The Design of Citizen-Centric Green IS in Sustainable Smart Districts

The switch to a perspective with a greater focus on the citizen brings us from SSCs to the concept of SSDs, a planning approach where the citizen-centric service perspective is particularly well aligned with the core idea of sustainability (Ahvenniemi et al. 2017; Graf-Drasch et al. 2022). As defined by Keller et al. (2019, p. 1404), an SSD is “a district performing in a forward-looking way in economy, people, governance, mobility, environment, energy, and living, built on a sophisticated, smart ICT infrastructure that ensures benefits for every stakeholder, in particular a high quality of life for every citizen.” In line with this definition, we see SSDs as integral, normative, and visionary subsystems of SSCs that bridge the gap between two sets of goals, the digital and the sustainable ultimately leading to a better quality of life (Martin et al. 2019).

In this study, we place a special focus on residential SSDs to look at how they perform to the mutual benefit of citizens and other city stakeholders, such as investors and service providers (Bisello 2020). Many SSDs already implement sustainability-related projects, ranging from the creation of new mobility projects or the sharing of concepts to the smart use of renewable energy sources and collaboration platforms that allow citizens to connect with one another more efficiently (Cappellaro et al. 2020; Hamari et al. 2016; Hosseini et al. 2018). It is important to note, however, that these projects rely on adequate technology infrastructure (Keller et al. 2019), such as mobile apps that make it possible not only to connect ever more citizens digitally but to provide and process a growing range of services (Anttiroiko et al. 2014; Staletić et al. 2020). Other examples include smart meters and sensors that facilitate the meticulous tracking of energy consumption in individual households, while advanced algorithms or artificial intelligence can provide greater transparency of energy supply and demand (Bonenberger et al. 2021; Mazur et al. 2019). Yet, even if a district makes full use of such technology, it is insufficient to make life more sustainable. Indeed, there is a clear need for approaches that ensure meaningful integration of the various characteristics and requirements of technologies, services, and human beings.

A subfield of ISs research, Green ISs can be defined as “information systems to transform organizations and society into more sustainable entities” (Seidel et al. 2017, p. 41). In this context, ‘sustainable’ denotes conserving, deploying, and reusing resources responsibly (Malhotra et al. 2013; Seidel et al. 2017). Green ISs are understood to have far greater potential to achieve this as they leverage advanced technology to provide services to individuals, groups, or organizations while keeping the focus firmly on sustainability-related goals (Watson et al. 2008; Schoormann and Kutzner 2020). Initially examined in 2007, Green IS research largely remains devoted to environmental sustainability, which is also strongly related to energy informatics (vom Brocke et al. 2013; Wang et al. 2015; Watson et al. 2010). These days, however, Green ISs are also appreciated for their ability to automate, inform, and transform processes in all areas of sustainability, be it by estimating energy consumption or providing guidance for behavioral change in line with the social perspective of an SSC (Brauer et al. 2015). For Green ISs to support sustainable actions within an SSD, it has to be planned, designed, implemented, and managed with in-depth understanding (Watson et al. 2008; Graf-Drasch et al. 2022; Melville 2010), and this understanding extends to its core constructs, as posited by Alter (2013). A summary of these constructs is presented in Fig. 1.

It is worth pointing out that the workings of Green ISs are significantly affected by three contextual factors: environment, infrastructure, and strategy (Alter 2013). As for the internal dynamics of a Green IS, the interplay of services & processes and technologies & information among the district’s stakeholders – especially its citizens – has to be brought into close alignment to create benefits for all stakeholders (Bisello and Vettorato 2018; Corbett and Mellouli 2017). When this is achieved, the Green IS can support an SSD in reaching its sustainability-related goals, in the process of which it facilitates notable improvements to the citizens’ quality of life (Bisello and Vettorato 2018; Gimpel et al. 2021).

Designing such a Green IS has proven to be a constructive research process, one by which to find an effective way of achieving the goals of SSDs (Hevner and Park 2004). Unlike in the corporate context, users who engage with IS in the urban context do not tend to rely on the one provided by a central authority as they go about their daily lives. Suppose, then, that the goal of a more sustainable lifestyle is to be reached by a critical mass of citizens within an SSD. In that case, it is essential to let citizens and stakeholders in the respective district participate in this research process. Indeed, the best results are achieved if they also participate in the co-creation of the Green IS (Golla et al. 2020). Dalén and Krämer (2017), for example, state that the attitudes and norms of individuals could be targeted with user-centered Green ISs. So far, however, many Green ISs have been pushed on citizens from the top-down, the main focus being firmly placed on introducing the technology itself without considering the needs or priorities of citizens (Trencher 2019). One unfortunate risk associated with this introduction strategy is that it can diminish citizen support, render the service provision within an SSD inadequate, or cause malfunctions, all of which can ultimately reduce the number of interactions and touchpoints (Graf-Drasch et al. 2022).

In Table 3, we present our seven final DPs for Green ISs in SSDs, all of which were systematically derived from the literature and refined over the course of 15 interviews (see Evaluation Episode 2). The DPs are formulated in accordance with the conceptual scheme for DP, as proposed by Gregor et al. (2020). This scheme structures each DP in terms of title, aim, context, mechanism, and rationale. Specifically, the aim is what we want to achieve with the respective DP. The context clarifies the stage of the software cycle at which a given DP receives the greatest attention (Alter 2013). The mechanism denotes how a DP wants to achieve a given aim. The rationale answers the question as to why we should pay attention to a particular DP and which area of the current literature is most pertinent to it. To ensure better understanding and practical application, we detailed at least five exemplary actions concerning each DP (see Appendix A). These exemplary actions indicate which practical measures ought to be implemented for each DP. In the following, we provide a detailed description of the structural interdependencies between the various DPs. We also discuss the application of each at its particular stage of the implementation process. While the general focus of these DPs is on the non-technological aspects of SSDs, DPs 1 to 4 are more qualitative in their specific focus, DPs 5 to 7 more technological. As we identify the relationships between the individual levels of these DPs and the literature that influenced their respective development, we expand on the content of Table 3 and deal with the most pertinent issues noted by other academic researchers.

DP 1 – Involvement of Citizens: The first DP is an integral aspect of a bottom-up approach to the human-centered design of Green ISs in SSDs (van der Bijl-Brouwer 2017). Involving citizens at the stages of initiation, development, and implementation of a Green IS brings its design into far closer alignment with the real needs of its users and the objectives of their districts (DP 2) (Kumar et al. 2020). Such active involvement can be beneficial at different stages of the SSD design. At the initiation stage, for instance, citizens can be involved to ensure that they identify their real needs before the focus of the development and implementation has a chance to shift away from the true purpose of the design. Once the active operation is underway, this involvement of citizens remains an important quality control measure as the feedback loop can keep the focus on their real needs and any necessary adjustments can be made accordingly (DP 3). Subsequently, say at the implementation stage, a designer can involve special target groups to account for the needs of various demographics and redirect efforts to the proper channels to ensure that the Green IS will integrate diverse perspectives (Renyi et al. 2019). It is worth noting, however, that Green IS designers might struggle with certain challenges arising from this involvement process. In most cases, designers must work with large groups of citizens, not all of whom will automatically understand the design process or appreciate how it may benefit them. This makes it all the more important to involve them as early as possible, allowing them to identify their concerns in light of which a healthy environment can be designed and redesigned accordingly.

DP 2 – Realization of District Objectives: Modern urbanization has led to a complex fragmentation of cities into various districts that can differ in multiple ways, be it in terms of their infrastructure, their geography or, most importantly, their demographic. For instance, a district close to the city center may undergo transformation while being home to a hugely heterogeneous group of citizens from various social classes. Such districts tend to have objectives that differ significantly from newly developed districts in the suburbs of a city, where a typically rather wealthy and homogenous older population has taken up residence. Given these disparities, SSDs can have a host of highly differentiated objectives, all of which may have specific causal links to the particular characteristics of the respective district. At the same time, however, these SSDs are part of a larger city and its overarching objectives and values. Information systems must, therefore, meet a range of requirements to help the district achieve its individual objectives as well as the greater interests of the city (Bastidas et al. 2018; Heaton and Parlikad 2019). This, however, requires that the district objectives are known and that they have been formulated according to the needs and desires of the citizens. If sufficient attention is not paid to the latter, citizens are likely not to use the IS on a regular basis, nor are the potential uses of specific technologies likely to be fully explored (DP 5 & 6). SSDs mitigate this risk by combining top-down and bottom-up perspectives on a district level (Mora et al. 2019). It is advisable, therefore, that a Green IS in an SSD brings its use of various technologies into alignment to overcome this challenge (DP 5 & 6) (Kumar et al. 2020). SSDs would then ensure accessibility across all ages and backgrounds (Majchrzak et al. 2018).

DP 3 – Response to the Feedback of Citizens: The involvement of citizens should not end with the initiation of the Green IS, nor with its development and implementation. Indeed, it is crucial to involve them further, first to keep ascertaining their needs and then to improve the Green IS accordingly. The focus of this involvement, however, ought to change in the operation phase; from a design to an assessment perspective. This should happen for the entirety of the operation phase, during which a range of communication channels should remain open to ensure that multifaceted feedback can be received from a representative cross-section of the district’s citizen (Kendel et al. 2017; Marsal-Llacuna and Segal 2016). These communication channels should be set up to account for the district’s particular context and special attention ought to be devoted to reaching citizens who are not already using the Green IS, given that its development is an opportunity to convince as yet disconnected groups of the benefits of the Green IS. The SSD can collect such feedback with quantitative monitoring by using smart meters or similar technology (DP 6), or it can do so with more conventional qualitative approaches, such as district meetings (Hopf et al. 2018; Kumar et al. 2020; Palumbo et al. 2021). Crucial to note in this context is the associated workload for the respective administration, and this should not be underestimated. The survey, evaluation, and implementation of any adjustments will require resources in terms of staffing and financing.

DP 4 – Adoption of a Holistic District Perspective: The designer in charge would be well advised to plan the Green IS for a connected district environment (Marsal-Llacuna and Segal 2016). An SSD is a complex construct of built infrastructure, diverse stakeholders, and various services (Keller et al. 2019; Palumbo et al. 2021). By designing the IS with an awareness of the interdependencies that exist between these district components, a Green IS can facilitate constructive interaction (Palumbo et al. 2021). What is more, by adopting a holistic district perspective from its initiation through its operation, the Green IS can leverage synergy effects and avoid island solutions that only work for specific services (Belanche et al. 2016; Palumbo et al. 2021). After all, an SSD connects a wide range of stakeholders, each of whom may wish it to serve a different purpose (Brauer and Kolbe 2016; Keller et al. 2019). Indeed, its many diverse stakeholders request a multitude of services while others provide just as many, including traditional housing services like waste disposal and more innovative ones such as car-sharing services (DP 1) (Brauer and Kolbe 2016). Since many of these depend on the SSD’s built infrastructure (Kendel et al. 2017), the specific objectives of a district, as defined in DP 2, must not be disregarded. Instead, a holistic perspective must take in the full complexity and conflicting goals of different stakeholders. This poses significant challenges in terms of feedback analysis. It also has notable implications for the ongoing re-prioritization of goals within the district.

DP 5 – Facilitation of a Flexible IT Architecture: Given these potential peculiarities of an SSD, a Green IS requires a modular IT architecture to guarantee a scalable and reliable system for its services. Since a holistic perspective promotes interaction between the infrastructural levels (DP 5) (Bastidas et al. 2018), the current consensus in the academic community favors open interfaces in conjunction with approaches that facilitate data transferability (DP 6) (Ruutu et al. 2017). As those in charge of designing the Green IS lay the foundations for these requirements at the development stage, they are of major importance at the operation stage. It is worth remembering, however, how fast these requirements change, especially when dealing with an application in a built infrastructure, designed to last for a long time. It is, therefore, a considerable challenge to design the system in such a way that changing requirements can continue to be implemented meaningfully, even after years of use.

DP 6 – Exploitation of the Full Potential of District Data: Due to the complexity of SSDs, the amount of data is high even in basic SSD systems. Given the rapid ongoing development of modern technologies and services, we can watch in real-time as this wealth of data increases further (Bastidas et al. 2018). There is, therefore, not only an urgent need for an adequate IT architecture, as described in DP 5. There is also a provision of various potential services and applications that can help us tackle the different challenges of an SSD. They may, for example, tap into the potential of DP 3 by meeting the need to monitor the use of the Green IS in order to collect valuable feedback for the improvement of the system and its services (Bibri 2018a). Doing so often requires real-time data sources, which opens possibilities for advanced data processing approaches, such as process mining, pattern recognition, and machine learning (Bastidas et al. 2018; Kumar et al. 2020). It is advisable, therefore, that the Green IS should integrate various data sources, both within the district and beyond (Ji et al. 2021; Lim et al. 2018; Wolff et al. 2020). The large amount of data thus generated requires the Green IS to have a robust data-management approach (Bastidas et al. 2018).

DP 7 – Preservation of Privacy and Security: Privacy and security are key issues in the digital environment of an SSD. The Green IS must guarantee each citizen’s digital and physical integrity in all contexts (Vandercruysse et al. 2020). This means it has to comply with current laws, regulations, and ethical standards to preserve their privacy and security (Keller et al. 2019; Yeh 2017). Such preservation is aided by the ability to include user accounts as well as authentication and authorization functionalities (Majchrzak et al. 2018). It can also be aided by further means, such as rewards for voluntary disclosure of data. This can be done by means of economic incentives or by enabling new functionalities within the Green IS, although these measures require careful consideration as they bear the risk of putting considerable pressure on citizens.

Of the many different types of citizen-centric ISs, mobile applications have shown particular promise in getting close to the needs and user habits of the citizens (Naous et al. 2022). To date, Green ISs have often used this implementation to address multiple areas that bear relevance to an SSC or SSD, including mobility, pollution, recycling, water, or energy. Since members of the public carry mobile devices on their person most of the time, it is easier to reach them via these devices, and easier for them to provide feedback that way, for instance on issues like energy consumption (Bonenberger et al. 2021). With this in mind, we opted for incremental and iterative development of a Green IS mobile app within the project Stadtquartier 2050, working closely with relevant district stakeholders. The goal of this publicly funded project was to supply citizens of two German SSDs in Stuttgart and Überlingen with climate-neutral energy in a socially acceptable way. To this end, 960 residential units were constructed or renovated, whereupon the mobile app aimed to support all of their citizens in living climate-conscious lifestyles. During its development over several years, our design principles were treated as imperatives as far as their contextual factors allowed. Upon completion, the mobile app helped us to reconcile the practical and theoretical insights in the course of a summative evaluation of our DVs (see Evaluation Episode 3).

Throughout this project, and indeed the mobile app development, we made a concerted effort to involve citizens in multiple ways. This focus also heavily influenced the entire process of mobile app development. For example, we held several events and focus groups that gave citizens opportunities to actively participate in the app development (DP1). To optimize the ways in which we ingrained the goals of the project and the objectives of the SSDs in the mobile app, we held two initial requirements workshops with representatives of the citizens and other SSD stakeholders. Here, the main functionalities were aligned with the district objectives (DP2). As a result, the mobile app had the best chance of helping citizens within the SSDs to embrace more climate-conscious lifestyles by making use of these three core functionalities (see Fig. 3): First, the app provided users with near real-time monitoring of their energy and water consumption, not only in detail but also in relation to its ecological and economic impact. Second, it allowed users to set small tasks (e.g., changing the older light bulbs to energy-efficient LED lighting) as actionable goals, such as those used to level up in a gamification component. Third, it drip-fed small bits of information about a more sustainable way of living. Rather than put users off with an overwhelming glut of data, this strategy let them digest snack-sized lessons on energy-saving lifestyle choices by means of an interactive training component. After multiple feedback loops (DP3), additional functionalities were requested by users, such as the integration of local mobility offers (e.g., e-carsharing) and social features, including conveniences like access to the menu of a restaurant in the district. We believe these features should not only incentivize residents to use the app even more. They should also broaden the holistic perspective on the SSD in general (DP4).

What all of these frontend functionalities have in common is that they require a flexible architecture in the backend (DP5). This, in turn, requires consideration of multiple basic factors rarely discussed with the citizens. Examples include a baseline at which various elements of the distributed system interact. Further examples include compliance with standard data protection regulations and data security measures. In the case of the Stadtquartier 2050 project, three main non-functional requirements emerged in early discussions with district stakeholders. First, the mobile app should be used by citizens of two geographically-separate SSDs with their own technical hardware infrastructure. Second, the citizens of these SSDs should be able to use the app irrespective of the operating system of their mobile device. Third, the consumption data should only be accessible to the respective user upon initial registration and ongoing authorization. These requirements had a significant impact on the resulting architecture, as depicted in Fig. 4.

To ensure the architecture would remain flexible and expandable, we developed a communication layer able to combine multiple data streams from various (future) service systems. One of these is the artificial-intelligence-enabled self-learning grid system that leverages anonymized consumption data on the district to generate sustainability-improving notifications for users of the mobile app (DP6). To make this possible, individual measurement sensors provide data, a virtual machine uses particular scripts to import this data into a database, and that database is made accessible via a secured application programming interface. Another virtual machine receives all requests from the mobile app, stores the core and transaction data required to operate all of the main functionalities, and forms the communication layer. At regular intervals, the system receives anonymized consumption data and offers updated notifications to specific citizen groups based on the respective data. During the application of the DPs, we were able to note several challenges (see Table 4) that affected the progress of the mobile app development. One such challenge in implementing the DPs was to observe all data protection laws pertaining to the rights and interests of citizens (DP7) as the legal situation and jurisdiction in the project Stadtquartier 2050 are neither transparent for the layperson nor designed for novel uses such as an app that leverages district data for sustainability purposes. While external experts were consulted to guide the development of an extensive data protection and security concept, and while the necessary revisions increased the quality of the work for all concerned stakeholders, they introduced extra challenges by delaying the development process for some functionalities (e.g., testing the detailed energy and water consumption display with real-world data). We resolved this by establishing a registration server located outside the district server infrastructure. This server made it possible for citizens to register with a pseudonym when using the app for the first time. Meanwhile, a clear data separation was implemented in the backend of the mobile app by tri-partitioning the data on the app virtual machine, the measuring data virtual machine, and the registration server. This proved very helpful in handling sensitive energy consumption data with care. It is worth noting, however, that this mobile app is only one of many possible manifestations of a Green IS in the SSD. In this case, we focused on energy and mobility, but the DPs that governed this application can also be implemented in other areas of sustainability.

Our theoretical contribution is a nascent design theory that produces design knowledge in the form of operational guidelines for a Green IS in an SSD environment. Current research stands to benefit from this district perspective as it applies the fundamental pillars of IS Design on a smaller scale than the much-discussed SSCs. As such, it is more relevant to the everyday lives of most citizens (Graf-Drasch et al. 2022; Keller et al. 2019; Mattoni et al. 2019). Following the rationale of Gregor and Jones (2007), namely that a design theory should consist of eight components, we list the specific components that relate to design knowledge and originate from this study in Table 6. This design knowledge is the essential theoretical contribution of the present paper. It extends the prevailing top-down approaches by introducing a citizen-centric bottom-up approach to the design of Green ISs. Ultimately, this promises to advance research in the domain of sustainable systems engineering (van der Aalst et al. 2023).

The seven DPs presented in these pages have a range of practical implications for the various stakeholders in SSDs, including citizens and service providers. Citizen-centric development benefits the design of Green ISs in SSDs in a number of new ways. First, it reduces the risk that the developed Green ISs will remain unused, as has happened with many top-down approaches. Second, it fosters transparency as well as a positive relationship with Green ISs, given that the citizens who are meant to use it were part of its creation (Heaton and Parlikad 2019). Third, it helps citizens understand how to use the Green IS as they come to appreciate its underlying principles (DP 2). In part, this is achieved by actively involving them in the design process (DP 1), specifically in the initiation, development, and implementation phase. The benefits are leveraged, however, by further involving citizens in the operation of the Green IS, where they provide feedback on the status quo (DP 3). If such feedback is collected continuously and subsequent updates made accordingly, the Green IS remains citizen-centric and up-to-date, despite changes in the needs of the citizens or the specifics of their environment (Bibri 2018a; Keller et al. 2019). Further worth noting is the fact that these DPs also provide guidance for system designers at the various stages of the IS development process. Meanwhile, adherence to the DPs assures a certain level of user-centricity (DP 1,3), a holistic perspective of the district (DP 4), and the use of safe, expandable technology (DP 5,7). As our research indicates, a Green IS must be scalable to respond to changing requirements, for example, substantial increases in data volume due to the use of big data-related technology or new functionalities (Bastidas et al. 2018). It must also be scalable in case the district undergoes spatial expansion (Kumar et al. 2020), and, as we have seen, it must remain flexible and extendible to enable the independent development of its components and the interoperability with other systems (Del M. Esposte et al. 2019). The Green IS designed for this project complied not only with the seven DPs discussed above but also with specific district objectives, by virtue of which it will be especially helpful to SSD managers wishing to realize these objectives (DP 2). As our findings indicate, Green IS designers must also be aware of potential conflicts of interest between different demographics. Older or disabled citizens, for example, may have the same need for Green IS services in an SSD yet find such services harder to access (Renyi et al. 2019). Associated benefits include those reaped from the fact that, since Green ISs often supports citizens in their daily lives, the SSD manager stands to make a profit from satisfied citizens in the SSD if an appropriate level of privacy and security can be assured for services accessed through the Green IS (DP 7). A solid privacy and security concept is, therefore, not merely a legal necessity but also a lucrative opportunity to foster trust and service usage among the citizenship of the SSD (Bastidas et al. 2018). Without confidence in its safety and security, many people may refuse to use any Green IS that impacts their personal life and property (Lim et al. 2018). In short, data protection is key to a durable and sustainable use of a Green IS (Vandercruysse et al. 2020). To bridge the gap in the current research on Green ISs in SSDs, we established a range of robust DPs, as depicted in Fig. 1. DP1 and DP3 are located at the center, where they act as touchpoints between citizens and services & processes and technologies & information. DP2 is linked with goals, as opposed to DP4 which is linked with contextual factors. DP5, DP6, and DP7 are most closely connected to services & processes or technologies & information, depending on their particular manifestation when implemented.

There are, of course, certain limitations to the general validity of our results, and these limitations, inevitable in a field as novel and vast as the science of sustainability, indicate promising starting points for future research. The first one to note is that we conducted 15 interviews with experts from research and practice, all of whom were originally from Germany, meaning that further interviews with experts from abroad will have to be added to ensure the quality of the DPs when used in an international context. Similarly, our prototype was developed for use in locations around Germany, a geographic limitation that could lead to problems when applying the DPs in different countries. The international background of our literature indicates that our DPs are usable in most Western civilizations, but this assumption will have to be evaluated in future research. Secondly, the vast number of districts meant that our DPs were formulated very broadly. When observing DP 2 at the stages of initiation and development, they will have to be adapted to suit the SSD under consideration. Future research could do so by using case studies for standardized district types and real-world implementations. Also worth noting here is that challenges remain with regard to achieving district goals (DP 2) in harmony with a holistic perspective (DP 4), and resulting interdependencies will have to be subject to future research. Thirdly, the prototype presented in these pages is an instantiation limited to the interests of citizens and does not involve service providers, utilities, or the city government. Furthermore, it is currently in the implementation phase where its focus is limited to ecological sustainability, without due attention being given to significant social aspects. At this early stage, we have no information on how the prototype’s approval rating among citizens will develop over time. Fellow researchers may, therefore, wish to evaluate our DPs in different cultural and national environments to ascertain their validity for a broad range of districts and a fully implemented, holistic prototype. A final important consideration to mention here is that the public sector exerts a notable influence on districts by means of regulation. Despite explicit demands voiced by the research community, contributions to the Green ISs literature are still few and far between, particularly with regard to the public sector perspective (Fernandez et al. 2020; Lehnhoff et al. 2021; vom Brocke et al. 2013). Since these calls have largely gone unheeded, many questions remain unanswered, particularly concerning the design of citizen-centered Green ISs in an urban context (Corbett and Mellouli 2017; Harnischmacher et al. 2020). Research in this area should help to appreciate how restrictions or incentives affect the design of Green ISs in SSDs.