2. Materials and Methods
This bibliometric study utilized Scopus databases as the primary source for data retrieval. Despite Scopus containing historical content dating back to 1788, the search was restricted to the period between 2000 and 2022. The research methodology encompasses a comprehensive approach involving data collection, preprocessing, and multifaceted analysis to unveil and comprehend scientific communities and their collaborative patterns, see
Figure 1.
In the initial phase, scholarly publications and their associated metadata were retrieved from the globally recognized Scopus database. The Scopus API was leveraged to conduct a targeted search focused on documents related to smart water meters, aligning with established research methodologies [
11]. The search query employed was as follows:
TITLE-ABS-KEY (smart* AND (water* OR flow*) AND meter*). All documents were downloaded from Scopus leading to the annotated search.
Crafted to cover a broad spectrum of literature, the query specifically addressed terms related to “smart water meter” or “smart flow meter” and their practical applications across various communities. However, the methodology introduced in this study effectively manages this potential issue, thanks to the resulting network graph’s structure and which can be visualized with the Gephi (open-source tool).
After acquiring the raw dataset, a preprocessing phase ensued, employing Microsoft Excel and OpenRefine for basic data organization and cleansing. This ensured that the dataset was primed for advanced analysis. For the core analysis, the open-source tool Gephi was employed, capitalizing on its robust capabilities in statistical analysis and data visualization. The data underwent refinement to eliminate documents lacking meaningful connections within the main body of scientific literature. This curation process allowed this research to focus specifically on the most pertinent documents and their relationships.
For the primary screening, the open-source tool Gephi was employed, taking advantage of its robust statistical analysis and data visualization capabilities. The dataset was subjected to a filtering process to eliminate documents that did not establish any connection with the rest of the documents. This filtering allowed us to focus the cluster analysis on the most relevant documents and their associations. For this purpose, we employed a community detection algorithm based on modularity in large networks, as detailed in [
12]. Specifically, Gephi’s algorithm used Louvain’s method, a widely accepted approach for community detection. Louvain’s method is optimized for large networks, with the goal of identifying modular structures by maximizing a modularity score within the range of −0.5 to 1. This score evaluates the density of edges within clusters compared to those between clusters. The Louvain method was chosen because of its computational efficiency and its ability to delineate clear structures in intricate datasets. Although alternative algorithms such as Girvan–Newman, Infomap, and Label Propagation were considered, they proved less suitable due to their computational intensity, limited relevance to citation networks, or production of less cohesive structures, respectively.
For the sorting by year, by subject category, and by country, these data are obtained directly from the Scopus search in the results analysis section of the website itself. It is possible to export them in CSV format. Later, year and subject category data were imported and represented using Excel. Country data have been mapped with QGIS software 3.28.
A comprehensive bibliometric analysis will be undertaken to delve into the evolution of publications over the years. This includes a meticulous examination of the countries and affiliations that make the most substantial contributions to this scientific field. Additionally, this study will focus on dissecting the key words employed in the articles, aiming to discern the scientific communities within which these works can be categorized. The overarching aim is to provide a comprehensive and insightful overview of the landscape of smart water meter research.
4. Worldwide Research Trends: Cluster Analysis
This bibliometric study has identified several clusters in the research landscape related to smart water meters, each characterized by its percentage weight, indicating its relative importance. The most substantial cluster is “Urban Water Meters” with a weight of 39.21%, underscoring a predominant focus on the application and development of smart water metering technologies in urban settings. The “IOT Connection” cluster follows closely with a weight of 36.39%, emphasizing the significant role of Internet of Things (IoT) connectivity in the context of water metering systems. “Communication and Security” constitute a noteworthy cluster, comprising 10.72% of the research focus, highlighting the importance of secure and efficient communication protocols in smart water meter networks. The “Grid Management” cluster, with a weight of 7.48%, suggests a substantial interest in the effective management and optimization of smart grids in water distribution systems. “Water Networks,” “Hot Water,” “Groundwater Monitoring,” and “Smart Irrigation” clusters contribute to the research landscape with weights ranging from 1.97% to 0.85%, indicating more specialized areas of study within the broader context of smart water metering. The distribution of these clusters reflects a diversified research landscape, encompassing urban applications, connectivity, communication protocols, grid management, and specialized topics such as hot water, groundwater monitoring, and smart irrigation. Overall, these clusters provide a comprehensive overview of the multifaceted aspects explored in smart water meter research, highlighting the interdisciplinary nature of the field and the varied priorities within the academic community.
Cluster 1, identified as Urban Water Meters, stands out as the most significant in terms of size.
Table 4 outlines the frequency (N) of occurrences of primary keywords associated with this cluster. It is noteworthy that this cluster leads in 18 out of the 20 main keywords, indicating a higher concentration of these terms within this community compared to others. In fact, eight of the 20 main keywords, such as Residential Water and Water Efficiency, are exclusive to this cluster.
Over the past two decades, more or less detailed measurement programs, referred to as smart measures, have been developed [
13]. This progress has led to nearly continuous monitoring of water consumption. The abundance of available data has prompted the exploration of strategies for modeling and managing water in residential areas [
14]. Periodic resource scarcity drives interest in limiting and controlling losses; leading studies toward water, notably by Rahim et al. [
15], demand planning, socio-economic analysis, behavioral analysis, classification of water-related events, and user feedback. The cluster’s connection with information technology is evident, sharing techniques with other supply systems such as electricity or gas [
16]. These meters enable the identification of the final use of water, a crucial aspect for making utility predictions [
17]. The technology has matured sufficiently to transition from pilot trials to real-world implementation in large supply networks [
18] and urban supply planning, closely intertwined with city planning itself [
19]. Additionally, the current analyses aim to disaggregate data to reach individual events [
20]. Reengineering of existing systems can be undertaken in light of the results already achieved [
21].
Cluster 2, denoted as IOT Connection, forms a community centered on Internet of Things (IoT) connection algorithms and network implementation. Ranking as the second-largest community,
Table 4 details the frequency (N) of primary keywords associated with this cluster, establishing connections with the Urban Water metering and Communication and Security communities, while maintaining ties with the Energy consumption cluster. Among its 20 main keywords, 12 are intrinsic to this cluster, and it holds significance as the leading community in 8 of them. Its primary associations lie with Clusters 1, 3, and 4, positioning it as a substantial and interconnected family with promising prospects for future development. Current meters are acknowledged for their sluggishness and time-intensive nature, promoting water wastage, thus advocating for the adoption of smart meters to enhance resource conservation and early fault detection [
22,
23]. Advanced Metering Infrastructure (AMI) architectures are anticipated to facilitate the integration of multiple networks simultaneously [
24], aiming for user interaction [
25]. The configuration of networks through smart water meters (SWM) and the advantageous use of IoT are highlighted in literature [
26]. Case studies explore innovative applications, such as using pumps as generators or direct interaction with users [
27], aligning with user interest in these devices [
28]. Efficient data transmission mechanisms within the network involve relaying data from each node to the nearest node [
29]. Two critical aspects, namely sensor network energy consumption and security issues, are underscored in the literature [
30], emphasizing the multifaceted nature of research within the IOT Connection cluster.
Cluster 3 was identified as Communication and Security and stands as the third-largest community.
Table 4 delineates the frequency (N) of principal keywords associated with this cluster, which revolves around communication and security aspects in network meters, particularly emphasizing energy meters. Remarkably, it encapsulates the top 10 keywords among the first 20 across all clusters, asserting its significance in this domain. In half of these 20 keywords, this community holds the utmost importance, while in the remaining 10, it shares keywords with other communities, primarily with Clusters 1 and 3. This community is predominantly concerned with communication protocols and applications of automatic meters, with a focus on reducing peak consumption, a growingly attractive topic [
31]. Traditional systems, effective thus far in electricity and water distribution, are deemed insecure and lacking user feedback capabilities [
32]. Smart energy meters can be leveraged to monitor other types of meters, integrated into networks [
33,
34]. The computerized control of these networks [
35] presents a challenge in terms of vulnerability to intrusions [
36], necessitating resolution in future developments [
37]. Data aggregation allows for processing while maintaining user privacy [
38]. The Communication and Security cluster, thus, contributes significantly to the exploration of secure and effective communication protocols in smart metering systems.
Cluster 4 was identified as Grid Management.
Table 5 provides an overview of the primary keywords associated with this cluster. Positioned as a peripheral cluster, it primarily addresses the challenges in managing vast amounts of grid data. Although it shares keywords with other communities, it assumes a leading role in what is termed Big Data, highlighting its auxiliary nature in supporting other fields. This specialized community focuses on the implementation of intelligent algorithms for information acquisition, data processing, and swift action in extensive supply networks, predominantly in energy but extending to other domains such as water or gas [
39,
40,
41]. The characterization of overall consumption holds a prominent place in this cluster, particularly in the context of water, leveraging time series [
42] or neural networks [
43]. Grid Management emerges as a crucial contributor to the application of intelligent algorithms for efficient data management and consumption characterization in large-scale supply networks.
Cluster 5, labelled as Water Network Partitioning, is outlined in
Table 5, featuring the primary keywords associated with this cluster. Characterized as an isolated cluster, it addresses the challenges of metering an extensive network through the analysis of smaller network components. Despite being a smaller cluster, it holds significance, representing 1.97% of the keywords. Within this group, all authors are closely interconnected, collaborating on a specific strategy for distribution network analysis. This strategy involves studying reduced sectors of the network to simplify the analysis [
44,
45,
46]. The methodology employed includes the installation of valves and smart meters on main lines, accompanied by the development of specialized software for these meters, referred to as SWAM [
47,
48]. This technique has been applied in fault detection and pollution control [
49] through recursive sectorization [
50]. The Water Network Partitioning cluster, although smaller in size, plays a vital role in the exploration of innovative strategies for network analysis, particularly in the context of simplifying analysis through the study of smaller network components.
Cluster 6 focuses on Water Heating and the associated metering of this valuable resource, as detailed in
Table 5 with the primary keywords for this cluster. Comprising 1.13% of the keywords, this well-defined community primarily interacts with Cluster 1. The majority of the keywords within this family are unique to it, although none stands out as particularly abundant. The substantial cost of this resource amplifies the significance of this family, as highlighted by Bacher et al. [
51]. The research within this cluster involves the development of prediction models for resource needs, particularly emphasizing scenarios in cold climates [
52]. Detection of leaks is of particular importance due to the high energy costs [
53]. This family extends its exploration to related networks such as domestic heat and gas, demonstrating a comprehensive approach to resource management [
54,
55]. The Water Heating and metering cluster, while relatively small, delves into crucial aspects of resource prediction, environmental considerations, and the efficient management of expensive resources.
The remaining clusters are of limited size but represent emerging sectors with promising prospects for future studies. Cluster 7 is a small cluster centered on Groundwater Monitoring, utilizing various approaches. It occupies a peripheral position in the diagram and is related to Cluster 1, accounting for 0.99% of the keywords. Similar to Cluster 8, the majority of the keywords within this group are specific to it, sharing few words with other clusters.
Table 6 lists the most representative keywords for this cluster.
Water crisis and the threat of resource shortage are identified as direct consequences of factors such as limited precipitation, traditional irrigation methods, and inefficient monitoring and control systems in agriculture [
56]. This cluster serves as a unique case within monitoring, often supported by simulation models [
57,
58,
59]. It places special emphasis on control elements and leverages other components, such as energy meters, to enhance information completeness [
60]. Additionally, Cluster 7 connects with the use of unconventional water sources, such as desalinated water for urban use and reclaimed water for agricultural purposes [
61].
Cluster 8, also a small cluster, is dedicated to Smart Irrigation control, representing 0.85% of the keywords. This group features highly specific keywords, occasionally shared with Cluster 1 and Cluster 3.
Table 6 displays the most representative keywords for this cluster.
Focused on intelligent control of irrigation systems to promote water savings and efficiency [
62], this cluster emphasizes user satisfaction as a crucial parameter [
63]. Initially applied in residential landscaping [
62,
63], smart irrigation is currently being implemented in agricultural areas [
64,
65,
66]. This family exhibits significant development potential, with numerous sensors of this type being deployed in agricultural areas [
67].
In the analysis, two clusters with minimal representation emerge—one focused on data acquisition, sensors, and pattern recognition of consumption. For instance, Grigoras et al. [
68] present an innovative solution in the form of a software platform for sustainable water supply system management. This platform incorporates advanced ICT solutions, including Blockchain and Artificial Intelligence, along with smart concepts such as smart metering and demand response, offering benefits to both the energy and water sectors.
Additionally, there is another cluster specialized in automated decision making, utilizing learning programs, decision support programs, etc. [
69,
70].
6. Conclusions
In conclusion, the analysis of smart water metering research based on Scopus data reveals a dynamic landscape with significant growth in publications over the years. The evolution of articles in the “smart water meter” topic indicates a rising interest and emphasis on this field, with a substantial increase in publications from 2000 to 2023. The distribution of articles across scientific categories highlights the multidisciplinary nature of smart water meter research, with a predominant focus on engineering, computer science, and energy.
Examining the geographic distribution of articles underscores global participation, with the United States, India, and China leading in the number of publications. This international collaboration contributes to a diverse and comprehensive understanding of smart water metering technologies and applications. The analysis of affiliations sheds light on key institutions involved in smart water meter research, with Griffith University, Politecnico di Milano, and the University of Salerno among the leading contributors. These institutions play a pivotal role in advancing knowledge and innovation in the field.
Exploring keyword frequencies reveals the prevalent themes in smart water meter research, with a strong emphasis on topics like Smart Metering, Water Demand Management, Smart Grid, and Internet of Things (IoT). Clustering analysis further categorizes research areas, showcasing specialized communities such as Urban Water Meters, IOT Connection, Communication and Security, and others.
Challenges and prospects in the field include the need for further exploration of emerging clusters, integration of diverse research areas, and addressing critical issues like communication protocols, security vulnerabilities, and efficient data management in smart water networks. On a positive note, the increasing focus on smart irrigation control, advancements in network partitioning, and exploration of unconventional water sources present promising avenues for sustainable water management. Overall, the research landscape in smart water metering is vibrant, collaborative, and evolving, with researchers and institutions worldwide contributing to advancements in technology, data analytics, and strategies for efficient water resource management.