Digitizing Complex Tasks in Water Management with Multilevel Analysis

Abstract

In the course of the debate about suitable digitalization solutions in the water industry, it is becoming clear that planning and implementation are characterized by growing complexity. Aspects such as sensor selection, IoT, cyber security, and artificial intelligence are shaping the public debate when it comes to digitalization solutions in the water industry. However, they only represent a small part of a comprehensive and holistic package of measures for a necessary digitalization system. Rather, a functional-systemic approach and implementation method is required, in which, in addition to purely digital aspects, water management-related topics and questions of organization as well as the role of people in these processes must be answered. As a consequence of this need, the Hof University of Applied Sciences has developed a special form of a “multi-level analysis” method for the parallel survey of complex water management challenges, the identification of digitalization priorities, and the definition of an ideal starting point for the implementation of measures.

1. Introduction

Digitalization has become a key issue in many areas of the economy and society [1, 2, 3]. This development can also be observed internationally, the biggest developments were found in the sector of product development and solution offerings [4]. The results of the first meta-study on digitalization in the German-speaking water industry, conducted in 2021, confirmed that digitalization in the water industry is no longer a marginal topic, as was the case just a few years ago [5]. However, there is still a noticeable gap between scientific project ideas/industrial products and consistent implementation on the municipal side in the water industry. Of more than 1000 identified digitalization products, ideas, and projects in the biggest meta-studies about digitization in the German-speaking water industry, named WaterExe4.0 and DigiNaX [6].

Only 11% were identified as realized solutions in the municipal environment. In addition, many of the scientific ideas are still too far away from being ready for the market. More than 70% of the scientific topics identified are still at the study stage and have yet to be implemented (see Figure 1).

Figure 1.

Allocation of digitization products, ideas, and projects in WaterExe4.0 [1].

Although a large number of digital solutions are now available on the market for almost every task in the water industry, the meta-studies confirmed the still strong reluctance on the part of public clients in the water industry when it comes to digitalization issues. Even if the water industry is more willing to innovate than the general public sector due to the technologies required [2], many of the benefits of digitalization are not being used to improve measures to address the challenges of climate change [3]. This was already demonstrated by the first major research projects on digitalization in the German-speaking water sector, SMADIWA [4] and KOMMUNAL 4.0 [5], which were able to show between 2016 and 2018 that many obstacles were due to non-technical causes. This was confirmed by the meta-studies. However, the vast majority of the projects identified as part of WaterExe4.0 and DigiNaX focused on technical feasibility, solution development, and application goals [6]; non-technical aspects hardly played a role. Topics that affect users mainly relate to information on user-friendliness, the standardization of data and structures, application-related settings at the software level, or, in some cases, legal issues. One exception was the topic of cybersecurity, which was analyzed in depth from both a technical and application-related perspective. Aspects of training and the added value of data were also considered in some studies [7].

Overall, it is noticeable that there is a lack of scientific studies on innovation dynamics in the water sector, both nationally [8] as well as internationally [9]. Although the urgent need for water innovation is becoming increasingly clear, especially in the context of climate change [10, 11] financial investment in the water sector lags far behind that in sectors such as energy [12]. It is not only in Germany that the water sector is considered less innovative than other sectors.

The water sector appears to invest far less in research and development than other sectors [13]. One reason for this is that the water sector worldwide is dominated by suppliers rather than consumers [9]. In addition, the water sector is dominated by municipal organizations, which leads to a lower willingness to innovate and is often even associated with a rejection of change [14]. This has a direct impact on the willingness to use innovative digital solutions as part of necessary changes.

It is not only the German-language meta-studies that show that a sole focus on technical aspects of digitization projects in infrastructures is considered insufficient. International studies also point to this problem [15, 16]. Although research into smart water solutions is increasing worldwide [17, 18, 19], the conceptual, technical, and practical gaps between providers and customers have not yet been sufficiently closed [20]. In particular, there is a lack of studies and concepts that illustrate the added value of a digitalization solution and clarify how a digitalization solution can be integrated into everyday working life and does not present any additional barriers [21].

In supplementary online surveys and expert interviews conducted at meta-studies WaterExe4.0 and DigiNaX, the question of possible reasons for the previously described strong reluctance on the municipal side to implement digitalization projects was investigated. A total of almost 400 survey participants and 30 experts from the German-speaking water industry interviewed between November 2020 and August 2022 provided insight into the motives and reasons why there is still a relatively strong reluctance to tackle many digitalization issues. Based on the results of preliminary studies and the meta-study authors’ own professional experience [8], it was known that innovation decisions are usually based on multi-criteria decisions (see e.g. [22, 23]). This was also to be expected in the context of digitalization in the water industry and was particularly evident in the free expression of opinion in the expert interviews, where the complexity of decision-making processes in relation to digitalization processes in the German-speaking water industry became clear. In the WaterExe4.0 and DigiNax studies, both the participants in the online survey and the experts interviewed identified various factors as important influencing factors for potential users. These assessments are strongly based on the personal experience and professional training of the survey participants and experts. This must be taken into account in the assessment.

Due to the limited scope and heterogeneity of the studies, a simplified multi-criteria analysis was created to adequately evaluate and compare the two parts of the study. To this end, all quantitative queries and qualitative statements were checked for consistency of content. The focus was on the success factors of digitalization, so consequently, only the success factors mentioned by the study participants were considered for the multi-criteria evaluation. A total of 19 common criteria were identified by all study participants (see Table 1). The importance of the expert statements was considered by multiplying the number of expert arguments by a factor of two (one argument = two points). The arguments of the study participants in the online survey were not adjusted (one argument = one point). The increase in the rating of the expert statements was necessary in order to appropriately assess the special knowledge and experience of the experts in comparison to the survey participants.

Success factors mentioned by participants from online surveys and experts	Online survey ratings	Expert ratings	Total ratings
Opportunities to drive digitization forward
Pilot project/best practice	3	16	19
Generational change/cultural change/readiness for change	2	6	8
Change in cooperation	1	4	5
Developing the management level	3	2	5
Engaging and listening to staff	4	4	8
Common database for all areas	3	2	5
Success factors in digitization projects
Competence/know-how	3	8	11
Willingness of staff	1	8	9
Recognizable added value	15	22	37
IT security	4	10	14
Key person (CEO/responsible person)	2	14	18
Transparency	5	2	7
Acceptance (staff and people)	10	22	32
Economic aspects	1	4	5
Connectivity/networkability	1	2	3
Interfaces	1	2	3
Overall strategy	4	14	18
Suitable (external) partners	2	10	12
Promising technologies/solutions of the future
IoT	1	2	3

Table 1.

Comparative results of online survey and expert interviews from WaterExe4.0 [8].

Table 1 shows that the factors “Visible added value” with 37 points and “User acceptance” with 32 points are the most important success factors from the perspective of the respondents and experts and are considered not enough in most digitalization projects. In the KOMMUNAL 4.0 research project¹ the lack of a clearly describable added value of a digital solution for the individual use case was already seen as one of the biggest obstacles to decision-making [24]. Participants in special workshops on the topic of cost-benefit and competence requirements, which were carried out as part of WaterExe4.0, emphasized the need for a tool or method to be able to carry out a validated benefit assessment for a digitalization solution in relation to the individual case [25].

2. More than 400 individual criteria lead to a high degree of complexity in digitization projects

As part of a detailed multi-criteria and interdisciplinary consideration of influencing factors in digitization projects in water management [26] all content and concept components were analyzed for their significance or (impact) effect on a benefit analysis against the background of a high degree of complexity. In total, more than 400 individual criteria were identified.

These were divided by [26] into several main and sub-categories. In addition to the selection of suitable evaluation criteria, the evaluation of modern and innovative digitalization solutions also depends strongly on a sufficient understanding of the associated complex interactions between the digitalization solution and the area of application. Whether such an evaluation is expedient also depends on the existing competence of the evaluators, who may be the only users capable of identifying and evaluating the sum of the interactions of the individual use case (see Figure 2).

Figure 2.

Impact triangle evaluation result [26].

The user’s background plays an important role in learning new information technologies. [27] emphasize, for example, in an analysis of application training for new information systems, that three areas of knowledge are decisive, which should be taught in training programs:

• Application knowledge that relates to the commands and tools embedded in IT applications

• Knowledge of the business context with reference to the use of IT applications to perform business tasks effectively and

• Knowledge of collaborative tasks that relate to how others use the application in their tasks.

The fact that these findings are neither taken into account in plans for the digitalization of water management nor in the associated user training courses was confirmed by the results of the expert interviews in the WaterExe4.0 study, in which the experts described their experiences of digitalization projects that had been carried out [28]. They emphasized that both users and solution providers need a systemic understanding of the impact triangle according to Figure 2 is hardly developed or trained and leads to corresponding reductions in success. Against this background, the evaluation criteria in question must be directly linked to the competence of the evaluator or the missing competence areas of the impact triangle must be included in the digitization system.

The importance of non-technical complexity in particular as an influencing factor in the course of decisions for or against a digital solution was one of the key findings of the WaterExe4.0 and DigiNaX meta-studies. The underlying complexity must therefore be analyzed in detail for each evaluated project. Every definition of complexity depends on the perspective that is applied to it [29]. Many definitions of the underlying complexity theory can be found in the scientific literature, but there is no standardized description for complexity [30, 31] that has attempted to identify a core of complexity theory based on various studies to illustrate the understanding of complexity.

From the perspective of organizations, [31] describes the core consisting of the five aspects of action, fitness, strategies, predictability, and emergence (see Figure 3) and emphasizes the specific causal relationships between them.

Figure 3.

Causal relationships between aspects of complexity theory [31].

Baccarini [32] identifies two key aspects of complexity from a project management perspective (see Table 2), whose perspective was particularly considered in the development of the multi-level analysis presented in this paper.

Definition approach complexity	Description
Many different, interconnected parts	A clearly defined description that allows the associated complexity to be operationalized into the various elements of the system under consideration (in the case of project management, e.g. tasks, specialists, components, dependencies, connections, etc.). This interpretation of complexity reflects systems theory, as a complex system is often defined in terms of differentiation and connectivity.
Complicated, intricate, intertwined	This rather older meaning of complexity is open to a broad and diverse interpretation. For example, it can be interpreted as encompassing everything that is characterized by difficulties. This interpretation represents a predominantly subjective connotation and describes, among other things, difficulties in understanding and dealing with an object. Therefore, this interpretation of complexity is in the eyes of the beholder. In many cases, this meaning of complexity can be better subsumed under the term uncertainty.

Table 2.

Overview of complexity definitions and forms [32].

Furthermore, volatile markets in which a high level of product diversity mitigates unfavorable relationships between fluctuations in demand and inventories/capacity utilization must also be considered [33]. This is typical of the water industry. An increase in product diversity is one of the other main reasons leading to an increase in complexity in technology-defining markets such as the water industry. The main aspects that lead to an increase in product diversity are, according to [34]:

• The market is forcing companies to introduce new product variants that must comply with different national or regional regulations and requirements.

• Customers are using their power and exerting pressure on manufacturers to further differentiate their range.

• The high level of market saturation is driving companies to open to niche markets and thus serve smaller customer groups with special requirements.

• Customers want greater customization and a larger selection of new or additional product functions.

As a result of climate change in particular, water management is increasingly being confronted with previously less pronounced aspects such as conflicting interests, additional stakeholders, new roles, and a different distribution of resources [35] in addition to the previously known human and physical (or technical) dimensions. Capturing the physical dimensions alone has its own high degree of complexity, as the problems become increasingly interconnected across spatial and temporal scales (for example, climate-related impacts such as flooding or water scarcity have a cross-border effect). It is a challenge to determine exactly where interventions lead to desired outcomes [36]. The supporting human systems are also already highly complex regardless of the sphere of influence (local, regional, national, or international). Water systems are rarely controlled by a single actor or institution. Moreover, water systems such as drinking water systems, rivers/water navigation, or watersheds extend across political boundaries, but responsible organizational departments continue to focus their jurisdiction on politically defined areas. Effective water management therefore requires the coordination of decisions between different stakeholders, including across national borders.

The more attempts are made to map this complexity in digital solutions, the greater the diversity of products and variants. On the other hand, digital solutions are created with an enormous scope of services that in many cases are not fully required. However, the user side is also actively involved in the increase in complexity. On the one hand, if requirements are described too imprecisely, whereby the solution is developed through speculation on the provider side, which is sometimes necessary [37].

On the other hand, for example, by listing as many requirements as possible when creating specifications or service descriptions for the creation of digital solutions without first examining their relevance and depth of impact in detail.

Another reason for the increase in complexity is heterogeneous IT landscapes, some of which have grown over decades [38]. This is typically for the water industry. In addition to a large number of different applications for various tasks, the number of variants is also increased by different core technologies, operating systems, database systems, or data integration technologies [39]. Even in cases where solution providers suggest combining IT heterogeneity in a centrally accessible system using intelligent integration systems such as Microsoft Azure in order to at least reduce operating and data processing complexity, municipal users leave it at the existing heterogeneity, as they want to avoid actively analyzing and dealing with the existing complexity due to a feeling of being overwhelmed [40].

When adapting or expanding in complex, heterogeneous IT landscapes, it should therefore be considered in advance whether a complete replacement of the outdated systems should be planned (so-called refactoring) or whether a gradual replacement that takes place in parallel to the existing system is the better approach (so-called strangler application pattern) [41]. This decision can have a significant impact on future complexity. In order to capture the high degree and variety of complexity in water-related projects, make it visible, and transfer it into an initial structure, researchers at Hof University of Applied Sciences have developed a special multi-level analysis [42]. On this basis, the existing complexity can be made more transparent and leads to better and faster decisions.

3. Transparency of complexity increases the success of digitalization

Cities and municipalities are increasingly confronted with various crises that occur at ever shorter intervals and have a parallel impact over longer periods of time. Whether it is the (still ongoing) financial crisis, the coronavirus pandemic, the shortage of skilled workers, the tangible consequences of the war in Ukraine (energy and refugee crisis), or municipal over-indebtedness, all crises require attention and generate pressure to act. Since the 1992 Framework Convention on Climate Change [43, 44], the consideration of climate change at the local level has been an essential part of municipal tasks (e.g. flood protection in the event of heavy rainfall). Climate change has only been recognized as an acute crisis in Europe since 2019 [44] especially since the heavy rainfall events in Belgium, the Netherlands, and Germany in the summer of 2021. On the one hand, this was triggered by the extreme damage caused by the heavy rainfall events in July 2021 (including in the German Ahr Valley), while on the other hand, the last few dry years have shown how much groundwater levels have fallen (and continue to fall), making the water supply in many parts of northern Europe more vulnerable than previously assumed. No matter what local, national, or global measures are taken to reduce global warming and implemented in the near future, the current consequences of climate change can only be mitigated in years or decades.

During several projects on the digitalization of water management, researchers at Hof University of Applied Sciences were confronted with the perceived excessive demands, coupled with diffuse fears, of around 100 representatives from German municipalities and specialist planners, which are caused by an increasing complexity of tasks and topics. Those responsible for local authorities often spoke of excessive complexity and of not knowing how and where to tackle it. The high complexity of aforementioned issues with their strong interdependence leads to great fears on the municipal side that mistakes will be made and wrong decisions taken.

Many of the project participants wanted a way to make the permanent increased complexity more transparent. Many years ago, normal day-to-day work in the municipalities was still characterized by the technical separation into transport, construction, wastewater disposal, water supply, social affairs, economic development, health, etc. Municipal tasks were usually handled independently of each other in the respective specialist departments. As a result of the increasing number of multiple crises, it is becoming more and more clear to local authority managers that issues that were previously considered in isolation are highly interconnected, although the structures are often still the same. As a result, increasing inter-thematic cooperation is required, but many local authorities are overwhelmed by the need for holistic processing, which on the one hand leads to an excessive focus on specialist responsibilities and on the other hand to a strong inhibition in implementation. In view of these increasingly noticeable obstacles to implementation, a special multi-level analysis (MLA) was developed by Hof University of Applied Science as a solution measure to make the various levels and types of complexity visible to the background of digitalization projects in water management.

With a complexity analysis, the affected fields of action can be better identified, the problem areas to be addressed can be defined more precisely and the necessary measures can be derived in a more targeted manner. The first step is a project-specific description of the various levels of complexity as shown in Figure 4.

Figure 4.

Municipal complexity cycle [26].

This means:

• Extra complexity:

  • Complexity that is external to an organization, e.g. laws, social developments, or legislation.

• Intra-complexity:

  • The complexity within the organizational structures and the processes of an organization.

• Inter-complexity:

  • Requirements to be fulfilled to meet the increasing needs of interested parties (e.g. citizens, authorities, interest groups).

The approach to dealing with the existing complexity is based on various key questions. Exemplary questions are:

• How can complexity be captured and described?

• How do selection and filtering work?

• How does a change in dynamics become visible?

• How can weighting and prioritization be meaningful?

• How can intra- and inter-complexity be managed effectively and efficiently?

• How are potential solution concepts integrated into existing structures and processes?

In various research projects at Hof University of Applied Sciences on digitalization in water management^2,3,4,5,6,7 it emerged that, in addition to the type of complexity, two main perspectives determine how complexity is dealt with and how action measures are derived. The first is the spatial perspective, i.e. where digital solutions affect crises or challenges. Here necessary digital solutions must be placed. Secondly, there are target perspectives to be considered, i.e. how and what is to be achieved with digital measures. An important aspect of the study was analyzing which municipal departments should be included against the background of the main perspectives presented before, as well as which cross-cutting issues play a role and should be considered in the further course (see Figure 5).

Figure 5.

First step toward complexity transparency in the municipal approach to climate change [26].

In this way, the municipalities involved in the projects can be offered a multidimensional complexity analysis for further processing. The multi-level analysis was developed as a suitable tool to take into account the overarching municipal complexity aspects and to be able to go into sufficient technical depth.

4. The multi-level analysis tool explained by the example of digital rainwater management at an elementary school

The spatial complexity perspective is the core of the multi-level analysis (at which local point digital solutions have a positive effect on the respective problem or addressed challenge). Against the background of various projects at Hof University of Applied Sciences, it became clear that almost all stakeholders approach the problem discussion from a spatial perspective.

The spatial approach is understandable. In water management and similar infrastructure projects, the spatial object is generally assumed. This applies both to the observation that water flows from A to B and that wastewater has to be transported from C to D before it can be treated in E. Spatial observation is also the focus when analyzing precipitation, as rain falls locally. It is becoming increasingly clear that heavy rainfall does not necessarily occur where the greatest damage is later detected. The placement of the sensors is a purely spatial decision.

Since a new tool like the described multi-level analysis, which also involves a new way of thinking and analyzing, is associated with fears of the new/unknown, reference should be made to a familiar way of behaving [45]. Because of this, the spatial perspective is placed at the center of the processing tool as a well-known way to do analysis in the water industry. Usually, water management and environmental measures start at a local point, e.g. a groundwater well, a pumping station, or an emission source (e.g. a machine at a production site). This smallest point of observation is named as the micro level in the multi-level analysis. From this point, further levels are defined as meso-, macro-, and meta-level. This is done in the form of larger spatial “circles” that are drawn around the starting point, the micro level (see example in Figure 6).

Figure 6.

“Circular” spatial level assignment of the multi-level analysis in the Schauenstein elementary school digital rainwater management project [26].

Table 3 shows the initial definition of the chosen levels that have proven to be suitable for analyzing municipal infrastructure tasks. In the previous use cases examined at Hof University of Applied Sciences, these four levels have proven to be expedient. Further levels can be added if required.

Level	Description
Micro level	Smallest action area (e.g. building/object/property)
Meso level	Surrounding area of the micro level (e.g. neighboring properties, residential area, city district, commercial area)
Macro level	Superordinate area of the micro level (e.g. city/municipality/transition area to the next city/municipality)
Meta level	Superordinate macro-level area (e.g. inter-municipal action area of two or more municipalities/cities, region, district)

Table 3.

Level definition of multi-level analysis [26].

The methodology of the multi-level analysis will be presented based on the digital rainwater management of the Schauenstein elementary school, a research project of the Hof University of Applied Sciences. The subject of the research project is the central rainwater management for the school properties including all buildings. The core elements of the investment project are a special green roof on the gymnasium and a sufficiently large rainwater storage tank which, monitored by sensors, provides the necessary flood protection during heavy rainfall, and supplies the necessary water to green areas and the green roof of the gym during dry periods. In Schauenstein, a Purple Roof from the manufacturer Sempergreen⁸ was installed in 2023 as the first installation in Germany on the roof of the gym. In contrast to conventional intensive green roofs, the Purple Roof has an integrating and runoff-reducing fabric mat at the base of the roof (see Figure 7). The arrangement of densely packed polyester threads creates the effect of a friction layer. A honeycomb arrangement of vertically aligned tubes creates additional storage capacity. Both components provide runoff resistance over the entire profile. Conventional (intensive) green roof structures also have a (lower) retention function, but this is achieved exclusively through the function of runoff throttling [46].

Figure 7.

Types of superstructures for intensive green roofs [46].

The overall digital system is to be designed in such a way that different rain scenarios can be simulated with the water storage tank and the connected irrigation system (also for teaching purposes) and the effects can be measured and thus made visible with the help of the high degree of digitization.

The spatial level assignment is the first step of multi-level analysis (Figure 6). As a starting point, the elementary school building was chosen and defined as the micro level. It was also important to consider the direct influences on and from the primary school (micro-level) on the immediate surroundings. For this purpose, the residential and property area neighboring the school was selected as the next spatial level (meso-level). Other possible influences were considered by selecting the center of Schauenstein as the macro level and the other districts of Schauenstein as the meta-level.

Once the four spatial levels had been determined, the next focus was placed on the level at which the central complexity analysis was to be carried out. Once the spatial levels had been assigned, the starting point and object of investigation were mainly at the micro level. The next step involved asking general questions to identify aspects of complexity and integrating the three types of complexity (external, internal, and intra), and linking the types of complexity and the four levels (see Figure 8). The analysis was supplemented by the time factor, which represents a separate dimension or level. When considering time, the past, present, and future were analyzed in relation to the task. Individual aspects of complexity or action can then be assigned to these three time periods. Typical examples of a past assignment are contracts or ongoing approvals that need to be considered when implementing measures (e.g. lack of rights of way for required pipelines).

Figure 8.

Relationship between complexity types and multi-level analysis [26].

For the rapid realization of selected measures, the present must be appropriately included as a dimension. This has been shown by several studies involving the Hof University of Applied Sciences. The ideal approach here is to tie in with infrastructure projects that are already planned, approved, or underway. This demonstrably facilitates the implementation of measures, accelerates projects, and leads to a significantly higher level of acceptance among the local authority representatives concerned [1, 5, 24]. This strategy, known as so-called “Anyway strategy” [6], is strongly oriented toward the conditions of everyday working life and thus significantly reduces intra-complexity. Further details on this can be found at [6]. The inclusion of the future dimension in the multi-level analysis resulted from another project at Hof University of Applied Sciences, which involved the consideration of five German municipalities in the northern Bavarian region of Upper Franconia that want to jointly develop into a water-sensitive region. In the various discussions, the aspect of “dealing with climate change” was increasingly raised as a kind of ethical or social value. The municipal representatives were asked to present good measures toward a water-sensitive municipality/region in such a way that they can be used as a vision for the future, making the municipality particularly livable and attractive, especially for young families looking for a new home.

The digital rainwater management system, for the green roof and the rainwater storage tank, is being developed as a digital twin to try out various case studies before a real-life test,

In particular, the adaptation of irrigation and drainage of the green roof should be analyzed with the scenario option. The question to be answered here is how long the green roof of the gymnasium can be supplied with rainwater when drinking water must be supplied (and how much), and under what conditions all other plants survive in different dry periods. This is followed by the process management of the storage tanks. The impact of various heavy rainfall events on the elementary school facilities will also be analyzed. To ensure that primary school pupils can also use this system, a child/primary school-friendly version will be added. Table 4 shows an excerpt from the multi-level analysis carried out at Schauenstein elementary school [26] assigned to the four levels.

Level	Spatial allocation	Aspects of the investigation
Micro level	School grounds with buildings, paths, and areas	Green roof on gym; water accumulation and collection from paved areas (roofs, paths, etc.); digital rainwater management; digital twin; primary school variant of rainwater management; water requirements of green roof and plantings and determination of water supply security; creation of a special water balance
Meso level	Residential and property area adjacent to the school	Influence of surface runoff to the micro level incl. Supply line to the rainwater tank (if rainfall on the micro level is too low); influence of water retention of the micro level on the combined sewer; securing the water supply if drinking water is required for irrigation on the micro level; transfer of the results to other properties in the neighborhood.
Macro level	Schauenstein city	Influence of micro-level water retention on combined sewer; securing water supply if drinking water is required for irrigation at the micro-level; transfer of the results to other properties in the city.
Meta level	Schauenstein center with districts	No detailed consideration

Table 4.

Summary of results of the multi-level analysis of the Schauenstein digital elementary school [26].

When using green roofs as part of a (heavy) rain management system, it is particularly important that green roofs remain sufficiently moist, otherwise, their functions will be lost. For example, intensive green roofs in particular require proactive irrigation during prolonged dry periods [47]. The development of the water balance required for Schauenstein elementary school yielded the following results (Table 5) [48]:

	3a	5a	10a	100a
Retention memory (safety value for calculation)	21.12 (2 × 10.56) m^3
Detention storage	37.027.6 l
Required size of collection container	67.23 m^3	76.66 m^3	90.35 m^3	142.82 m^3

Table 5.

Water balance calculation results for Schauenstein elementary school [48].

Taking into account the functional requirements of the Purple Roof (green roof) and the water balance, the following requirements result in the technical implementation and thus for the digitalization:

• Remote-controlled pump (submersible pressure pump)

• Remote-controlled valves/fittings/gate valves

• Level measurement for cistern

• Remote-controlled drip irrigation system

• Drinking water replenishment for the cistern

• Weather station

• If possible, very flat moisture sensors for the roof

• Separate sensors for all trees

• Flat roof monitoring

  • Constant monitoring of the condition of the roof cladding in order to be notified immediately in the event of leaks

From the analysis of the water balance, indications for irrigation (see Table 6) and various control scenarios according to Figure 9 were derived.

Purple roof	Other plants/vegetation
In extreme heat Max. 2 weeks until substantial plant loss occurs In moderate heat Up to max. 1 month until substantial plant loss occurs If the substrate will be very dry (< 10% retention volume) Deep soaking Fill 1 × total retention volume In normal weather: Water once a week With a desired cooling capacity 10–20 mm/m^2	Possibly young/ very young trees Water once a week Very likely lawn area Water once a week To stimulate deep roots

Table 6.

Instructions for the irrigation of Schauenstein elementary school [49].

Figure 9.

Process diagram for rainwater management at Schauenstein elementary school [49].

Even for a small project such as the renovation of Schauenstein elementary school, the number of criteria to be considered when applying the multi-level analysis should not be underestimated. For this reason, Hof University of Applied Sciences developed its own Excel tool for using multi-level analysis. With this tool, many individual criteria per level and thematic aspect can be considered or processed, including the evaluation of the individual criteria and weighting of the thematic aspects.

For the actual complexity analysis with the multi-level analysis [26] divided the approximately 400 individual criteria into 17 standard categories, as shown in Table 7. These are supplemented by an 18th category, “Project-specific requirements”, which can be used to record the project-specific criteria. Table 8 shows the selected standard categories and the individual criteria identified for all categories (including the 18th category “Project-specific requirements”).

Standard category	Description
(1) Meta category	This category can be used to show the stage of development of a potential digital solution (e.g. “sufficiently market-ready” or still in the “concept phase”).
(2) Area of application	Selection of the area of application, e.g. “drinking water” or “waste water”.
(3) Area of application/possible use	In relation to (2), this category can be used to determine what the solution is to be used for (e.g. “water extraction” or “distribution/drainage”).
(4) Target definition of the digitization task	This category is used to determine how well the required target definition of the digitalization task has already been described or what is still missing (e.g. “Needs assessment on the user side has been performed by the user” and/or “Expectations of third parties are known and defined”).
(5) Functional objectives	In addition to the previous categories, extended objectives can be specified, such as “pollutant-free or pollutant-reduced mixed water drainage” or “measure against shortage of skilled workers in relation to staff costs in the company”.
(6) Benefit aspects	In order to emphasize the benefits of the objectives in particular, appropriate additions can be made in this category (e.g. “cost savings” or “transparency of processes”).
(7) Variants	The variants category allows statements to be made on flexibility and expandability (e.g. “freedom of choice” or “expandability”).
(8) Basis of the decisions	In order to reflect on the decision-making process and assess its influence on a solution, it can be categorized (e.g. “low level of uncertainty” or “high level of uncertainty”).
(9) Data basis	In this category, the data to be used for a potential solution can be evaluated (e.g. “Historical or parallel scenario analysis” or “Forecast data from external sources (e.g. precipitation)”).
(10) Risk analysis	Here you can specify whether and in what form the digital solution can perform risk assessments (e.g. “water pollution” or “runoff damage potential”).
(11) Control methods	If a digital solution goes beyond pure data analysis and is intended to intervene in machine control systems, this can be defined (e.g. “integrated (system-wide) control” or “predictive control”).
(12) Checking the plausibility of results and operational safety	This category can be used to assess data reliability and the determination of results from data (e.g. “Frequency and effort of verification (calibration)” or “Effort for determining the basic data and models”).
(13) Business-as-usual potential	Assessment of whether and to what extent a digital solution already considers or can consider business-as-usual potential (e.g. “are parts of the new digital solution already a task of the business-as-usual project?” or “according to which criteria?”).
(14) Network width	Category for assessing the scope of application (e.g. “intermunicipal” or “subsystem”).
(15) User	User description (e.g. “small municipality” or “group equivalent”).
(16) Existing information systems	Assessment of the existing information system (e.g. “current status insufficient” or “integration capability of local controls”).
(17) Organization	Assessment of the influence of organizational aspects on the digital solution (e.g. “organizational culture appropriate” or “organizational structure sufficient”)

Table 7.

Overview of the 17 standard categories of the multi-level analysis [26].

Category	Individual criteria
(3) Area of application/possibility of use	Utilization/utilization; distribution/drainage; dynamic sewer network management; data accuracy; reliability; maintenance reduction; protection against vandalism; improve microclimate (E3–4); use for political decisions (E-4); increase in value of the municipality (E4); role model function (E-4); starting point for “smart city” development (E4);
(4) Target definition of the digitization task	Needs assessment on the user side is partially available; external consulting is necessary to assess user needs; user expectations known and defined; decision criteria 50% facts, 50% trust; benefit/risk assessment based on facts; habitual adaptation becomes apparent during implementation; involve citizens/interest groups (E4); increase attractiveness as an employer (E4);
(5) Functional objectives	harmless or Pollutant-reduced combined water drainage; better management of strong fluctuations; functional contribution to the combination of hydrological processes storage, infiltration, evapotranspiration, and delayed runoff; infrastructure protection during extreme weather events; influence on the expansion of existing structures in the combined sewer network, including pipes, basins and overflow structures; Avoiding discharge events and thus water pollution; increasing process and energy efficiency; determining weather forecasts; compensating for additional loads caused by extraneous water; controlling water quality (E2–4); increasing quality of well-being (E3–4); supporting volume activation in the sewer (E4); supporting water resource protection (E4);
(6) Benefit aspects	Ensure sustainability of municipality/organization; data for better decision-making basis; clarity through individual dashboards; creation and optimization of scenarios; prediction/forecasting function (supports future decisions; adaptation to climate change; adaptation to climate change; create robustness of systems; increase learning of users and third parties; take over sponge function of neighboring properties (E3–4); employee attractiveness (E4); learning object for further projects (E4);
(7) Variants	Volume-based control; expandability; rule-based control; weather-dependent control
(8) Basis of the decisions	small degree of uncertainty; committee decision; referendum (E4)
(9) Data basis	Local measurement data; forecast data from external sources (e.g. precipitation); scenario analysis historical or parallel; real inventory changes can be taken into account; experience values of operating personnel; data analysis of handheld devices, e.g. smartphones; soil mechanical characteristics (E3–4); data of installed soil material (E3–4);
(10) Risk analysis	Overflow and overflow; runoff damage potential; consider pollutant release from the roof (E2–4); overflow control to neighboring properties (E3–4); precedent (E4);
(11) Control methods	Real-time based; local; optimization-based control; predictive control; emergency operation for dynamic variant; digital precipitation data; experience of operating personnel
(12) Checking the plausibility of results and operational safety	Frequency and effort of verification (calibration); user training; integration of drone surveillance/monitoring; experience of operating personnel; data analysis of handheld devices, e.g. from smartphones; population surveys (E4)
(13) Business-as-usual potential	To be checked; in which planning phase; according to which criteria; funding potential; influence of employees on process; support, participation or active decision; comparison with B-plan (E4); comparison with drainage planning (E4); comparison with other infrastructure planning (E4); comparison with flood protection concept (E4)
(14) Network width	Cross-system
(15) User	Newcomer; small municipality; the high need for further training; role of end user; lack of experience with intended technologies; usage perspectives; usage context
(17) Organization	Organizational culture appropriate; organizational structure sufficient; participation provided for in the project
(18) Project-specific requirements	Cost situation school; everyday work of teachers:Inside; accessibility of measuring systems on green roof; use as a test facility at the University of Applied Sciences; integrability of PV control for PV/green roof combination; lighthouse character; recording de-/retention effect of green roof; analysis of moisture requirement and functional maintenance of green roof; analysis of moisture requirement for other planting; Analysis of plant life support; analysis of influence of green roof on microclimate; analysis of influence of planting on microclimate; determination of minimum equipment for basic operation; determination of equipment for innovation promotion; selection of own weather station or purchase of digital data; storage basin size (E3–4); calculation of extinguishing water reserve (E3–4).

Table 8.

Application testing Schauenstein—selection of individual criteria [26].

E2–4 = only meso-, macro-, and meta-level.

E3–4 = macro- and meta-level only.

E4 = meta-level only.

5. Conclusion

With the help of the Excel tool, all individual criteria can be evaluated at each of the four levels in terms of their general significance (rating) and their impact as a risk (inhibiting the digitalization project) and/or as an opportunity (promoting the digitalization project). Figure 10 uses the example of standard category No. 4 to show the possibility of selecting individual criteria and their individual evaluation. After the selection of criteria and their individual evaluation for all selected categories (according to Table 7), the evaluation overview (see Figure 11) can be used to check which categories are most likely to have a positive (green) or negative (red) influence on the selected digitalization project. The Excel tool also offers further result display options for visualizing the influences of the four levels or even each individual criterion on the selected digitization project (see Figure 12 as an example).

Figure 10.

Evaluation example of individual criteria of the 4th standard category at the micro level with the excel tool [26].

Figure 11.

Evaluation overview of several standard categories at the micro level with the excel tool [26].

Figure 12.

Evaluation overview of level influence with the excel tool [26].

The multi-level analysis presented here can close a key gap in the field of digitalization in water management, as it not only allows the necessary criteria for the added value of a digitalization solution to be defined more precisely. This method also makes it possible to compare different criteria with each other and recognize their connection. The lack of such a method has so far meant that many potential digitalization projects in the water industry have not been started or have had to be carried out to the dissatisfaction of involved persons. The practical tests carried out by Hof University of Applied Sciences with around 100 people showed that the basic system is accepted and is considered to be effective in the processing of digitization projects in water management [50]. In future projects, the multi-level analysis will be transferred to a web version so that the application can be used in a more user-friendly and flexible way.

Conflict of interest

The authors declare no conflict of interest.