AI-Driven Bibliometric Analysis of Bacterial Concrete Research (2020–2025)

Abstract

This investigation examines the novel application of bacterial concrete as a sustainable substitute for traditional concrete within the construction sector, utilizing bibliometric analysis in conjunction with machine learning. The main aim of the study is to gain insights into the application and potential benefits of using bio-based concrete in the construction industry. A comprehensive search of all publications indexed in Scopus was carried out for the period spanning from 2020 to 14 March 2025, followed by meticulous screening and extraction of relevant documents. The dataset obtained from Scopus was processed in strict accordance with the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines to uphold transparency and replicability throughout the systematic review process. A descriptive analysis was undertaken to evaluate publication trends over time. The research on bio-concrete combined with machine learning is highly concentrated in Asia, Europe, and the USA; in contrast, vast areas of Africa show no research output regarding self-healing concrete based on this data extraction. Various types of bacteria, including Bacillus species, are explored for their calcium carbonate precipitation capabilities in this review. Microbial-induced calcite precipitation process reduces carbon emissions associated with cement production and extends concrete lifespan by sealing cracks.

1. Introduction

Concrete has conventionally been favored as a material within the construction sector due to its notable strength, durability, and versatility. Its cost-effectiveness and compatibility with steel reinforcement render it essential in the advancement of contemporary infrastructure [1,2]. However, conventional concrete is criticized for its environmental impact, tendency to degrade, and low tensile strength, leading to cracking [3,4].

Sustainable construction practices emphasize minimizing environmental impact while ensuring social and economic outcomes for both present and future generations [5,6]. The construction sector is one of the largest contributors to global carbon dioxide emissions and energy consumption [7]. Cement production is a fundamental component of concrete and significantly contributes to CO₂ emissions [8,9]. Additionally, the extraction of raw materials leads to habitat destruction and resource depletion [10]. Sustainable concrete structures aim to mitigate the negative impacts by adopting a life-cycle approach that reduces environmental burdens while maintaining structural integrity [11]. As a result, sustainability in concrete has become a major focus in the construction industry to ensure long-term viability.

Awareness of alternative cementitious materials is a very important issue to enhance concrete’s sustainability [9,12]. Performance-based approaches are being explored to improve durability and service life predictions [13,14]. To address sustainable and environmental challenges, various techniques have been developed to enhance concrete’s strength and durability. One promising approach is microbial-induced calcite precipitation (MICP), which leverages urease-producing bacteria to enhance concrete properties. Bio-based bacterial concrete offers an eco-friendly alternative to traditional concrete by utilizing bacterial self-healing mechanisms to generate calcium carbonate, effectively sealing cracks and improving durability while reducing maintenance cost and frequency [15,16].

Supplementary cementitious materials, such as fly ash and slag, are being incorporated into concrete to reduce CO₂ emissions [17,18]. Additionally, concrete recycling and reuse strategies help conserve resources and minimize waste generation [19,20]. These sustainable initiatives align with green building certification programs such as LEED and BREEAM, which promote eco-friendly construction [21]. Corrosion of reinforcement steel is another major issue affecting concrete structures [22]. Traditional treatment methods often pose environmental and health risks and provide only short-term solutions. In contrast, microbial self-healing techniques offer a long-lasting, environmentally friendly, and efficient method for crack repair, surpassing conventional approaches in durability and compatibility with concrete composition [1].

The self-healing capability of concrete is an emerging area of research, with microbial-induced calcium carbonate precipitation receiving significant attention for its potential applications in building structures [23]. Current bio-cementation techniques focus on microbially induced carbonate precipitation (MICP) and enzyme-induced carbonate precipitation (EICP) to enhance durability and eliminate cracks [24,25]. Self-healing concrete has the potential to extend infrastructure lifespan, reduce maintenance costs, and promote sustainability. It can also restore its original properties after cracking, behaving as if it had never sustained damage [26,27,28].

Cracks in concrete structures occur due to temperature variations, structural stress, and freeze–thaw cycles, compromising durability [29,30]. However, bacterial activity can seal these cracks, preventing further expansion and reducing repair and maintenance costs [31,32]. Bacterial self-healing capabilities enhance structural durability and sustainability while minimizing maintenance expenses [29,33]. Despite these advantages, challenges such as thermal expansion effects and potential environmental concerns persist, necessitating further research [34,35].

This review paper contributes to the existing body of knowledge by exploring the concept of bio-based bacterial concrete and its fundamental mechanisms. Additionally, it serves as a valuable tool for policy support by offering data-driven insights that guide decision-making in research funding, regulations, and strategic planning.

Policymakers can use bibliometric and machine learning findings to align research priorities with societal and economic needs. For researchers seeking funding, bibliometric analysis strengthens grant proposals by demonstrating the significance of their work within the global research landscape. It highlights impact metrics, emerging trends, and potential areas for innovation, increasing the chances of financial support. Furthermore, bibliometric studies facilitate collaboration by mapping networks of researchers, institutions, and countries working on similar topics. This enhances opportunities for partnerships, knowledge exchange, and interdisciplinary research, ultimately driving scientific progress. The key questions aim to address the bio-concrete application and sustainability of concrete in construction. These questions guide the review by focusing on critical aspects that influence the feasibility, application, and challenges of using bacteria in concrete production.

▪

How is the distribution of self-healing concrete research currently conducted and applied in the construction industry?

▪

How are bibliometric analysis and machine learning combined to show the integrity of self-healing concrete research?

▪

What is the performance of self-healing concrete in terms of durability and mechanical properties of concrete?

▪

Performance of self-healing concrete.

▪

What are the current practices, regulatory and standardization considerations, research gaps, and future directions of bacterial concrete?

2. Methodology

Data acquisition from the existing literature was significantly important in the research, especially for the conclusions drawn from the Scientometric analysis [36]. The literature collection strategy involved searching all Scopus publications by relevance and from 2020 to 14 March 2025, and the papers were screened by reviewing. The document search was performed using the search filters “Search in article title, abstract, keywords” and “Search documents,” with the descriptors “AND” and “OR” [37].

2.1. Search Database

The first step is selecting and searching relevant bibliographic databases to gather research articles. Scopus database is used for this research work. A well-defined search strategy is essential to retrieving high-quality research articles. This involves identifying relevant keywords and Boolean operators (AND, OR, NOT) to refine search queries. Searches are usually conducted within the title, abstract, and keywords fields to enhance precision. Additionally, search filters such as publication year, document type, and language restrictions help refine results further. Once relevant articles are identified, citation data is exported in formats of CSV and RIS for further processing and analysis.

The literature on self-healing concrete in Scopus was retrieved by searching keywords within the publications’ title/abstract/keywords. Based on the research question of this review, the selected keywords were TITLE-ABS-KEY ((Bio-Concrete OR Self-healing Concrete) AND (Sustainab* OR Application) was taken in the advanced search. In conducting a systematic review, selecting the Scopus database is crucial to ensuring comprehensive and high-quality literature coverage.

Scopus is wider Coverage and multidisciplinary which is one of the largest abstracts and citation databases, covering a broad range of disciplines, including engineering, environmental sciences, material sciences, and construction research. Unlike PubMed and web of science which focuses primarily on medical, life sciences, and more selective indexing policy, respectively, Scopus offers more extensive and interdisciplinary coverage suitable for research in sustainable construction, materials, and Engineering applications.

Higher Number of Indexed Journals are published in Scopus peer-reviewed journals relevant to engineering, materials science, and sustainability. It has a more comprehensive citation and allows better citation-based search refinement. User-Friendly Interface and Advanced Search Features and enabling precise search filtering by keywords, subject area, author, affiliation, funding sources, and publication year.

2.2. Screening Process

The screening process involves filtering and refining the collected dataset to ensure accuracy and relevance. This step begins with the definition of inclusion and exclusion criteria. Typically, the analysis focuses on peer-reviewed journal articles, reviews, and conference papers, while excluding duplicates, non-English documents, or irrelevant studies that do not align with the research objectives. The final dataset is cleaned and structured, ensuring that only high-quality research articles proceed to the next stage.

It entails systematic filtration and refinement of the compiled dataset to guarantee accuracy and relevance. This phase commences with the establishment of precise inclusion and exclusion criteria. Typically, the analytical focus is directed towards peer-reviewed journal articles, review papers, and conference proceedings, while excluding duplicates, non-English documents, and studies that are incongruent with the research objectives. Subsequent to this step, the dataset undergoes a cleaning and organization process to ensure the retention of only high-quality research articles for further analysis. In this investigation, certain document types are deemed non-pertinent and are thus omitted from consideration during the screening phase. Initially, any publications lacking a direct association with the principal research topics, such as self-healing concrete, bacterial-based healing mechanisms, or the integration of machine learning within this context, are eliminated. Furthermore, studies pertaining to disparate disciplines, including medicine, agriculture, microbiology without direct application to cementitious materials, or general concrete research devoid of a healing component, are similarly designated as irrelevant.

Furthermore, to uphold methodological rigor, materials that have not undergone peer-review, such as blogs, news articles, informal online content, editorials, letters to the editor, or conference summaries lacking full papers, are systematically excluded. To prevent redundancy, duplicate records manifesting as the same study appearing across multiple indexing formats or as an initial conference version preceding a subsequent journal article are meticulously removed. Additionally, publications in languages other than English are excluded when the research is exclusively focused on English-language sources, thereby ensuring consistency in the analytical process.

2.3. Data Extraction and Analysis

Data extraction involves collecting key bibliometric indicators and metadata from the selected documents. The essential fields extracted include the title, authors, year of publication, journal name, keywords, abstract, number of citations, author affiliations, funding sources, and research domains. These fields provide valuable insights into publication trends, influential authors, and research impact. Bibliometric software tools, such as VOS viewer, for data processing and visualization were applied. The extracted data serves as the foundation for further analysis, revealing collaboration networks, citation structures, and thematic trends in the research field.

The analysis phase applies various bibliometric indicators and visualization techniques to understand research trends and networks. A descriptive analysis is conducted to examine publication trends over time, the most productive authors, institutions, and countries, as well as the most cited articles and journals. Additionally, network analysis explores relationships between different elements of the research landscape. Key network analyses include co-authorship analysis (examining collaborations among researchers), co-citation analysis (identifying frequently cited papers together), and keyword co-occurrence analysis.

2.4. Reporting and Interpretation

The final step is to communicate the findings through structured reporting. A summary of key findings highlighting research gaps, emerging topics, and influential works in the field is reported. The results are typically presented with graphical representations, such as trend graphs, co-citation maps, and bibliometric networks, making the findings more accessible and interpretable. The reporting process also involves preparing a manuscript that aligns with journal guidelines, ensuring clarity and coherence. Finally, a discussion of the implications of the findings and future research directions, which strengthens the contribution of the bibliometric study to the academic community, is presented. The whole structure of the bibliometric analysis methods for this research is summarized below in Figure 1.

2.5. Limitations

Bibliometric analysis of self-healing concrete faces limitations, including database bias due to a lack of coverage in Web of Science, leading to incomplete global research representation. Inconsistent author keywords and citation biases favoring older articles distort the analysis. Variability in software settings affects reproducibility, and reliance on metadata overlooks technical details in full texts.

Machine learning in self-healing concrete research faces challenges, like small, varied datasets that risk overfitting. Models often do not generalize due to specific experimental settings, and the lack of standardized features hinders comparability. Complex algorithms struggle with interpretability, and biases in training data can skew results. Machine learning’s dependency on pre-processing and tuning makes it sensitive, and the absence of experimental validation limits practical use.

Citation-based indicators have limitations, like citation age bias, disciplinary citation differences, and self-citations, which can distort research impact and country rankings. Additionally, bacterial concrete research’s interdisciplinary nature might not be fully captured by tools like VOS viewer, which are sensitive to settings, thresholds, and data density [38].

3. Results of Bibliometric Analysis

3.1. Geographical Distribution and Key Studies

The analysis identifies 119 documents that meet the specified criteria. At the core of the visualization is a corresponding line graph illustrating the publication count per year from 2020 to 2025. The provided figures present a compelling overview of document publication trends, analyzed through various lenses. The bar graph of documents by year charts the progression of publications from 2020 to 2025, illustrating year-to-year fluctuations in output. Documents by affiliation offer insight into the contributions of different institutions, including those with distinct regional presences, such as Universiteit Gent, Politecnica di Milano, Delft University of Technology, and Shenzhen University, highlighting the diverse geographical distribution of contributing organizations.

Documents by Country/Territory” provides a clear depiction of region-based document output. This graph allows for direct comparison of contributions from countries across different continents, including significant representation from Asia (China, India, Malaysia, Saudi Arabia), Europe (United Kingdom, Belgium, Netherlands, Italy), and North America (United States), alongside contributions from Australia. Collectively, Figure 2 offers a multifaceted analysis of document publication trends, examining them by year, affiliation, author, and with a specific emphasis on the geographical distribution of publications.

The second set of figures presents a metadata-driven analysis of the documents. The “documents by type” graph classifies documents according to their respective types and provides a comparative visualization of the relative frequency and distribution of each document type within the dataset. The documents by subject area graph classifies the documents by subject area, enabling the analysis of the representation of different subject areas. The documents by funding sponsor graph highlights the financial support behind the research, showcasing contributions from entities such as the European Commission, Ministry of Science, Horizon 2020 Framework Programme, UK Research and Innovation, and National Key Research and Development Program. Specific programs like Horizon 2020, H2020 Excellent Science, and H2020 Marie Skłodowska-Curie Actions are also referenced in Figure 3.

Figure 4 presents a world map illustrating the regional distribution of research documents, using a color gradient to represent the number of documents per country. The map employs a dark red scale, where lighter shades indicate a lower volume of documents, while deep red shades signify a higher concentration of documents. This visual representation effectively highlights the geographical disparities in document production, emphasizing the dominance of certain regions within the dataset. Additionally, a legend labeled “Series 0 to 48” clarifies that the color gradient corresponds to the number of documents per country within this range, enhancing the interpretability of the data.

China and the USA appear in the deep red shade, indicating the highest number of research documents. Other regions, such as India, Australia, and parts of South America, also show significant concentrations. In contrast, vast areas of Africa reflect no research output regarding self-healing concrete.

The findings presented in Figure 4 offer valuable insights into various stakeholders and institutions. Researchers can use these data to identify publication trends, leading institutions, and key subject areas within their field of interest. Academic institutions can assess their research output in comparison to global trends, identifying strengths and potential areas for improvement. Policymakers can leverage these insights to analyze research patterns and make informed decisions regarding funding allocation, institutional support, and strategic research priorities.

Table 1. Citation count and rate (%) of occurrence of documents per country.

S, No	Countries	Occurrence	Rate (%)	Citations	Total Link Strength
1	Belgium	15	4.79	529	39
2	Italy	16	5.11	472	39
3	United Kingdom	29	9.27	469	35
4	Malaysia	12	3.83	426	30
5	Poland	8	2.56	157	25
6	Lativa	2	0.64	88	19
7	Turkey	6	1.92	116	19
8	Spain	7	2.24	183	18
9	China	48	15.34	709	15
10	Saudi Arabia	12	3.83	291	14
11	United States	25	7.99	627	14
12	Bangladesh	2	0.64	92	13
13	Greece	2	0.64	50	13
14	Canada	6	1.92	66	12
15	Germany	7	2.24	191	12
16	Australia	11	3.51	208	11
17	Netherland	13	4.15	178	8
18	South Korea	5	1.60	93	6
19	Iraq	7	2.24	195	5
20	Lithuania	2	0.64	39	5
21	Russia	4	1.28	107	5
22	Egypt	8	2.56	205	4
23	India	19	6.07	298	4
24	Philippines	2	0.64	67	4
25	Brazil	3	0.96	9	2
26	Ireland	3	0.96	95	2
27	Pakistan	5	1.60	109	2
28	Czech Republic	3	0.96	32	1
29	Cyprus	3	0.96	60	0
30	Denmark	2	0.64	50	0
31	Iran	4	1.28	84	0
32	Japan	5	1.60	53	0
33	Serbia	4	1.28	99	0
34	Singapore	2	0.64	58	0
35	Sweden	4	1.28	165	0
36	Switzerland	3	0.96	175	0
37	Taiwan	2	0.64	20	0
38	Thailand	2	0.64	154	0

offers a comprehensive scientific overview of various countries’ contributions to research, likely in terms of their output and influence within the global scientific community.

Each row is indicative of a specific country, featuring a set of metrics intended to quantify its scientific engagement and impact. Among these metrics, ‘Occurrence’ appears to denote the frequency or volume of scientific output from a country, including publications or research initiatives. Complementary to “Occurrence,” the “Rate (%)” provides a proportional representation of each nation’s contribution, delineating its share of the total scientific activity documented across all countries included in the dataset. This percentage offers essential insights into the relative significance of each country’s scientific undertakings within the framework of the dataset. An important indicator of scientific impact, “Citations,” counts the number of times a country’s scientific works have been referenced by other researchers. A higher citation count typically denotes increased influence, relevance, and recognition within academic discourse.

Additionally, although not explicitly defined in

Table 2. Top affiliations with the highest number of documents and citations per country.

Top Affiliations with Higher Number of Documents and Citations	Countries
Built Environment and Sustainable Technologies (BEST) Research Institute, Liverpool, United Kingdom	United Kingdom
Built Environment and Sustainable Technologies Research Institute, Liverpool John Moors University, United Kingdom
Department of Civil Engineering and Industrial Design, The University of Liverpool, Liverpool, United Kingdom
Department of Civil Engineering, University of Birmingham, Birmingham, United Kingdom
Department of Engineering, University of Cambridge, United Kingdom
Laboratory for Track Engineering and Operations for Future Uncertainties, University of Birmingham, Birmingham, United Kingdom
School of Engineering, Cardiff University, Cardiff, United Kingdom
Centre for Infrastructure Engineering and Safety, Australia	Australia
School of Civil and Environmental Engineering, University of Technology Sydney, Australia
School of Civil and Mechanical Engineering, Curtin University, Australia
Batir, University Libre de Bruxelles, Belgium	Belgium
Department of Mechanics of Materials and Constructions, Vrije Universiteit Brussel, Belgium
Department of Structural Engineering and Building Materials, Belgium
School of Highway, Changan University, China	China
School of Traffic and Environment, Shenzhen Institute of Information Technology, Shenzhen, China
State Key Laboratory of Silicate Materials for Architecture, Wuhan University of Technology, Wuhan, China
Department of Civil and Environmental Engineering, Polytechnic Di Milano, Milan, Italy	Italy
Polytechnic Di Milano, Department of Civil and Environmental Engineering, Italy
Civil Engineering and Geosciences, Delft University of Technology, The Netherlands	The Netherland
Faculty of Civil Engineering and Geosciences, Delft University of Technology, Delft, The Netherlands

, “Total link strength” likely represents a measure of the breadth and robustness of scientific collaborations or connections associated with each country. This may encompass various forms of linkage, including co-authorship networks, collaborative research initiatives, or the distribution of research funding. Collectively, these metrics offer a nuanced assessment of countries’ roles within the global scientific framework, extending beyond mere publication counts to encompass the broader impact and interconnectedness of their research endeavors. Such data are pivotal to the disciplines of bibliometrics and scientometrics, providing a quantitative foundation for the analysis and comprehension of scientific output. Researchers and policymakers can employ this information to assess national research performance, discern emerging scientific trends, and inform strategic decision-making regarding research funding and international scientific collaborations. Furthermore, it aids in the examination of collaborative research activities and the tracking of scientific progress over time, serving as an indispensable tool for apprehending the dynamics of global science.

The depicted visualization illustrates countries with a substantial number of institutions. The image presents a network visualization, presumably generated by VOS viewer, which demonstrates the patterns of collaboration or relationships between countries as determined by scientific output or institutional affiliations. This form of visualization, commonly employed in scientometrics, features nodes (circles) that represent countries and lines (edges) that connect them, thus indicating the strength or frequency of collaborations. Node size: Each country’s node size is typically proportional to a specific quantitative metric, which, in this case, likely denotes the number of institutions from that country participating in the studies or the volume of their scientific output. Larger nodes, such as those representing China, the United States, the United Kingdom, and India, imply a greater number of contributing institutions or a higher research volume originating from these nations.

Connecting lines (edges) on the map represent collaborations or co-authorships between countries. The thickness of these lines typically denotes the magnitude or regularity of these collaborative endeavors. For instance, more substantial lines, such as those between the United States and the United Kingdom or China and Pakistan, imply a higher frequency of joint research activities. Color-coding: The nodes are assigned colors ranging from purple/blue to green/yellow, accompanied by a legend reflecting the timeline from 2021 to 2023. This implies that the coloration of a node signifies either the mean publication year of the research associated with that nation or the recency of its notable collaborations. Nodes with darker hues (purple/blue) denote earlier activities within the specified interval, whereas those with lighter hues (green/yellow) denote more recent activities. For instance, nations such as Pakistan and Egypt exhibit more yellow tones, signifying relatively recent contributions within the period visualized.

Clustering: Countries that often work together or have similar research interests tend to form clusters. Noteworthy clusters include a prominent Western group that consists of the United States, the United Kingdom, the Netherlands, Belgium, and Italy, as well as another cluster featuring China, India, Pakistan, and Egypt. In summary, the figure provides a visual representation of a network of countries. The size of each country’s circle signifies either the number of participating institutions or the volume of research produced, while the connecting lines depict collaboration ties. The colors may represent the recency of scientific contributions within the time frame from mid-2021 to 2023, as seen in Figure 5.

The image presents a visualization derived from VOS viewer, a software instrument employed for the construction and visualization of bibliometric networks. These networks typically depict relationships among keywords, authors, or publications, predicated on patterns of co-occurrence or citation as shown below in Figure 6. In the present image, a network of keywords is likely depicted. Each circle presumably represents a keyword, with the lines connecting them suggesting their co-occurrence within the analyzed documents. The dimensions of the circles may correspond to the frequency or significance of the keywords, whereas the colors may denote clusters of interrelated terms.

The co-occurrence network of keywords is recommended for analyzing research trends, knowledge structures, and thematic evolution in bacterial or self-healing concrete. By mapping terms like self-healing concrete, bacteriology, calcium carbonate, microcapsule, bacteria, hydrogel, crack closure, healing process, healing agent, biomineralization, crack width, chlorine compounds, encapsulation, and sulfur compounds, the network shows conceptual links and development over time. This analysis identifies dominant clusters, emerging themes, and interdisciplinary connections. The co-occurrence structure reveals key knowledge areas in bacterial concrete research, like microbial-induced calcite precipitation, material encapsulation, and advanced healing agents. Bibliometric visualization shows a shift from basic bacteriological methods to advanced approaches using hydrogels and microcapsules. Thus, a keyword co-occurrence network deepens bibliometric insights and enhances understanding of bacterial self-healing concrete’s evolution and alignment with trends like machine learning–assisted material optimization.

Self-healing concrete is a growing focus in sustainable construction research, with increased publications since 2020 due to interest in durable, low-maintenance infrastructure. It connects biology, healing mechanisms, and material innovations, highlighting its interdisciplinary nature. Network analysis shows it as a central theme linking engineering, microbiology, and materials science. Bacteriology, particularly microbiologically induced calcium carbonate precipitation (MICP), is crucial in connecting biological sciences to construction engineering, marking a shift towards bio-based healing, especially in biotech-strong countries. Co-citation analysis confirms bacteriology’s pivotal role in bridging microbial studies and civil engineering.

Calcium carbonate is vital in microbial self-healing for crack repair via CaCO₃ precipitation, linked to structural performance, microstructural analysis, and durability. Microcapsule research, focusing on healing agents, encapsulation, and crack closure, emphasizes capsule design, release mechanisms, survivability, and cost-effectiveness. Publication trends show growth as technologies become more durable, responsive, and scalable. Biomineralization, a key biological mechanism, intersects with bacteriology and calcium carbonate, highlighting interest in MICP pathways, enzymatic reactions, and mineral formation. Bibliometric trends show expansion into genetic engineering, biofilm-mediated healing, and hybrid biomineral systems.

Bacteria, particularly strains like Bacillus sphaericus, Bacillus pasteurii, and ureolytic bacteria, intersect with bacteriology and biomineralization. Their frequent mention indicates interest in bacterial screening, metabolic optimization, and survival in challenging cement environments. Bibliometric coupling reveals collaboration between microbiologists and civil engineers on MICP-based self-healing. Hydrogel research, focusing on carriers, water reservoirs, and bio-retention systems, is increasing, and it is crucial for moisture retention and healing in dry conditions. Their association with terms like bacteria, healing agent, and crack closure highlights their multifunctionality, marking a shift towards hybrid biological and polymer-based healing technologies. Crack closure is crucial for evaluating experimental studies, strongly linked to crack width, healing, and calcium carbonate formation. It Is essential for quantifying healing efficiency using metrics like closure percentage, mechanical recovery, and durability. Research focuses on real-time monitoring and standardized methods, shown by increased citations. Crack width is key to healing efficiency, affecting bacterial activation, microcapsule rupture, and CaCO₃ deposition. In keyword clusters, it appears with crack closure and the healing process, indicating applicability limits. Bibliometric studies highlight the focus on maximum healable crack size, impacting material design and codes.

The healing process encompasses physical, chemical, and biological activities for repairing cracks, linked to mechanisms like healing agents and encapsulation. Bibliometric data shows a focus on mechanistic studies using microscopy and modeling to understand healing kinetics. Its central role in thematic maps marks a shift from conceptual to application-focused research. Healing agents connect microcapsules and self-healing systems, appearing in networks with materials like epoxy and bacterial spores, highlighting their importance in biological and chemical methods. Trends indicate growing interest in sustainable, responsive healing agents. Chlorine compounds in bibliometric networks highlight interest in chloride exposure’s impact on healing, bacteria, and calcium carbonate stability, especially in marine, de-icing, and corrosion contexts. Sulfur compounds, linked to microbial metabolism and environmental resilience, show interest in non-ureolytic bacteria for mineral precipitation via sulfate. Trends suggest exploring healing chemistries to address ureolysis’s ammonia release and environmental issues. Encapsulation is key in technology, connecting microcapsules, healing agents, and carriers. Bibliometric evidence shows strong growth, focusing on polymeric capsules, ceramic shells, hydrogels, and inorganic carriers. It is crucial for enhancing agent protection, controlled release, and long-term functionality in self-healing concrete systems.

3.2. Prisma Protocol

The dataset exported from Scopus was processed in accordance with the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines to ensure full transparency and replicability of the systematic review process. Initially, 1119 records were retrieved from Scopus. After applying a temporal filter (2020–14 March 2025), the number of eligible records was reduced to 780, indicating that 339 documents fell outside the specified timeframe. To further refine the scope, a subject-area filter limited to Engineering, Materials Science, and Chemical Engineering was subsequently applied. This step reduced the dataset to 202 documents that are directly pertinent to the core research fields. This final number is consistent with the progressive narrowing of the initial 1119 records to 780 time-restricted records and, subsequently, to a more focused subset of 202 subject-specific documents.

Subsequently, the document type was restricted to Articles and Reviews. A correction is necessary at this stage: the original description indicates an increase from 202 to 578 records, which is logically inconsistent. The corrected interpretation is as follows: From the 780 records remaining after application of the time filter, limiting the document type to Articles and Reviews yielded 578 records. Thereafter, application of the subject-area filter to these 578 records resulted in a final set of 202 subject-relevant articles. Accordingly, the corrected sequence of records is Records retrieved (1119), After time filter (780), After document-type filter (Articles + Reviews) (578), and After subject-area filter (202).

Following this, only English-language documents were retained, reducing the eligible set. A further filter selecting only open-access records narrowed the dataset to 119 documents. During the screening phase, titles, abstracts, and full texts were evaluated using Covidence. After removing duplicates, non-relevant articles, and studies failing to meet inclusion criteria, 54 documents were finally selected for the systematic review.

Generally, the Relationship Between Sample Sizes Across Stages in the PRISMA protocol constitutes a progressively restrictive selection procedure, aimed at preserving the relevance, quality, and methodological rigor of the final corpus:

• Time filter (1119 → 780): Excluded older publications to ensure that the review captures the most recent developments in the field.
• Document-type filter (780 → 578): Restricted the dataset to peer-reviewed scholarly outputs (articles and reviews), thereby enhancing the robustness and credibility of the evidence base.
• Subject-area filter (578 → 202): Delimited the corpus to publications within research domains directly related to bacterial concrete, ensuring precise thematic alignment with the review’s objectives.
• Language filter (202 → reduced set): Retained only English-language publications to maintain linguistic consistency and enhance comparability in the interpretation and synthesis of findings.
• Open-access filter (English-language set → 119): Limited the dataset to open-access articles, promoting transparency and unrestricted verification of the full texts.
• Screening (119 → 54): Further excluded duplicates, thematically irrelevant records, and studies exhibiting inadequate methodological rigor, based on detailed full-text appraisal. This stepwise reduction procedure reflects the central tenet of the PRISMA framework: at each stage, studies that do not meet predefined eligibility criteria are systematically removed, culminating in a final analytical corpus of 54 methodologically sound and substantively pertinent documents for comprehensive systematic evaluation. See Figure 7 below.

4. Discussion on Systematic Review

4.1. Machine Learning Techniques Used in Predicting Self-Healing Concrete Properties

Self-healing concrete represents a rapidly advancing material that enhances the sustainability and durability of concrete structures through the autonomous rectification of cracks. For the practical application of self-healing concrete, accurate performance prediction is crucial. Nevertheless, traditional analytical models encounter difficulties due to the complexity and variability of influencing factors, such as crack width, type of healing agent, environmental exposure, and curing duration. In order to simulate these complex interactions, machine learning (ML) techniques have proven exceedingly effective. They offer precise predictive capabilities in addition to significant insights into the behavior of self-healing concrete (SHC) under various conditions [23]. Concrete exhibits self-healing capabilities that facilitate the sealing of microcracks contingent upon diverse testing conditions and proportions of the concrete mix. Self-healing concrete is categorized into two principal types: autogenous healing concrete and agent-based healing concrete. Autogenous healing refers to the innate capability to heal cracks based on the components inherent within the concrete matrix. Consequently, autogenous healing concrete encompasses both intrinsic healing concrete and enhanced autogenous healing concrete. Conversely, agent-based healing concrete involves the employment of healing agents, such as polymers or bacteria, to effectuate the healing process [39].

Regression models like linear regression, support vector regression, and decision tree regression are early machine learning techniques used to predict self-healing concrete properties. Favored for their interpretability and ability to represent linear and some non-linear relationships, these models use inputs like healing time, agent dosage, and crack width to estimate compressive strength recovery and crack closure rates. Accurately predicting concrete’s self-healing, including post-fire recovery, is crucial to reduce costly destructive testing and assess intelligent cementitious materials’ performance. However, there are few attempts to model post-fire auto-repair of concrete [40].

The first step in model generation is detailed study and selection of influential input parameters affecting self-healing concrete output. Experimental data for prediction includes mineral content of FA, SF, LP, CWB, and crack thickness prior to self-healing, representing the influential variable. Post-healing concrete crack thickness can be predicted using these variables [41].

CWA = f (FA, SF, LP, and CWB)

(1)

It explains the process involved in creating a predictive model for self-healing concrete. The first step is to carefully examine various variables associated with the material. This detailed study helps in identifying input parameters that significantly influence the self-healing capability of the concrete. Specific experimental elements, such as the mineral content of fly ash (FA), silica fume (SF), limestone powder (LP), and cement waste byproducts (CWB), are measured. In addition, the initial crack thickness prior to any self-healing process is considered a key influential variable. The resulting model aims to predict how the thickness of concrete cracks changes and reduces after undergoing the self-healing process by considering these identified parameters.

The artificial neural network (ANN), alternatively referred to as a neural network, has its origins in the emulation of biological neural networks. Typically, it comprises numerous neurons arranged in layers, which include one input layer, multiple hidden layers, and an output layer. The neurons are thoroughly interconnected between adjacent layers via weighted connections, and, customarily, interconnections between neurons within the same layer are absent [42,43].

Artificial Neural Networks exhibit significant potential in the modeling of non-linear, multivariate relationships characteristic of self-healing concrete systems. ANNs possess the capability to learn from experimental datasets to accurately predict properties such as compressive strength recovery, healing index, and durability parameters [44]. In order to improve the precision of predictions and the robustness of models, scholars have investigated hybrid models that integrate machine learning with optimization algorithms [45,46].

Machine learning is categorized by data labeling into supervised, unsupervised, and semi-supervised learning. Supervised learning, the most common in radiotherapy for planning evaluation or outcomes prediction, estimates unknown mappings using labeled samples from experts or clinical endpoints [47]. Model validation strategies, like the holdout method, prevent overfitting by splitting data into training, validation, and testing sets. The training set builds models, while the validation set selects the model with the smallest prediction error. The testing set assesses the final model’s prediction error independently, minimizing test error through the bias-variance trade-off [48].

To assess the efficacy of bacterial-based self-healing concrete (BSHC), the concept of healing performance (HP) is introduced. HP quantifies the proportion of cracks that can be effectively repaired, calculated using the following equation based on initial and final crack condition assessments. The evaluation of cracking conditions is performed utilizing five distinct measurement techniques: crack width measurement, crack area measurement, ultrasound evaluation, regained strength assessment, and anti-seepage repair measurement [39,49].

$$HP={\displaystyle \frac{\left(Cwi-Cwt\right)}{Cwi}}\times 100$$

(2)

where Cwi is the initial cracking condition, Cwt is the final cracking condition measured at a specific curing time, and HP is the healing performance.

4.2. Machine Learning and Bibliometrics for Self-Healing Concrete

Self-healing concrete has grown significantly in the last two decades due to rising demand for sustainable, low-maintenance materials [50]. As the corpus of literature continues to expand, researchers are increasingly utilizing sophisticated tools to traverse the intricate landscape of extant knowledge. The concurrent application of machine learning (ML) and bibliometric analysis constitutes a formidable methodology for comprehending the existing research milieu and directing subsequent investigations in the domain of self-healing concrete [51]. Machine learning predicts material properties and performance, while bibliometrics analyzes scientific trends, collaborations, and thematic evolution [52,53].

Bibliometrics encompasses the quantitative examination of scholarly texts, focusing on publication trends, author relationships, keyword co-occurrences, and citation behaviors. Within the realm of self-healing concrete, the bibliometric tool VOS viewer has been employed to pinpoint research on bacterial healing, encapsulation techniques, and durability assessments, as well as to follow advancements in materials and technologies [54]. Bibliometric analyses have the capacity to elucidate key authors, institutions, and journals, thereby offering a comprehensive overview of the research ecosystem. Such data is crucial for researchers aiming to discern gaps, establish collaborative efforts, and prioritize significant research domains. The synthesis of machine learning and bibliometrics generates a feedback mechanism that augments both scientific innovation and the efficacy of modeling practices [55,56].

Combining machine learning and bibliometrics poses challenges, such as the need for standardized data and complex integration of diverse information sources. Researchers must also ensure technical rigor and interpretability, especially with deep learning. However, advancements in AI-driven bibliometrics, semantic analysis, and open-access databases are expected to mitigate these challenges [57]. Through the integration of knowledge discovery and performance prediction, the utilization of machine learning in conjunction with bibliometrics presents a promising avenue for enhanced precision, expedited advancement, and more focused innovation within self-healing concrete technologies.

The prognostic efficacy of machine learning (ML) models employing five distinct algorithms for estimating the horsepower (HP) of Bamboo Scrimber Hybrid Composites (BSHC) is quantitatively evaluated using the coefficient of determination (R²) and the root mean square error (RMSE). The RMSE, which is defined as the square root of the mean square error (MSE), is equivalent to the standard error. Importantly, the RMSE demonstrates heightened sensitivity to significant deviations in predictive values, thus providing a precise measure of predictive accuracy. Sensitivity analysis (SA) serves as a method for the interpretation of machine learning models. Additionally, it functions as an uncertainty analysis method to examine the impact of variables on the output derived from quantitative analysis [49].

The Artificial Neural Network (ANN) represents a computational model that is profoundly inspired by biological systems and holds significant applicability within the domain of civil engineering. It emulates the cognitive processes observed in the human brain, which comprises billions of neurons interconnected through axonal pathways. Furthermore, the efficacy of both the ANN and the Response Surface Methodology (RSM) was assessed via various statistical indicators. Specifically, metrics such as residuals, mean square error (MSE), and root mean square error (RMSE) were employed as evaluative estimators within the context of statistical analysis [58].

To enhance the development of bacteria-based self-healing concrete, six machine learning (ML) methods are used to predict calcium carbonate precipitation (CCP) due to their effectiveness in nonlinear forecasting. Particle Swarm Optimization (PSO) is used to optimize hyperparameters for these ML models. The training set aids in training the models, while the testing set evaluates their accuracy. Leveraging historical data, artificial neural networks (ANNs) predict new data by forming synaptic connections like neurons. ANNs comprise input, output, and possibly multiple hidden layers with interconnected neurons. They approximate nonlinear functions without explicit model equations and are used in pattern classification, optimization, and prediction across various fields, including load forecasting and strength prediction [59].

The integration of machine learning techniques and bibliometric analysis is progressively advancing within the domain of bacterial concrete research, facilitating a more profound, data-driven comprehension of the discipline. Bibliometric analysis offers a systematic examination of scientific literature by delineating publication trends, identifying highly cited publications, influential authors, organizations, and tracing the evolution of pivotal concepts such as bacteriology, biomineralization, crack remediation, microencapsulation, and hydrogel-based systems. Employing methodologies such as keyword co-occurrence networks, citation analysis, and collaboration mapping, bibliometrics elucidates research hotspots and identifies lacunae within the global epistemic landscape of bacterial concrete research. This structured analytical framework enables the subsequent application of machine learning methodologies.

Incorporating machine learning methodologies markedly enhances the bibliometric framework by uncovering patterns and predictive associations that are frequently overlooked by conventional statistical approaches. The deployment of advanced methodologies such as clustering, topic modeling, supervised learning, and natural language processing permits researchers to systematically classify scholarly publications, discern emerging thematic domains, and predict future research trajectories. Topic modeling is particularly adept at autonomously identifying thematic structures within extensive literature datasets pertaining to bacterial concrete, while clustering algorithms enable the aggregation of analogous research domains, including healing efficiency, calcium carbonate precipitation, bacterial encapsulation, and durability performance. Furthermore, machine learning can forecast emergent keywords or research trends, thus empowering scholars to anticipate and adapt to forthcoming shifts within the academic landscape.

The integration of bibliometric analysis with machine learning methodologies facilitates significant advancements in the domain of bacterial concrete research. Bibliometric analysis serves to map the extant body of knowledge, whereas machine learning augments this by offering predictive insights and automating the recognition of patterns. Collectively, these tools establish a robust framework for comprehending the evolution of the field, identifying existing knowledge gaps, highlighting current innovations, and proposing future research trajectories. This synergistic approach supports decision-making processes for researchers, funding bodies, and industry stakeholders, thereby aligning scientific progress with the requirements of sustainable construction and the development of self-healing concrete materials.

4.3. Performance of Self-Healing Concrete

The mechanical properties of concrete, including its resistance to water penetration, are affected by the bacterial self-healing process. The efficacy of crack repair in bacteria-based concrete specimens is directly proportional to the extent of mineral precipitation coverage [60]. Microbially induced carbonate precipitation (MICP) or bacterial healing demonstrates optimal efficacy in the remediation of narrow fissures, with commonly reported effective limits ranging approximately from 0.1 to 1.0 mm. Numerous studies have indicated that effective sealing is frequently achieved within the narrower spectrum of approximately 0.3 to 1.0 mm. Conversely, the sealing of wider fissures poses a greater challenge, even when high surface coverage is attained, due to critical factors such as crack depth, the continuity of the deposit, and the density of the precipitate [61].

Within this framework, the geometry of the crack effectively retains both bacterial cells and nutrient solutions, thereby promoting sustained biomineralization along the fissure walls. Furthermore, the diffusion distance for nutrients and ions is minimized, which facilitates homogeneous calcium carbonate (CaCO₃) precipitation capable of bridging the crack edges efficiently. Consequently, the healing process is expedited and exhibits greater structural continuity, often achieving near-complete closure when the crack width is reduced to the interval of 0.3–1.0 mm, as documented in numerous experimental studies. In contrast, broader cracks create a significantly different micro-environment that diminishes the efficacy of bacterial sealing, even in cases where mineral precipitation coverage on the crack surface is extensive. Firstly, the increased depth and width of the crack introduce a larger volume that necessitates filling, demanding a more substantial and sustained production of CaCO₃ than is typically achievable by bacteria under standard nutrient conditions. Secondly, precipitation in wider cracks tends to occur predominantly along the exposed surfaces, resulting in the formation of a thin superficial layer rather than a dense, mechanically interlocking deposit throughout the crack depth. This leads to discontinuous sealing, leaving internal voids that compromise structural integrity and permit water ingress.

The chemical makeup of various matrix elements aids in self-repair, with the subsequent addition of bacteria to the matrix boosting this self-healing capability. Nevertheless, the document does not define the optimal bacterial percentage or examine how different concentrations affect durability and mechanical properties. It does, however, present a study employing machine learning to predict the self-healing potential of concrete that integrates diverse types of industrial slag waste to restore its mechanical properties [62,63]. Unhydrated particles undergo dissolution in water, leading to the generation of a supersaturated ionic solution, subsequently resulting in the formation of hydrates that act as nucleation sites for hydration. This phenomenon is recognized as the dissolution precipitation reaction of cement with water [64].

There are various self-healing material approaches, mainly using chemical agents in brittle fibers or microcapsules. When cracks occur, these release an air-hardening agent to fill them. Notably, this method allows for diverse chemical compositions for crack filling [65]. The precipitation of calcium carbonate (CaCO₃) serves as an effective and compatible bonding agent within the cement matrix, facilitating its densification through the filling of pores and cracks. Consequently, this process can result in the restoration of mechanical properties and a reduction in water permeability [66,67]. Self-healing technologies mimic biological systems’ ability to repair damage. Using intrinsic, capsule-based, and vascular mechanisms, these technologies and their self-healing agents can be added to concrete mixtures [68].

The preservation of conditions favorable to microbial proliferation and calcium carbonate deposition is crucial for effective autogenous healing, thereby facilitating the development of infrastructure that is durable, resilient, and requires minimal maintenance [69,70]. Bacteria have demonstrated enhanced calcium-precipitating capabilities in relation to concrete durability [60,71]. Researchers seek creative solutions to shield concrete structures from harsh environments and lower repair costs. Self-healing concrete is an innovative solution that automatically fixes cracks [72].

Different bacterial strains have different efficiencies and capabilities when it comes to producing calcium carbonate, which is crucial for the self-healing properties of concrete. The choice of bacterial strain can affect the effectiveness of crack healing in concrete by influencing the extent and quality of calcium carbonate precipitation, which is a key mechanism in the self-healing process [73]. Optimizing urea and calcium-based material concentrations is crucial to provide enough carbonate ions for precipitation while maintaining the concrete’s structural integrity [60,72]. Various self-healing technologies are employed within large concrete structures, aiming to understand and enhance their ability to repair cracks autonomously, thereby ensuring structural integrity and longevity. These systems are likely separate from, or in addition to, the bacterial systems focusing on calcium carbonate precipitation for crack repair [74]. The permeability of water is reduced in concrete samples that have incurred cracks due to the autogenous process of crack healing [75,76]. Upon the occurrence of cracks in hardened concrete, the healing mechanism is initiated by the ingress of water and is maintained by the presence of oxygen. The metabolic activity of bacteria converts organic compounds into calcium carbonate, and such precipitation facilitates the healing of the cracks [77,78]. This metabolic activity of bacteria converts organic compounds into calcium carbonate, and such precipitation facilitates the healing of the cracks. It refers to the process by which bacteria, when introduced into concrete, can aid in its self-repair. When cracks occur in concrete, the water that enters the cracks activates the bacteria present. These bacteria use organic compounds as a nutritional source and through their metabolic processes, convert these compounds into calcium carbonate. This calcium carbonate forms as a precipitation within the crack, effectively filling it. By filling in the crack, calcium carbonate aids in the self-healing process of the concrete, helping to restore its structural integrity and reduce water permeability.

4.4. Current Practices, Research Gap, and Future Directions

Emerging trends in environmentally friendly concrete materials are transforming the construction industry, addressing ecological concerns while improving performance and durability [35]. A notable trend involves the incorporation of industrial by-products such as fly ash, slag, and silica fume as supplementary cementitious materials. These materials not only mitigate the carbon footprint associated with concrete production but also augment its mechanical properties and long-term durability [79]. Alternative binders like geopolymers and calcium sulfoaluminate cement provide significant environmental advantages over Portland cement [80]. Calcium sulfoaluminate cements offer lower energy and CO₂ emissions, serving as an eco-friendly alternative for concrete [81].

Nano-additives like nano-silica and nano-titania improve concrete strength, reduce permeability, and mitigate environmental impacts. These additives also enable self-healing concrete, repairing cracks autonomously via calcium carbonate or other agents [82]. Biomimetic mineralization, using CaCO₃ crystal modifiers, governs calcium salt reactions [53]. Biomineralization in nature, seen in organisms like mollusks and sponges, provides sustainable mortar options [83]. Eco-friendly concrete materials are increasingly adopted in construction, driven by sustainability needs. Bacterial concrete has emerged as a potential solution to address the inherent weaknesses of conventional concrete structures. This innovative material possesses the distinctive ability to autonomously repair cracks, thereby enhancing the durability and longevity of infrastructure systems [64].

Recent advancements in microbiology, materials science, and civil engineering have directed research efforts towards the comprehensive utilization of bacterial concrete in various applications. The employment of genetically modified bacteria for improved crack healing in concrete structures has significantly increased healing efficiency and long-term durability. This breakthrough highlights the importance of exploring novel microbial technologies to optimize the performance of bacterial concrete in real-world scenarios [40,84]. Moreover, ongoing research is concentrated on identifying environmentally sustainable approaches to the production and application of bacterial concrete. Recent studies exploring alternative nutrient sources, fermentation processes, and manufacturing techniques to improve the sustainability profile of bacterial concrete [85] constitutes an additional promising research direction within the domain of bacterial concrete. Through the integration of sensors and actuators into bio-concrete structures, researchers endeavor to develop self-diagnostic systems proficient in monitoring structural health and identifying precursors to damage. This interdisciplinary approach has the potential to transform maintenance methodologies and prolong the operational lifespan of infrastructure assets [86]. The establishment of standardization and regulatory frameworks is pivotal in ensuring the quality, safety, and reliability of bacterial concrete within construction applications. Recent initiatives have been focused on the creation of standardized testing protocols and industry guidelines, aimed at facilitating the widespread adoption of bio-concrete technology [87].

Concrete is characterized by a highly alkaline pore solution (pH 12–13), high mechanical density, and exposure to substantial thermal and chemical stresses. Maintaining microbial viability (or the reactivatability of spores) over multi-year timescales—and ensuring that these microorganisms germinate and induce localized CaCO₃ precipitation specifically at sites of cracking—constitutes a central technical challenge in the development of bio-based self-healing systems. Various approaches, including the use of spore-forming microorganisms, protective encapsulation strategies, carrier materials, and modifications to concrete mix design and chemistry, have been proposed to mitigate these constraints; however, their performance exhibits considerable variability between controlled laboratory conditions and in situ field applications [88]. Bacteria-mediated calcite precipitation requires suitable substrates (e.g., calcium sources, organic carbon, and sometimes urea). Supplying nutrients that remain stable in the cementitious matrix, do not impair early-age mechanical properties, and, in addition, become bioavailable only after cracking is technically challenging. Some substrates, notably urea-based ones, can also cause undesirable effects such as ammonia emissions and related environmental impacts. Encapsulation strategies (e.g., microcapsules, lightweight aggregates, pelletized carriers) have been widely studied to mitigate these issues, but reliable control of nutrient release, mechanical compatibility with the host matrix, and cost-effective large-scale production are still unresolved challenges [89].

Certain MICP pathways, particularly those based on ureolysis, generate ammonia and ammonium as by-products, which raises concerns regarding localized gaseous emissions, odor nuisance, and potential environmental impacts when deployed at large scale. Existing life cycle assessment (LCA) studies indicate that some bio-concrete systems may offer net CO₂ mitigation benefits; however, these outcomes are highly sensitive to assumptions about nutrient sourcing, encapsulation production processes, and transportation requirements. To date, comprehensive cradle-to-grave LCA and environmental fate analyses of these technologies remain limited [89]. Although numerous studies employ non-pathogenic Bacillus spp.—which are spore-forming and generally recognized as safe—regulatory authorities and the public continue to express concerns regarding the deliberate release of engineered microorganisms, including the potential for horizontal gene transfer and unintended, persistent colonization of environmental niches. Consequently, robust demonstration of negligible ecological risk, along with evidence of effective biological and physical containment, is a prerequisite for large-scale or widespread field deployment [90].

Delivering consistent, shelf-stable biological additives (viable spores, encapsulated pellets, nutrient blends) at industrial scale requires controlled bioprocessing, stabilization (drying, freeze-drying), QC, cold-chain or validated ambient-stable formulations, and compatibility with ready-mix plants. These supply chain and QA/QC steps are complex and add cost [90]. Industry favors solutions that fit existing workflows (ready-mix batching, precast production, pumping). Methods needing extra on-site steps, special curing, or special handling reduce contractor adoption. Some field demonstrations exist, but broad use requires drop-in solutions or clear cost–benefit cases [91].

High upfront material costs and uncertain long-term performance weaken the commercial case despite potential lifecycle savings. Owners, insurers, and asset managers need robust long-term trials demonstrating reduced maintenance and extended service life to justify premium pricing. Without recognized self-healing performance standards or test protocols, industry cannot reliably specify products. Standardized test methods (e.g., crack healing under realistic environmental cycles, corrosion suppression metrics) and their inclusion in national standards or procurement specifications are needed for scale [92]. It is imperative that collaboration between researchers, industry stakeholders, and regulatory bodies be fostered to expedite the development and commercialization of bacterial concrete products. See

Table 3. Current practices and future directions.

Focus	Recent Progress	Challenges	Future Directions	Authors
Supplementary Cementitious Materials	The incorporation of fly ash, slag, and silica fume effectively contributes to the reduction of CO2 emissions, while concurrently enhancing the mechanical properties and longevity of the material in bacterial concrete.	Optimization of mixture proportions and consistency of performance across diverse environmental conditions in bacterial concrete.	Augmented utilization in high-performance bacterial concrete; lifecycle evaluation and integration with supplementary sustainable methodologies.	[79]
Alternative Binders (Geopolymers & cement)	The utilization of geopolymers derived from industrial by-products in conjunction with calcium sulfoaluminate cement presents a viable approach to reducing carbon dioxide emissions and conserving energy.	The capacity for scalability, comprehensive long-term performance data, and cost-effectiveness across diverse conditions in bacterial concrete.	Production on an industrial scale, compatibility assessments involving reinforcement, and the attainment of regulatory approval.	[80,81]
Nano-Additives	Utilization of nano-silica and nano-titania to augment strength, diminish permeability, and facilitate self-healing.	The safety of nanoparticles regarding health and the environment, coupled with the absence of standardized testing protocols, presents a significant concern.	Advancement in the creation of secure nano-concretes, the incorporation of sensor technologies, and the implementation of standardization measures with self-healing concrete.	[82]
Biomimetic/Biomineralization Techniques	Employment of calcium carbonate (CaCO3) crystal modifiers as derived from biological entities, such as mollusks.	Paucity of empirical research on real-world applications, and durability assessments under diverse conditions remain constrained.	Extensive fabrication of biomimetically engineered mortars for structural applications.	[53,83]
Effectiveness of Bacterial Concrete	Autonomous repair facilitated by bacterial activity, such as the precipitation of calcium carbonate (CaCO3), resulting in improved durability.	The supply of nutrients, the resilience of bacteria under imposed stress, and the effectiveness under fluctuating load conditions.	Employment of genetically engineered bacteria to enhance efficiency and resistance.	[40,84]
Sustainable Production of Bacterial Concrete	Investigation into alternative sources of nutrients and the processes of fermentation.	Enhancement of eco-friendly manufacturing techniques.	Sustainability analysis across the entire lifecycle; environmentally friendly additives and a minimized ecological footprint.	[85]
Smart Bacterial Concrete (Self-Diagnostic)	Incorporation of sensors and actuators within bio-concrete for the purpose of structural health monitoring.	The dependability and economic implications of embedded systems; real-time data processing.	Self-regulating and autonomous monitoring systems are crucial with advanced technological integration and essential infrastructure.	[86]
Standardization & Regulation	Formulation of testing protocols and establishment of industry guidelines.	Absence of harmonized global criteria.	Joint advancement in the formulation of codes, establishment of performance benchmarks, and development of certification frameworks.	[87]

for more clarification of current practices and future directions.

4.5. Regulatory and Standardization Considerations

The existing regulatory framework for bacterial concrete emphasizes the significance of safety, performance, and environmental compatibility, thereby imposing rigorous criteria concerning the viability, stability, and non-toxicity of bacterial strains [93]. These criteria are established to ensure the safety of the material for construction workers, building occupants, and the environment. Moreover, it is imperative for bacterial concrete to demonstrate adequate mechanical properties, durability, and resistance to various environmental conditions in order to fulfill structural application standards [94]. Of particular importance are the environmental regulations that pertain to genetically modified organisms (GMOs), which play a critical role in supervising the use of bacterial concrete. Regulatory authorities require thorough evaluations of the environmental impact and potential risks associated with the release of genetically modified bacteria [95].

As bacterial concrete becomes increasingly prominent for enhancing the durability and sustainability of concrete structures, standardization efforts have been initiated to establish guidelines and protocols for its production, testing, and application [96].

It is imperative that various organizations and committees engage proactively in standardization activities relevant to bacterial concrete. In addition to organizational standardization efforts, the establishment of collaborative research projects and consortia is crucial for advancing the standardization of bacterial concrete. These initiatives leverage multidisciplinary expertise to address technical challenges, validate performance metrics, and facilitate the dissemination of knowledge and best practices [97]. The implementation of policy measures, including grants and subsidies, has the potential to incentivize developers, contractors, and manufacturers to allocate resources towards the advancement of bacterial concrete technology, thereby facilitating its market proliferation and acceptance. Additionally, the inclusion of bacterial concrete specifications within public procurement policies can create substantial market demand, promoting its application in government-funded infrastructure projects [98]. Policymakers should promote training, certification, and knowledge-sharing to adopt best practices in bacterial concrete use.

The extensive adoption of bacterial concrete is contingent upon the formulation of robust policy frameworks specifically designed to address the multifaceted challenges and opportunities associated with this innovative construction material. The establishment of clear regulatory standards is imperative to ensure the safety, reliability, and quality of bacterial concrete in construction endeavors [15,99]. Collaborative endeavors among policymakers, industry stakeholders, and standardization bodies are indispensable for the development of comprehensive guidelines encompassing production, testing, and application procedures. Governmental investment in research and development (R&D) is crucial to advance the progression of bacterial concrete technology. Policymakers ought to allocate funding for collaborative R&D initiatives, academic research, and technology transfer activities. Such investment would facilitate accelerated innovation, address technical obstacles, and promote the development of cost-effective production methods, thereby advancing the extensive adoption of bacterial concrete [100,101].

Financial incentives and subsidies play a crucial role in enhancing the demand for bacterial concrete and promoting its integration into construction projects [68]. Policy measures such as grants or subsidies can motivate developers, contractors, and manufacturers to invest in bacterial concrete technology, thereby nurturing market growth and adoption. Furthermore, incorporating bacterial concrete specifications into public procurement policies can generate a significant market pull, encouraging its use in government-funded infrastructure projects [98].

Table 4. Regulatory and policy framework for bacterial concrete.

Category	Key Points	Research Area	Research Focus	Authors
Regulatory Framework	Prioritizes safety, performance, and environmental compatibility, necessitating the non-toxicity, viability, and stability of bacterial strains.	Regulatory Science, Microbiology	Bacterial viability and safety evaluation. Assess the ability of bacterial cells to survive under certain conditions (viability) and to determine if those conditions pose any risks to health or safety. This type of evaluation might be conducted to ensure that bacterial levels remain safe for human exposure.	[93,94]
	The mechanical performance and longevity of materials are crucial for their implementation in structural applications.	Structural Engineering	Mechanical assessment and durability analysis entail the evaluation of mechanical properties or the performance of a system or component. This process may include tests or evaluations related to strength, flexibility, efficiency, and functionality to ensure compliance with the required standards and specifications.	[94]
	The regulatory framework governing genetically modified organisms requires an extensive assessment of environmental risks prior to the release of genetically modified bacteria.	Environmental Science, Biosafety	Analysis of biosafety risks and environmental impact. The process of evaluating potential risks to biological safety and the environment. In this context, it involves examining how certain activities, substances, or technologies might pose hazards to living organisms and their surroundings. This analysis is critical for identifying any adverse effects and implementing strategies to mitigate them, ensuring the protection of both human health and the environment.	[95]
Standardization	Guarantees the safety, quality, and dependability of bacterial concrete as a construction material.	Materials Science, Concrete structure	Protocols for quality assurance and material specifications. Establish guidelines and standards designed to ensure that products or materials meet certain quality levels. In a broader context, it involves procedures for checking that the materials used in production conform to specific standards and specifications. Quality assurance protocols aim to maintain consistency, safety, and reliability in the manufacturing or production process of bacterial concrete.	[96]
	Structured guidelines and protocols are in the process of being formulated for the purposes of production, testing, and application.	Construction Standards, Metrology	Development of standardized test methods. The process of creating uniform procedures and criteria for testing. In a broader context, this involves establishing commonly accepted protocols to ensure that test results are consistent, reliable, and can be compared across different scenarios.	[96]
	The engagement of standardization entities and committees is essential for effective participation.	Policy & Governance	Adhering to regulations, interacting with stakeholders. emphasizes the importance of complying with rules and guidelines set by authoritative bodies while maintaining active communication and engagement.	[96]
	Collaborative research and consortiums facilitate the resolution of technical challenges and the advancement of best practices.	Multidisciplinary R&D	Collaboration across disciplines and sharing of knowledge. It suggests working together with individuals from different areas of expertise and exchanging information. This emphasizes the importance of integrating diverse perspectives and specialized knowledge to enhance understanding, problem-solving, and innovation in various contexts	[97]
Policy Framework and R&D Investment	Regulatory standards must be robust for widespread adoption.	Public Policy, Construction Law	Creation of frameworks and the establishment of legal standards. The process of developing structured guidelines and rules that govern behavior within a particular field or context. Frameworks provide a systematic way to structure and organize information or processes, while legal standards are formally adopted rules that are enforceable by law.	[15,99]
	Cooperation among policymakers, industry stakeholders, and standardization bodies is vital.	Innovation Policy, Urban Planning	Integration of policies, collaboration among stakeholders. The idea of combining various policies to work together effectively while ensuring that all parties work collaboratively towards common goals or solutions.	[15,99]
	Governmental R&D funding should support innovation, overcome technical barriers, and promote cost-effective methods.	Technology Development, Economics	Approaches to securing funds and reducing expenses. It refers to strategies or methods focused on obtaining financial resources while also minimizing costs. In the given context, it suggests exploring ways to increase financial inflows and decrease financial outflows in order to enhance financial stability.	[100,101]

clarifies the regulatory and policy framework for bacterial concrete.

5. Conclusions

The novel utilization of bio-based bacterial concrete as a sustainable alternative to conventional concrete in the construction industry represents a significant area of study. Bibliometric analysis assesses the distribution across research areas, citation patterns, and keyword usage, incorporating fundamental discussions. The research questions that guide the review by focusing on critical aspects that influence feasibility, application, and challenges of using bacteria in concrete production are as follows:

▪

How is the distribution of self-healing concrete research currently conducted and applied in the construction industry?

▪

How are bibliometric analysis and machine learning combined show the integrity of self-healing concrete research?

▪

What is the performance of self-healing concrete in terms of durability and mechanical properties of concrete?

▪

Performance of self-healing concrete.

▪

What are the current practices, regulatory and standardization considerations, research gap, and future directions of bacterial concrete?

The documentation, organized by country or territory, offers a clear regional breakdown for comparing contributions from continents like Asia (China, India, Malaysia, Saudi Arabia), Europe (UK, Belgium, Netherlands, Italy), North America (United States), and Australia. Researchers identify publication trends, top institutions, and key subjects, while academic institutions compare their output to global trends to identify strengths. Policymakers use research patterns to inform funding and strategic priorities.

The predictive model for self-healing concrete analyzes material variables to forecast crack thickness changes. Parameters include the mineral content of fly ash, silica fume, limestone powder, cement waste, and initial crack thickness. The artificial neural network (ANN), modeled after biological networks, has neurons in an input layer, multiple hidden layers, and an output layer, interconnected by weighted connections without same-layer links. Researchers use machine learning to predict properties and bibliometric analysis to track trends and collaborations with tools like VOS viewer. This highlights key authors and research gaps, promoting collaboration and research focus.

The integration of bibliometric analysis with machine learning methodologies facilitates significant advancements in the domain of bacterial concrete research. Bibliometric analysis serves to map the extant body of knowledge, whereas machine learning augments this by offering predictive insights and automating the recognition of patterns. Bacterial self-healing mechanisms in concrete (calcium carbonate precipitation) improve concrete’s mechanical properties and water resistance. Microbially induced carbonate precipitation (MICP), or bacterial healing, demonstrates optimal efficacy in the remediation of narrow fissures, with commonly reported effective limits ranging approximately from 0.1 to 1.0 mm. Numerous studies have indicated that effective sealing is frequently achieved within the narrower spectrum of approximately 0.3 to 1.0 mm.

The challenges associated with data standardization and complexity are anticipated to be alleviated through the application of artificial intelligence, particularly the amalgamation of machine learning techniques and bibliometric analyses. This integration has the potential to augment accuracy and foster innovation within the realm of self-healing concrete research. Identified research gaps include the use of genetically modified bacteria, the pursuit of sustainable production processes, the integration of smart materials, and the establishment of standardization protocols, all of which are vital to ensure the quality and reliability of bacterial concrete. Future research should focus on advancements in the development of secure nano-concrete, the integration of sensor technologies, and the implementation of standardization measures in conjunction with self-healing concrete. Furthermore, the extensive fabrication of biomimetically engineered mortars for structural applications, alongside the utilization of genetically engineered bacteria to improve efficiency and resistance, remains a pressing issue within the field.