be concluded that the current literature focuses on the individual characteristics describing the present context and related solution directions, no pragmatic solution is provided that synthesizes the aforementioned subgoals and design principles into a single method or tool to support EMs with the project scope definition. Therefore, this study aims to support the management of inter-organizational change integration in the sociotechnical railway system by designing an expert-based analysis that aids in jointly creating insight into inter-organizational SoIs, aiding the scope definition. This is done by (1) facilitating multidisciplinary stakeholder/expert involvement, (2) creating a shared high-level SoI among various stakeholders, (3) taking a broad perspective including more than merely the technical perspective, (4) facilitating the structured exchange of inter-organizational knowledge and information, and (5) coping with significant modeling changes.

Methodology

This study utilizes the design science research methodology (DSRM), which aims to design an artifact to address an unsolved and important problem (Hevner & Chatterjee, 2010). The DSRM involves a rigorous process for designing artifacts to solve observed problems, make research contributions, evaluate designs, and communicate results to appropriate audience members (Peffers et al., 2007). It constitutes a systematic but flexible methodology that aims to improve practices through iterative analysis, design, development, and implementation based on collaboration among researchers and practitioners in real-world settings, leading to context-sensitive design principles and theories (Wang & Hannafin, 2005). Peffers et al. (2007) designed a commonly accepted framework for carrying out the DSRM, as displayed in Exhibit 3. As this research resulted from observing the problem, this problem-centered approach is particularly appropriate.

Exhibit 3. DSRM Approach Adapted from Peffers et al. (2007)

Furthermore, this study used data triangulation by employing multiple methods to collect data on the same phenomenon of interest (Campbell et al., 2020). This type of triangulation, frequently used in qualitative studies, includes interviews, observations of expert sessions within the industry, and documents on existing models and modeling approaches in the industry, which aim at achieving nearly identical objectives but on a small scale and in a decentralized manner.

Design Objective and Principles

The main objective of this study (II in Exhibit 3) was to design an expert-based analysis that aids EMs in jointly creating SoIs. This design is broken down into multiple subgoals, which have been addressed using design principles: (1) provides high-level insights, (2) includes sociotechnical viewpoints, (3) utilizes integral expert-based sessions, (4) provides insight into inter-organizational interdependencies, (5) has a high degree of modeling flexibility, (6) focuses on a clear objective, and (7) provides insights into task sequencing. Moreover, the developed analysis should provide the ‘how’ and give insight into the process which results in the SoI.

The combination of the subgoals and principles is summarized in Exhibit 4. Moreover, emergent design principles arise from the iterative nature of the design cycle.

Exhibit 4. Design Subgoals and Related Design Principles

The combination of the design principles mentioned above and their iterations led to the design and development (III in Exhibit 3) of the MOSAIC analysis.

Design of MOSAIC

The developed MOSAIC analysis consisted of four steps, visually represented in Exhibit 5.

Exhibit 5. Proposed MOSAIC Steps

This is described in the following paragraphs.

Step 1: Goals & Scope Definition of the Change

The first step of MOSAIC is to develop an understanding of the main goal and preliminary assumptions regarding the change, the type of change, and an initial overview of the stakeholders impacted by it. This determines the initial scope of the change project, and which multidisciplinary experts from distinct organizations are required to be involved in MOSAIC Step 2. Additionally, the starting assumptions are important: they provide a record of circumstances for which the analysis is valid. If these assumptions change later, the MOSAIC steps should be reviewed and revisited, if necessary.

In Step 1, the following guiding questions from Bartolomei et al. (2012) and Maier and Rechtin (2000) can be useful: (1) Who affects/is affected by the change (2) Who decides on the project (3) Who carries out work in the project? (4) Who benefits from it? (5) Who provides what? (6) Who loses?

Step 2: Impact Identification by Means of Integral Expert Session

To meet the objective established in Step 1, an understanding of how the change affects the previously determined stakeholders needs to be reached. Therefore, this step strives to determine what impacts the change will have. These can include impacts on various sociotechnical domains: social (individual stakeholders, team, organizations), functional (functions), process (tasks, processes), technical (technical components, subsystems), and environmental domains (laws, policies, regulation).

These impacts can be identified by bringing together selected experts in an expert session, during which collective expertise and information are employed and synchronized, and diverse tasks required to be conducted by distinct entities to integrate the change effectively are determined. Exhibit 5 depicts these using rectangles, representing the sticky notes used in the expert sessions. Thus, this second step leads to an initial, synchronized, high-level overview of the SoI.

During this session, the following questions can be raised: (1) What is the effect of the change on the aforementioned domains? (2) What adjustments are required for change to be properly integrated? (3) Are these adjustments a part of the project? (4) Which tasks must be conducted by distinct departments and stakeholders to achieve the identified objective?

Step 3: Impact Analysis

After the information collection steps, the initial SoI can be analyzed to highlight important features. The information presented on sticky notes, such as grouping related tasks and impacts, should be checked for possible duplication and clustering. Furthermore, it is important to distinguish which individual, department, and organization is responsible for each task for effective follow-up.

Depending on the entity responsible, this can be done by recoloring the sticky notes. In addition, there is a difference in the sequence of the identified tasks. Although some tasks have already been completed, some tasks remain to be conducted to achieve the main objective. This differentiation can be made by moving existing sticky notes to the inner or outer circle, as shown in Exhibit 5. Furthermore, this information can be reorganized into distinct domains. Because preliminary research showed that the technical, process/organizational, and compliance/regulatory domains are the most significant in the railway context, the SoI is divided into quadrants reflecting this. This demonstrates which sociotechnical domains the current focus is on, whether there is a tendency toward a specific domain, or whether more attention should be paid to other domains.

Therefore, Step 3 leads to a more detailed, synchronized, high-level overview of the SoI, providing insights into the responsible stakeholders per identified task (colors), task sequencing (inner and outer circles), and any tendencies toward different sociotechnical domains (quadrants), highlighting directions for further analysis.

Step 4: Identification of Inter-Organizational Interdependencies

Because integration mostly fails at interfaces, it is essential to identify the interfaces between the elements mentioned above and manage them accordingly. Especially in inter-organizational contexts, it could be the case that some responsibilities are shared between organizations and, as such, require close collaboration on that interface for effective integration of the change. These interdependencies can be inter- and intra-domain as well as inter- or intra-organizational.

Relevant questions to ask during this step include (Browning, 2002): (1) What inputs are needed to do the work? (2) What is the source or supplier of each input or task? (3) What is the destination or customer for each output? (4) Which tasks depend on inputs from other departments/organizations?

Step 4 leads to an overview of interfaces in the SoI. These can include (1) tasks that are interdependent (i.e., sequential or reciprocal) or (2) tasks in which responsibilities are shared between stakeholders (i.e., there is an interface between responsible parties, and collaboration is necessary).

After carrying out steps 1–4, the SoI is defined.

Demonstration of MOSAIC: A Case Study on a Multipurpose Train Towing Vehicle

MOSAIC was applied to a case study in the Dutch railway sector to demonstrate its practical applicability. This concerned a project on a multipurpose train towing vehicle. The purpose of the vehicle’s deployment by the Incident Response Department of the infrastructure managing organization is to tow trains belonging to the main railway operator when stranded on the high-frequency railway network. This towing process should be conducted as quickly as possible in situations where a train is stranded owing to problems with the traction systems or in case of problems with the overhead contact line. Furthermore, this vehicle is powerful enough to pull or push a full passenger train and is unique in its ability to operate on both roads and rail, allowing easier access to stranded trains. Because of its inter-organizational nature, this project included multidisciplinary representatives from the infrastructure managing organization and the main railway operating organization.

Based on structured interviews to explore challenges in this case study, conducted with all project team members, it became apparent that a standardized process for integrating this type of joint initiative in existing railway operations has not yet existed. The newly introduced vehicle affected the operations of the main railway operator and the infrastructure manager and their responsibilities in towing stranded trains. Mutual responsibilities and ownership needed to be well established during this phase and after commissioning. The problem investigation indicated that the multipurpose train towing vehicle was (1) a large project with a broad scope involving different departments and organizations, which made its scope unclear. Moreover, because of its large scope, (2) numerous multidisciplinary stakeholders with different perspectives and a lack of shared understanding were involved. Additionally, (3) predicting the impact of this change in the inter-organizational context was challenging.

The following subsections demonstrate the step-by-step application of MOSAIC to the case study, which is briefly summarized in Exhibit 6.

Exhibit 6. Demonstration of MOSAIC Steps Applied to the Project Concerning the Implementation of a Multipurpose Train Towing Vehicle

Step 1: Defining the Goal and Scope of the Project Concerning the Implementation of a Multipurpose Train Towing Vehicle

The type of change, goal and scope, and starting assumptions of the case study were determined ahead of the expert session, together with project leaders from the infrastructure managing organization and the main railway operating organization. The knowledge of their own organizations aided them in understanding which departments and processes would be affected by the project. This highlights the scope and indicates that collaboration is necessary to clarify the scope in the inter-organizational context. In this regard, Project Member A from the Safety Department mentioned:

In case of an internal project, it is easier to get the right parties together. It is important to look across departments and have a project interest, and not only a departmental interest.

The inter-organizational scope aided in identifying the most important experts required to be involved in Step 2. Because the project was quite large, the emphasis of this case study was ultimately on a subgoal of the project: to have a complete file finalized for the independent assessment body containing demonstrable evidence concerning safety requirements, as shown at the center of Exhibit 6.

Step 2: Impact Identification by Integral Expert Session of the Project Concerning the Implementation of a Multipurpose Train Towing Vehicle

The experts identified in Step 1 were requested to join an integral expert session. These included inter-organizational representatives from the service and operations department, train testing center, incident management, legal counselors, project leaders, implementation managers, and safety experts, as depicted in Exhibit 6.

Due to the Covid-19 pandemic, the integral expert session could not be held in person and was required to take place online. Because of this, several online collaborative tools were tested to maintain optimal effectiveness and efficiency. The visual collaborative software MIRO was chosen to carry out the session because of its ability to support the collaboration of multiple users simultaneously, allowing the step-by-step building of the design, and providing opportunities for easy adjustments and analysis.

To reach the goal set in Step 1, the impacts of the change needed to be clear, and tasks required to be conducted by the stakeholders involved in the expert session were identified. Examples include access agreements in their proper order, arranging vehicle maintenance, certifications, and safety systems to be up to date. A project member working in service and operations stated the following:

He must pick up on signals for train drivers in this session and represent the interests of the drivers: determine what training is needed, whether safety agreements are covered, and arrange capacity for drivers for testing.

During impact identification, participants could use sticky notes to write down aspects of which they had knowledge and place them on the digital board. This resulted in an overview of the various impacts of the changes and tasks that needed to be arranged to integrate the change effectively. The facilitator then addressed the sticky notes to encourage open discussion within the group. This discussion aided in clarifying various tasks and provided insightful discussions within the group. An example of such a discussion dealt with the inter-organizational cooperation between the organizations during vehicle operation and shared responsibilities in the case of accidents.

Step 3: Impact Analysis of the Project Concerning the Implementation of a Multipurpose Train Towing Vehicle

After information collection, the SoI was checked for duplicates and clusters, such as tasks related to risk management. At this stage, participants could recolor the digital sticky notes produced in Step 2, depending on who was responsible for the task listed on them. Project member X, who focused on legal aspects, mentioned the following:

The different roles and responsibilities must be clear when the multipurpose towing vehicle is operational.

In addition to using different colors, the participants were required to tag each sticky note with a name. The layout distinguishing task sequencing (inner and outer circles) and sociotechnical domains (quadrants) were already prepared in MIRO to facilitate brainstorming; hence, sticky notes could be easily repositioned accordingly. This step revealed a strong tendency toward regulations/compliance and less toward the technical domain of the project. Project member Y from the Safety Department stated the following:

The project includes a lot of inter-organizational regulations, because of that, there is a lot more to be considered.

Step 4: Identification of Inter-Organizational Interdependencies of the Project Concerning the Implementation of a Multipurpose Train Towing Vehicle

To identify interdependencies, the facilitator opted to use multiple questions regarding tasks represented by sticky notes: Which tasks depend on others? What is required by other stakeholders to carry out these tasks? Are these inputs for other tasks? Do these factors lead to additional impacts (not previously identified)?

“There are quite a few challenges in this project, the train drivers and vehicle operator have to work together, which is very different compared to other projects. How are the responsibilities shared when a stranded train is towed by the vehicle? We should manage these dependencies correctly.” – Safety expert Y.

Because there was already substantial discussion in steps 2 and 3, interdependencies between different tasks had already been referred to, albeit indirectly. However, these were not explicitly stated in the SoI or assumptions.

“We should ensure that we know these interdependencies at the front end, so we can get our own processes up and running properly in time.” - Project member Z.

This demonstrates that each MOSAIC step leads to intermediate results. Moreover, further details of the SoI are obtained during the process. A representative focusing on project implementation stated the following:

In the beginning, it did not seem that complicated, but during the project it turned out that there were many more aspects that had to be considered.

Evaluation & Discussion

The next step in the DSRM process (V) presented in Exhibit 3 concerns design evaluation. This involves comparing the objectives of the proposed solution with the observed results from the use of the artifact in the case study. The MOSAIC was evaluated using information from preparatory meetings, observations during the case study, and evaluations after the expert session.

This section will evaluate each MOSAIC step in detail and describe to what extent the design principles worked to achieve the different subgoals.

Evaluation of MOSAIC Step 1: Goal & Scope Definition of the Change

Step 1 describes the main objective of the change and type of change. By mentioning this, the main reason for the change and the requirements for a successful assessment were made explicit. Moreover, the objective provided a useful context and helped determine which multidisciplinary stakeholders/experts (subgoal 1) were required to be included from the affected processes, departments, and stakeholders at a high level for each specific objective. To do this, project leaders relied on their knowledge of the organization and departments and their understanding of whether a change would affect different organizational units. A project member working in service and operations stated:

The key to reaching the objective is collaboration and connecting the right people.

Furthermore, the iterations in the case study application show that MOSAIC analysis is highly flexible in its application and can easily be adjusted to changing goals and, as such, determining which experts are required to be involved. Project member B:

One must very clearly define the scope within one’s own work, and a clear overall objective helps with this.

Multiple participants stated that it would be easy to include this approach in different projects due to its flexibility, and it would certainly help to carry out this session as early as possible in projects. Moreover, this indicates that the MOSAIC analysis aids significantly in clarifying the initial scope of the project by determining which departments, processes, and organizations are impacted by this change. This analysis can also respond to changes over time, which means that MOSAIC can cope with significant changes (subgoal 5).

Evaluation of MOSAIC Step 2: Impact Identification by Means of an Integral Expert Session

Step 2 determines the impact of the change by using an integral expert-based session. This session helped involve the identified multidisciplinary stakeholders and experts (subgoal 1). Project member B stated:

The people involved have different backgrounds, e.g., more ‘outside people’ who are more hands-on versus more ‘inside people’ who know less about the actual calamities and work more from a procedural perspective. The collaboration between them is essential.

The impacts included high-level impacts on sociotechnical domains. Moreover, because visual aids support these insights; subgoal 2: Create a shared SoI among various stakeholders was met. By asking guiding questions related to sociotechnical viewpoints, subgoal 3: Take a broad perspective that should include more than the technical perspective was also met. This combination led to an initial, synchronized, high-level overview of the SoI, which facilitated the structured exchange of inter-organizational knowledge and information (subgoal 4). Project member Y from the Safety Department stated the following:

Conducting such an exercise shows how complex the project is, and that every expert has their own understanding of impacts related to their field of work.

Moreover, according to several participants:

Doing such an exercise in a group provides more consensus.

In this step, it was essential to give participants time to reflect and write down the aspects they considered relevant. Additionally, the role of the facilitator is important; while it did help that the facilitator present had content-related knowledge of the project in this case, it would have been beneficial to have a process facilitator present who focused only on the MOSAIC process and analysis without engaging in content-related discussions.

Evaluation of MOSAIC Step 3: Impact Analysis

Although stakeholders from various disciplines each had their own models and approaches for describing and understanding an SoI, the integral expert session facilitated the involvement of multidisciplinary stakeholders/experts (subgoal 1) to a great extent. By focusing on high-level insights and multiple generalizable cross-sections such as sociotechnical perspectives (showing a tendency toward specific domains) and insights into task sequencing (to achieve the objective), the attendees could easily share knowledge from their own experiences, ensuring that the analysis aids in creating a shared SoI among various stakeholders (subgoal 2) and facilitates the structured exchange of inter-organizational knowledge and information (subgoal 4). One interviewee stated:

You always have an idea of the impacts of a change by and large, but not exactly what that would mean. Only after the impact analysis and discussion does it become clearer.

The evaluation revealed that MOSAIC created an understanding within the group that different domains play a role in reaching the defined objective. Despite this, the reorganization of sticky notes in one of the sociotechnical quadrants did not proceed smoothly. This suggests that, although acknowledging different sociotechnical perspectives is important, this way of thinking does not come naturally to the participants. Furthermore, the evaluations showed some inconsistencies regarding the tagging of sticky notes. It was unclear whether the tag referred to the individual responsible for the task mentioned on the sticky note or whether the tag referred to the participant who added the sticky note. Project member B mentioned:

Sometimes we are not explicit enough about the responsibilities, which results in a lack of ownership.

Overall, participants mentioned that MOSAIC helps to identify concrete tasks and identify the responsible stakeholders required to take actions. In addition, this process and the resulting overview create insight into conducted tasks and what the focus should be on next. Moreover, it helps prioritize important aspects and improve relationships among project members.

Evaluation of MOSAIC Step 4: Identification of Inter-Organizational Interdependencies

Close collaboration across interfaces is required to integrate changes effectively in an inter-organizational context. Evaluation with participants indicates that the insight into interdependencies to a large extent, aids in clarifying (1) how different tasks fit into the project as a whole, (2) how different tasks are linked to each other, and (3) how a change in adjacent tasks could impact dependent tasks, and as such, should be planned accordingly. This shows that making these insights explicit facilitates the structured exchange of inter-organizational knowledge and information (subgoal 4) on specific tasks/impacts. One attendee mentioned:

Applying regulations between 2 companies takes a lot of time. This shows that we need to start earlier with such exercises and have the right people join at the right times.

In some instances, the guiding questions asked by the facilitator led to impacts that had not been considered previously. However, because there were inconsistencies with the tagging of the sticky notes in steps 2 and 3, it was not explicitly stated who and which organization was responsible for which task. This made it more difficult to identify tasks in which responsibilities were shared.

An additional session is necessary for future applications to identify the interdependencies. Moreover, because a significant number of tasks with possible interdependencies were identified, a simple, compact, and visual representation is preferable for further analysis. Further research on this is required, which should include paying attention to the differences in pooled, sequential, and reciprocal interdependencies (Davies & Mackenzie, 2014), as well as how interdependent tasks should be followed up on.

During steps 2, 3, and 4, a significant number of assumptions and clarifications were discussed throughout the expert session; making these explicit and including them in the SoI could improve the follow-up in later sessions. In addition, the SoI which resulted from the application of MOSAIC is based on information contemporary to the execution of the different steps. This means that for the SoI to remain relevant and usable, it must be updated continuously based on the latest information available. Project member Q from the safety department stated:

If you want to plan everything in advance and not be flexible to changes, it is not going to work.

All interviewees mentioned that the resulting SoI and a structured analysis to get to such an integral, mutually agreed upon are important to facilitate the structured exchange of inter-organizational knowledge and information. According to project member A:

This method could be introduced in all other projects that want to move forward. If you do this well with the right delegation of people involved in the project or those who are going to be affected by it, there are always insights and/or confirmations in the vein of ‘we are on the right track,’ or ‘there is still work to do here.’ It gives direction and ensures connections regarding content and relationships.

Implications for Engineering Managers

EMs are frequently required to oversee the integration of changes in complex sociotechnical systems. Similar to the case study in this study, EMs in this context outside of the railway sector are often faced with large projects with a broad, ambiguous scope involving different departments and organizations. As a result, numerous multidisciplinary stakeholders with different perspectives and a lack of shared understanding are involved. Additionally, predicting the impacts of changes in an inter-organizational context can be challenging for EMs. Fortunately, the MOSAIC can provide useful guidance. The application and evaluation of MOSAIC in a practical case study have multiple implications for EMs:

1. By bringing together experts from different backgrounds and different organizations in the expert session with a clear, explicit goal and supported by visual aids, the impacts of the change on their respective disciplines and domains can be mapped and discussed. This ensures that diverse insights and domain-specific SoIs can be combined into an integral overview, facilitating multidisciplinary group communication and sharing of information across organizational boundaries.

2. Additionally, confusion between different practitioners working on the same change is reduced by not basing an SoI on assumptions related to the change, but building it on explicit information. Instead of having a single person providing information for the SoI, a multidisciplinary and inter-organizational group of experts can be employed to define the SoI collectively. Thus, conducting MOSAIC in a group setting helps create a consensus on the SoI.

3. By including an appropriate delegation of experts in the process, both from the project and those affected by it, MOSAIC provides a broad range of insights for EMs. These include the tendency toward specific domains and insights into task sequencing, which are relevant for EMs in steering and controlling the project.

4. Based on the discussions observed during the case study, the MOSAIC analysis, inherent discussions, and resulting SoI seem to have stimulated the working relationship among project members by providing opportunities to understand each other’s perspectives and expertise while still systematically synchronizing these in a high-level, visual, and condensed manner.

5. Carrying out MOSAIC can aid EMs in front-end planning, which is critical for uncovering project unknowns by focusing on explicit objectives, developing an adequate scope definition, and recognizing key stakeholders.

6. The MOSAIC results can be used to identify concrete tasks, and responsible stakeholders required to take action for the progress of the project. The case study shows that interdependencies between tasks are implicitly known to individual experts; however, making these explicit and placing appropriate levels of focus on these also clarifies topics/impacts requiring more detailed coordination between multiple stakeholders to support the EM with the integration of different parts.

7. MOSAIC can be easily generalized and applied by EMs to other projects within the railway industry and beyond. These projects (a) involve multidisciplinary experts; (b) concern a variety of (non)engineering stakeholders who are required to collaborate for effective change integration; (c) require information that is dispersed over (multiple) departments and organizations; and (d) concern a sociotechnical system change, where multiple domains play a role. While MOSAIC focuses mainly on the technical, process, and regulatory domains which are deemed important in the railway industry, it is possible that other domains play a significant role in other industries. In this sense, the MOSAIC is flexible and can be adjusted accordingly.

Conclusion

Sociotechnical systems are becoming more complex and inter-organizational, causing the number of inter-organizational components and interdependencies to increase drastically. As such, a change in one part may affect other system parts, impacting multiple domains across different organizations. This broadens the SoI and makes it more difficult to describe, understand, and manage EMs. This study designed the MOSAIC analysis to support EMs in managing inter-organizational change integration in a sociotechnical railway system to aid in achieving performance upgrades resulting from change integration.

This paper shows that MOSAIC contributes to this by (1) focusing on a clear objective and incorporating relevant stakeholders; (2) providing a structured process and analysis to determine the SoI; (3) synchronizing the perspectives of inter-organizational and multidisciplinary experts through an integral expert session; (3) enabling an open-yet focused discussion, supported by guiding questions and visual aids; (4) identifying diverse impacts of the change across sociotechnical domains, and tasks required to be carried out by diverse stakeholders to achieve the determined objective; (5) highlighting ownership and responsibilities; and (6) identifying interdependencies across which collaboration is essential. Furthermore, the evaluation shows that the designed analysis is pragmatic and flexible in dealing with the changes. Owing to the high-level insights it provides, MOSAIC is adaptable for use in other engineering management contexts. Moreover, because MOSAIC is an analysis in which experts exchange information, interact with, and discuss the impacts of the change based on their expertise and experience with similar changes, the analysis is less appropriate for changes related to radical innovations with which experts have no experience.

The world is becoming increasingly connected, inevitably causing challenges that are too difficult to solve by applying methods from a single discipline. This is reflected in the growing trend toward multidisciplinary collaboration. By applying the MOSAIC analysis, collective expertise and information are synchronized in high-level SoIs, highlighting what the change entails, its scope, inherent sociotechnical domains, and inter-organizational responsibilities, all of which require attention to reach the foreseen improvements. Moreover, by establishing a stepwise SoI in a group setting, mutual understanding and acceptance are built during the process, facilitating the exchange of information and resources across organizational boundaries, thus enhancing collaboration and supporting EMs in managing inter-organizational change integration.

Acknowledgments

This study was co-financed by the research and innovation contribution (PPP) of the Dutch Ministry of Economic Affairs and Climate. The authors acknowledge the support of NS and ProRail in making this study possible through the framework of the Systems Integration for Railway Advancement (SIRA) project. Furthermore, the authors acknowledge peer review feedback, which has improved this paper.

Exhibit 1 is reprinted from Research Policy, Vol. 31, Frank W. Geels, Technological transitions as evolutionary reconfiguration processes: a multi-level perspective and a case-study, 8-9, Copyright (2022), with permission from Elsevier.

Exhibit 2 is reprinted from Systems Engineering, Vol. 15, Donna H. Rhodes, Richard de Neufville, Daniel E. Hastings, et al., Engineering Systems Multiple‐Domain Matrix: An organizing framework for modeling large-scale complex systems, Page 79, Copyright (2011), with permission from John Wiley and Sons.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The work was supported by the Research and Innovation contribution (PPP) of the Dutch Ministry of Economic Affairs and Climate.

Notes on contributors

Merishna Ramtahalsing

Merishna Ramtahalsing is a Ph.D. candidate in the chair of Asset Management and Maintenance Engineering within the Department of Design, Production, and Management at the University of Twente. Her doctoral research focuses on the integration of systemic changes in complex, inter-organizational, and socio-technical contexts like railways, to improve system performances. She holds a master’s degree in Mechanical Engineering from the University of Twente and focuses on maintenance engineering and system integration.

Willem Haanstra

Willem Haanstra is an assistant professor in the Maintenance Engineering chair at the Department of Design, Production, and Management at the University of Twente. His research focuses on strategic Asset Life Cycle Management, mainly from systems and life cycle perspectives.

Jan Braaksma

Jan Braaksma is an associate professor in the chair of Asset Management and Maintenance Engineering at the University of Twente. Jan’s research focuses on Asset Management and Maintenance Engineering with special attention for sustainable Asset Management, Asset Life Cycle Planning, systems integration, and Design for Maintenance.

Mohammad Rajabalinejad

Mohammad Rajabalinejad is an Assistant Professor at the University of Twente. He is interested in safety as a designer, systems engineer, teacher, citizen, human, and as a father. He has practiced different disciplines including Civil, Mechanical, and Industrial Engineering. He is leading several projects including SIRA (system integration for railway advancement). He serves the scientific community as associate editor. He has organized several special sessions and chaired conferences. He is a member of the ISO committee for “Safety of Machinery”.

Leo van Dongen

Leo van Dongen has worked for the Netherlands Railways for 35 years. He retired as Chief Technology Officer, responsible for the asset management of the rolling stock fleet, workshops, and maintenance equipment, in August 2019. Since 2010 he is a professor in Maintenance Engineering at the Faculty of Engineering Technology, in the Department of Design, Production, and Management of the University of Twente. He is a mechanical engineer and completed his doctoral research at the Eindhoven University of Technology on the energy efficiency of drive trains for electric vehicles.