Typology-Specific Gaps in Building Fire Safety: A Scientometric Review of Technologies, Functions, and Research Trends

Abstract

Fires remain a critical threat to the resilience and safety of the built environment, yet current research is often fragmented across building types, technologies, and functions. This study investigates typology-specific gaps in fire safety by conducting a scientometric review of peer-reviewed articles published between 2010 and 2025. Following a PRISMA-guided protocol, a total of 83 studies indexed in the Web of Science were systematically screened and analyzed using VOSviewer (v1.6.19) and the R-based Bibliometrix package (version 4.2.1). The dataset was classified according to building typologies, fire safety functions—detection, suppression, and evacuation—and applied technologies such as BIM, simulation platforms, and AI-based models. The results show a strong research bias toward evacuation modeling in high-rise and general-purpose buildings, while critical typologies including healthcare facilities, heritage structures, and informal settlements remain underexplored. Suppression systems and real-time detection technologies are rarely integrated, and technological applications are often fragmented rather than interoperable. A conceptual matrix is proposed to align tools with typology-specific risk profiles, highlighting mismatches between research priorities and building functions. These findings emphasize the need for integrated, data-driven, and context-sensitive fire safety strategies that bridge methodological innovation with practical application, offering a roadmap for advancing resilient and adaptive fire safety in diverse urban settings.

1. Introduction

Fires, whether originating from natural or anthropogenic sources, represent one of the most critical threats to the safety, resilience, and sustainability of the built environment. “This risk is particularly pronounced in rapidly urbanizing regions”, where high-density and multifunctional constructions intensify urban vulnerabilities associated with fire risk [1,2]. Despite significant advancements in fire engineering and evolving regulatory frameworks, real-life fire incidents continue to reveal substantial performance gaps in fire safety systems [3,4,5]. Fires exert “multidimensional impacts” that extend beyond physical destruction, including threats to human life, environmental degradation, and major economic losses. “Scenarios involving delayed evacuations, toxic gas exposure, structural collapse, and post-fire pollution illustrate the need to reframe fire events as socio-environmental and governance challenges, rather than purely technical phenomena” [6,7]. Traditional fire safety models, often grounded in prescriptive design codes, are increasingly inadequate in the face of emerging construction typologies and vertical urban growth.

Contemporary fire safety research generally revolves around four interrelated thematic pillars: fire hazard, fire prevention, fire protection, and fire mitigation [8]. However, many studies remain narrowly focused on specific building types or isolated technologies, leading to a fragmented body of literature lacking a unified analytical framework [9,10]. Moreover, debates persist regarding the effectiveness of performance-based fire safety approaches; while some scholars advocate their flexibility and adaptability [11], others point out their implementation complexity and regulatory uncertainty [12]. To address these limitations and reveal deeper structural patterns, this study adopts a scientometric approach, combining “keyword co-occurrence analysis, co-authorship networks” and “typology-function cross-tabulation” using tools such as VOSviewer (v1.6.19) and the R-based Bibliometrix package (version 4.2.1) [13,14].

The main aim of this study is to construct a “typology-sensitive conceptual matrix” that links building types with fire safety system components detection, suppression, and evacuation mechanisms. This framework not only identifies critical gaps in underrepresented contexts such as healthcare and heritage buildings but also exposes functional imbalances and regional disparities in the application of fire safety technologies. Figure 1 illustrates the thematic structure that underpins this study, synthesizing the core research domains and their co-occurring sub-concepts.

This analytical structure serves as the foundation for mapping typology-function-technology relationships across the paper. In summary, the study provides a novel framework to bridge the persistent gap between academic research and real-world fire safety practice. The results reveal concentration biases in the literature, identify under-researched building typologies, and propose more integrated and context-sensitive safety strategies to inform future research, design practice, and policy.

1.1. Previous Studies

Fire safety constitutes a rapidly evolving and inherently multidisciplinary research domain situated at the intersection of structural engineering, disaster management, architecture, and environmental design. In its early stages, studies predominantly focused on engineering-oriented topics such as the physical behavior of fire, combustion mechanisms, material flammability, and flame propagation. During this period, concepts like combustion, heat release rate, and flame spread featured prominently in scholarly discourse [8,15].

Since the mid-2000s, the scope of fire safety research has expanded considerably, encompassing not only physical phenomena but also risk assessment, scenario modelling, fire management systems, and human behavior in emergencies [9,16]. This shift has been facilitated by the increasing presence of interdisciplinary approaches and computational technologies. Simulation tools such as FDS, PyroSim, and Pathfinder, along with Building Information Modelling (BIM), AI-assisted decision systems, and evacuation analytics, have become widespread in fire safety literature and practice [17,18,19,20].

Since 2015, several prominent thematic directions have emerged, including:

• Vertical evacuation strategies in high-rise buildings;
• Fire resilience of heritage and cultural sites;
• Fire safety in sustainable and green buildings;
• Digital fire management systems integrating BIM and IoT platforms [10,21,22,23].

Alongside these developments, scholars have increasingly examined the integration of sustainability principles such as life-cycle cost analysis, carbon footprint reduction, and post-disaster recovery within fire safety frameworks [7,24]. This indicates a conceptual transition from technically constrained analyses to broader system-oriented approaches, as summarized in

Table 1. Evolution of Fire Safety Research Themes, Tools, and Key Concepts (1990–2025).

Period	Thematic Focus	Key Concepts and Tools	References
1990–2004	Physical fire behavior, material response, flame propagation	Flame spread, combustion, fire load, heat release rate, structural collapse	[8,15,25,26,27,28,29]
2005–2014	Risk analysis, scenario modelling, performance-based systems, human behavior	Fire risk assessment, agent-based evacuation models, fire scenarios, CFD, PyroSim, Pathfinder	[9,11,16,30,31,32,33,34]
2015–2019	High-rise buildings, heritage buildings, BIM integration, green fire safety	BIM-FDS integration, neural network modelling, cultural heritage fire resilience	[10,18,19,20,22,35,36,37]
2020–2025	Sustainability, AI & IoT, multi-hazard resilience, digitalization in fire safety	Smart sensors, life-cycle cost, post-earthquake fire, VR evacuation drills, carbon footprint	[7,24,38,39,40,41,42,43]

.

Moreover, this period has witnessed growing academic attention towards the integration of environmental performance indicators, such as sustainability, life-cycle cost, and carbon footprint, into fire safety planning and analysis frameworks [7,24]. These developments reflect a conceptual shift from narrowly defined technical assessments to more holistic and system-oriented approaches tailored to the complex demands of contemporary built environments. Table 1 illustrates this thematic evolution by mapping the chronological progression of dominant research priorities, associated methodologies, and technological tools used in fire safety studies between 1990 and 2025. Fire safety practices in the literature are fundamentally shaped within two main approaches: prescriptive systems and performance-based designs. While prescriptive systems mandate compliance with predefined structural and technical requirements, performance-based systems provide more flexible solutions based on achieving functional safety objectives. In developed countries, the adoption of performance-based strategies has become increasingly widespread, enabling the development of adaptive, typology-sensitive fire safety frameworks [12,19].

Simulation-based tools have emerged as essential instruments for evaluating performance-based fire safety strategies. Computational platforms such as Fire Dynamics Simulator (FDS), Pathfinder, and PyroSim are widely employed to model fire spread, estimate evacuation times, and assess smoke control measures [17,23,39]. In particular, studies involving high-rise buildings frequently utilize these tools to analyze elevator-assisted egress, smoke propagation, and vertical circulation dynamics in realistic scenarios [20]. In addition to these applications, some studies also utilize AI-supported decision systems, visually supported planning, and behavior simulations [16,18]. These applications form the methodological backbone of the present study’s tool-function-typology framework. The effectiveness of fire safety measures varies significantly depending on building function. Each typology, residential, commercial, industrial, public, or heritage, presents distinct fire risk profiles and intervention requirements. In residential buildings, research has primarily focused on early warning systems, continuity of escape routes, and safe evacuation mechanisms [44]. In commercial and high-rise settings, active suppression systems, smoke control strategies, and vertical evacuation logistics become central considerations [23,45]. Healthcare facilities introduce complex challenges due to the presence of patients and individuals with limited mobility, necessitating tailored evacuation protocols [37,46]. Similarly, fire safety in heritage buildings must reconcile protection of life with the preservation of architectural and cultural value [9,10]. This typological diversity underscores the need for fire safety strategies that are not only function-specific but also adaptable to varying spatial, operational, and regulatory constraints. The present study builds upon this differentiation to examine how specific fire safety systems—detection, suppression, and evacuation—are distributed across building types, and to what extent functional gaps persist.

In recent years, fire safety research has increasingly incorporated digital technologies that offer new capabilities for both pre-incident planning and active response. Tools such as Building Information Modelling (BIM), sensor systems, Artificial Intelligence (AI), and the Internet of Things (IoT) enable real-time risk detection, integrated decision-making, and lifecycle-based fire management strategies [22,36]. Particularly, BIM-supported simulations have demonstrated the potential to identify fire risks during early design phases and maintain interoperability throughout the operational phase of buildings. Autonomous robotic interventions [40], AI-driven risk prediction systems [37], and IoT-based evacuation guidance solutions [47,48] have further expanded the technological frontier of fire safety practices. While these innovations present promising directions, the extent to which they are integrated into typology-specific safety strategies remains unclear. The current study systematically investigates how such technologies are adopted across different building types and whether their application aligns with functional safety priorities such as detection, suppression, and evacuation.

Despite the growing volume of publications, fire safety literature continues to exhibit several critical knowledge gaps, particularly in strategic and high-risk building types. In developing countries, comparative studies evaluating the practical implementation and institutional effectiveness of fire safety regulations remain limited. Similarly, the role of human behavior and social dynamics is often insufficiently modeled in evacuation scenarios, leading to overly deterministic representations [49]. Post-earthquake fire scenarios, which involve complex multi-hazard interactions affecting both structural integrity and emergency response, are still underexplored in modelling studies [32,42]. Furthermore, the integration of fire safety planning with sustainability goals, life-cycle cost analysis, and carbon footprint assessments is a relatively recent research frontier [7,24]. In this study, these gaps are systematically addressed by examining how building typologies correlate with the presence, absence, or partial implementation of fire safety systems. The analysis maps these inconsistencies and explores whether the adoption of advanced tools, such as BIM, AI, and IoT, varies by building function or remains fragmented across the sector.

1.2. Research Objectives and Questions

This study aims to investigate the intersection of fire safety technologies, building typologies, and functional system coverage through a scientometric and typology-specific mapping of peer-reviewed publications from 2010 to 2025. Rather than offering a general synthesis, the research develops a data-driven framework to analyze how safety components—detection, suppression, and evacuation—are distributed and emphasized across different building types. The following five objectives guide this investigation:

• To map the evolution and scope of fire safety research, including trends in publication year, document types, author networks, contributing countries, and source journals.

To assess the distribution and adequacy of fire safety measures across building typologies, such as residential, commercial, healthcare, and heritage structures, identifying typology-specific system coverage or neglect.

• To classify and evaluate the methodological and technological tools, including simulation platforms, BIM-integrated systems, risk assessment models, and AI-based applications, used in fire safety design and analysis.
• To identify critical research gaps and typology-function mismatches, particularly in areas like healthcare facilities, post-earthquake fire scenarios, and sustainability-linked planning.
• To develop a conceptual matrix that links building functions with fire safety technologies and implementation patterns, aiming to support future research and context-specific safety strategies.

To address these objectives, the following six research questions were formulated:

• Q1: How have “the research dynamics of fire safety in buildings” evolved in terms of publication years, types, authorship, contributing countries, and source journals?
• Q2: How effective and comprehensive are fire safety measures when assessed across different building functions?
• Q3: What methodological approaches are employed in fire safety management and risk assessment, and how do they align with typology-specific needs?
• Q4: Which technologies are applied in fire safety systems, and what is their functional distribution across typologies?
• Q5: How can pre-fire planning, emergency response, and post-fire recovery processes be enhanced in context-sensitive ways?
• Q6: What knowledge gaps exist in current fire safety literature, and how can future studies address these gaps through integrated and performance-based approaches?

By addressing these questions, the study not only maps the thematic and methodological evolution of fire safety research but also provides a critical framework for analyzing its practical, technological, and typology-specific dimensions. Ultimately, the aim is to bridge the persistent gap between academic exploration and real-world implementation, supporting engineers, architects, policymakers, and safety professionals in designing more context-sensitive, adaptive, and resilient built environments.

2. Materials and Methods

2.1. Research Design

This study adopts a data-driven research design that integrates structured literature sourcing with scientometric and typology-function analysis. The objective is to systematically investigate how fire safety technologies and methodological tools are distributed across building typologies, and to identify gaps and functional imbalances in their application. While the initial article selection process follows the logic of systematic literature protocols, guided by the PRISMA 2020 framework [50], the primary aim is not to conduct a general synthesis, but to extract measurable insights for typology-specific fire safety planning. The four-phase review structure, identification, screening, eligibility, and inclusion was adapted from established systematic review protocols [51,52] to ensure methodological transparency.

Research manuscripts reporting large datasets that are deposited in a publicly available database should specify where the data have been deposited and provide the relevant accession numbers. If the accession numbers have not yet been obtained at the time of submission, please state that they will be provided during review. They must be provided prior to publication.

2.2. Database Selection and Search Strategy

The Web of Science (WoS) was selected as the primary data source due to its high-quality, peer-reviewed coverage across engineering, architecture, safety science, and environmental disciplines. Boolean operators and thematic keyword strings were formulated directly from the research questions (Q1–Q6) to ensure relevance and coverage. Search expressions included combinations such as “fire safety AND building typology”, “fire protection AND simulation tools”, and “fire risk assessment AND BIM”.

2.3. Inclusion and Exclusion Criteria

The database search initially yielded 761 articles. These were screened in line with transparency and traceability principles derived from the PRISMA 2020 guidelines.

Inclusion criteria were:

(i)

peer-reviewed journal articles;

(ii)

published in English between 2010 and 2025;

(iii)

directly addressing fire safety in buildings;

(iv)

focused on risk analysis, simulation tools, or functional safety components.

Exclusion criteria involved duplicate records, non-academic sources (e.g., editorials, conference papers), and content misaligned with the six research questions. After abstract and full-text screening, 83 studies met all selection criteria and were included for analysis. Title and abstract screening were manually conducted to ensure thematic alignment with the research objectives. Studies were prioritized based on their relevance to the six research questions, with the top 120 articles selected for full-text evaluation. Each article was assessed in terms of methodological clarity, data applicability, and topical consistency. During the eligibility phase, 37 articles were excluded due to insufficient methodological transparency or lack of direct relevance to the research scope. Ultimately, a total of 83 peer-reviewed studies were retained as the final analytical dataset for typology-specific and scientometric investigation.

2.4. Data Extraction and Analysis Process

A custom-designed Excel matrix was used to extract metadata and classify each article based on publication year, building typology, safety components (detection, suppression, evacuation), and technological tools used. Two analytical paths were followed:

(a)

Scientometric Mapping: Using VOSviewer (v1.6.19) [13] and the Bibliometrix R package (v4.2.1) [14], co-authorship networks, keyword co-occurrence clusters, and thematic trends were visualized.

(b)

Typology-Function-Tool Cross-Tabulation: Each article was manually coded to determine which fire safety components were addressed, and how those aligned with specific building types (residential, commercial, high-rise, healthcare, heritage, etc.). This enabled the construction of matrices and figures linking building function with safety system coverage.

Title and abstract screening were manually conducted to ensure thematic alignment with the research objectives. Studies were prioritized based on their relevance to the six research questions, with the top 120 articles selected for full-text evaluation. Each article was assessed in terms of methodological clarity, data applicability, and topical consistency. During the eligibility phase, 37 articles were excluded due to insufficient methodological transparency or lack of direct relevance to the research scope. Ultimately, a total of 83 peer-reviewed studies were retained as the final analytical dataset for typology-specific and scientometric investigation.

The overall data sourcing, screening, and selection process is illustrated in Figure 2, which adapts PRISMA-guided principles to the structured collection of research articles and their classification for scientometric and typology-function analysis. The process includes objective formulation, keyword and database selection, multi-phase screening, and final dataset construction.

3. Results

3.1. Bibliometric Analysis

To address Research Question 1 (Q1), a comprehensive bibliometric analysis was conducted to trace the evolution of scholarly output in fire safety research between 1991 and 2025. This analysis includes metrics related to publication trends, document types, geographic distribution, author collaborations, disciplinary domains, and journal impact. The goal is to uncover patterns of growth, thematic concentration, and institutional or regional disparities that shape the academic landscape of fire safety in the built environment.

Figure 3 illustrates both the annual and cumulative number of publications retrieved from the Web of Science database. The growth trajectory remained modest during the 1990s, with a gradual acceleration starting in the early 2000s. A sharp rise is evident after 2015, aligning with global concerns about climate-induced disasters and the emergence of digital risk assessment tools such as FDS and BIM-integrated simulations. The steep incline in the cumulative curve, especially between 2020 and 2025, underscores the increasing prioritization of fire safety within the domains of architecture, engineering, and risk management.

A breakdown of document types, shown in Figure 4, reveals that peer-reviewed journal articles constitute the dominant category (n = 462), reflecting the prevalence of original research in the field. Conference proceedings account for 146 documents, highlighting the importance of scholarly exchange and real-time dissemination of emerging findings. Review articles remain relatively scarce (n = 26), indicating limited synthesis efforts. Early-access documents (n = 8) and book chapters (n = 3) represent marginal shares, suggesting that the field is primarily journal-driven.

Figure 5 presents the geographic distribution of publication outputs. China emerges as the most prolific contributor with 247 articles, followed by the United States (135) and the United Kingdom (47). However, when adjusted for citation impact, the United States and United Kingdom demonstrate significantly higher visibility, indicating their stronger influence on theoretical and methodological advancements. While China excels in quantitative output, the citation performance of Western countries suggests a more foundational role in shaping the intellectual direction of the field. Other contributors, such as Australia and Canada, show moderate levels of productivity. In contrast, countries like India and Russia remain underrepresented, pointing to potential disparities in research funding, institutional capacity, and regulatory enforcement.

Figure 6 highlights the disciplinary composition of fire safety research, with “Engineering” (464 publications) and “Construction Building Technology” (167) emerging as the dominant domains. This reflects the technical orientation of the field, where structural design, building systems, and safety technologies form the core of scholarly output. “Materials Science” and “Environmental Sciences–Ecology” also appear prominently, suggesting a growing emphasis on the fire performance of construction materials and the ecological implications of fire events. Other technical disciplines, such as “Energy Fuels” and “Mechanics,” contribute insights into fire dynamics, heat transfer, and prevention strategies. Notably, the presence of “Public Environmental Occupational Health” underscores the integration of human safety and occupational risk into the fire safety discourse. Overall, the distribution confirms that fire safety research is inherently multidisciplinary, situated at the intersection of engineering, materials science, environmental sustainability, and public health.

Figure 7 presents the distribution of academic journals contributing to fire safety research. Fire Safety Journal and Fire Technology lead the field, with 52 and 45 publications respectively, confirming their role as central platforms for disseminating empirical studies, simulation models, and regulatory discussions. These journals primarily focus on engineering-based analysis and computational methods and are often cited in high-impact technical literature. Other specialized venues, such as Fire and Materials and Process Safety Progress, highlight the growing importance of material performance and process reliability in fire safety systems. Their emphasis on fire-retardant materials, ignition thresholds, and suppression chemistry broadens the technological scope of the field.

Multidisciplinary journals like Sustainability and Journal of Building Engineering reveal a shift toward integrative approaches that connect fire safety with architectural design, environmental impact, and lifecycle optimization. Meanwhile, Safety Science and International Journal of Disaster Risk Reduction (IJDRR) address human behavior, disaster governance, and systemic resilience, elements that are increasingly recognized as integral to fire risk management in complex environments. This distribution reflects a diversification of publication strategies, with high-volume technical journals supporting foundational research, while interdisciplinary outlets facilitate dialogue on policy relevance, public health, and urban sustainability.

3.2. Keyword Analysis

To further address Research Question 1 (Q1), a two-tiered keyword analysis was conducted. The aim was to identify dominant research themes, evolving terminologies, and interconnections among conceptual clusters in the fire safety literature between 2010 and 2024. This section incorporates both thematic categorization and co-occurrence mapping to capture not only the frequency of terms but also the conceptual relationships they represent.

Figure 8 displays thematic clusters derived from author keywords and title-abstract terms. Each cluster reflects a distinct knowledge domain, including performance-based fire safety design, material behavior under fire conditions, risk mitigation strategies, sensor technologies, and evacuation modeling. These clusters demonstrate the structural organization of the field and signal emerging interdisciplinary territories, such as the integration of artificial intelligence and GIS into real-time fire response planning.

To deepen the analysis, a keyword co-occurrence network was constructed based on term co-appearance frequency. Figure 9 visualizes the conceptual proximity between keywords using node centrality and link strength. Core nodes such as “fire safety,” “performance,” “design,” and “risk management” are the most prominent, indicating their foundational role in the literature. Technical keywords like “flame spread,” “combustion,” and “heat release rate” are densely connected to engineering-centric publications, while “evacuation,” “compartment fire,” and “building fire” reflect structural and spatial dynamics within fire scenarios. The emergence of recent keywords, colored in yellow and green to represent studies from 2022 to 2024, signals a paradigmatic shift. Terms such as “predictive modeling,” “deep learning,” “artificial intelligence,” and “GIS” point to a growing digital transformation in fire safety research. These are often linked to data-driven risk prediction, real-time evacuation planning, and adaptive sensor networks.

In addition, terms like “health,” “burn injuries,” and “air pollution” highlight an expanding concern with human safety and environmental consequences. Their increasing frequency and network centrality suggest that fire safety is evolving from a purely technical field to a more holistic, systems-oriented discipline. From a scientometric perspective, articles that incorporate next-generation technologies, particularly those related to AI, smart sensors, and digital twins, tend to attract higher citation counts. This indicates not only their methodological novelty but also their perceived utility in addressing complex fire safety challenges across diverse building typologies.

Taken together, the results demonstrate a clear conceptual shift in the fire safety literature, from narrowly defined engineering concerns to broader, integrated research agendas. The convergence of computational tools, environmental health considerations, and resilience-oriented design underscores the increasing complexity of built environments and the multidisciplinary demands of effective fire safety strategies.

3.3. Building Typologies in Fire Safety Research

To address Research Questions 2 and 3 (Q2 and Q3), this section examines how different building typologies are represented in the fire safety literature and how these typologies correlate with the inclusion of core safety systems, namely, detection, suppression, and evacuation mechanisms. The findings not only reveal thematic concentration but also identify gaps in the functional integration of fire safety strategies across various architectural contexts. Figure 10 illustrates the distribution of building types investigated in fire safety publications between 2010 and 2024. General or mixed-use buildings (31 studies) dominate the dataset, reflecting an increasing emphasis on hybrid spatial scenarios that defy conventional prescriptive codes. These settings often involve overlapping occupancies, necessitating integrated evacuation and suppression strategies tailored to complex human behaviors and usage patterns.

High-rise buildings (18 studies) constitute the second most studied typology. Their prevalence stems from the heightened fire risk associated with vertical egress complexity, delayed evacuation, and pressure-driven smoke behavior. Fire safety research in this context often focuses on performance-based designs, evacuation simulation, and the role of ventilation systems. Residential structures (14 studies) and heritage or cultural buildings (12 studies) also emerge as key typologies. Research on residential buildings frequently addresses early detection, occupant risk perception, and domestic evacuation procedures. In contrast, studies on heritage buildings emphasize the balance between structural preservation and the retrofitting of modern fire protection technologies—a challenge compounded by material sensitivities and architectural constraints. Healthcare facilities, educational institutions, and industrial complexes are notably underrepresented in the dataset, despite their high occupancy densities and unique fire risk profiles. This thematic asymmetry suggests a significant research gap in addressing evacuation challenges and system redundancies in vulnerable building functions.

Figure 11 provides a temporal and typological overview of fire safety system integration across the building categories. It maps the presence or absence of three core components: detection systems, suppression systems, and evacuation routes. While evacuation mechanisms are consistently addressed across most building types, particularly in high-rise and general-use facilities, detection and suppression systems are often either partially covered or omitted altogether in published studies.

This gap is particularly acute in heritage and general-purpose buildings, where fire detection systems are rarely analyzed with the same depth as evacuation scenarios. The lack of uniformity in the treatment of safety systems across building types highlights a methodological imbalance in the literature, one that privileges human movement over technical infrastructure evaluation. The data also indicates that post-2020 publications are more likely to incorporate multi-system frameworks, often in conjunction with simulation tools and sensor technologies. However, despite these advancements, a significant portion of the literature still treats fire safety systems as isolated elements rather than integrated subsystems tailored to specific building functions. While certain building types have been extensively studied in terms of egress modeling and occupant behavior, there remains a notable deficiency in the comprehensive evaluation of detection and suppression systems, particularly in functionally complex or historically sensitive environments. Future research should aim to close these gaps by adopting typology-specific frameworks that holistically assess fire risk, operational feasibility, and retrofit compatibility.

3.4. Tools and Technological Classifications in Building Fire Safety Research

To address Research Q4 and Q5, this section provides a systematic classification of technological tools and methodological approaches employed in fire safety research from 2010 to 2024. Based on the bibliometric mapping of 83 selected studies, five core categories were identified: simulation tools, BIM-based platforms, risk analysis and optimization methods, experimental validation techniques, and decision-support frameworks. This typology is visualized in Figure 12.

The first category, Simulation and Software Tools, forms the computational backbone of modern fire safety analysis. Tools such as Computational Fluid Dynamics (CFD), Fire Dynamics Simulator (FDS), PyroSim, Pathfinder, and Autodesk Revit are extensively used to model flame propagation, smoke behavior, evacuation scenarios, and structural response under fire loads. FDS and PyroSim, in particular, are dominant in studies focusing on high-rise and heritage buildings, where prescriptive codes fall short. Pathfinder, an agent-based evacuation simulator, is often used in scenarios involving complex human behavior under duress. CFD and HVAC simulation tools have also been employed for analyzing ventilation-driven fire dynamics and stratification effects. These simulation platforms enable high-fidelity analyses but are often siloed from empirical validation or integrated design frameworks.

The second category, BIM-Based Applications, includes Building Information Modeling tools that support spatial integration, fire sensor layout planning, and data-driven visualization of fire scenarios. Revit-based systems have been increasingly used in conjunction with FDS or PyroSim to build composite digital environments for fire strategy simulations. Recent studies have combined BIM with neural network models to enhance real-time hazard detection and response optimization. However, full interoperability between BIM and real-time sensor feedback remains limited, suggesting a technological gap in smart building integration.

The third category comprises Data Analysis, Optimization, and Risk Assessment Methods. These include statistical models (e.g., logistic regression, Markov chains), artificial intelligence algorithms (e.g., deep learning, decision trees), and heuristic optimizers such as Particle Swarm Optimization or Ant Colony Optimization. Risk metrics like CEPI (fire hazard scalar index) and CFO (consequence-based fire intensity models) are particularly used in infrastructure with low regulatory coverage, such as heritage sites and informal settlements. These methods provide predictive capability but often lack calibration against empirical or full-scale scenarios.

Experimental and Sensor-Based Tools form the fourth category. These include cone calorimeters, flame detectors, thermal cameras, and full-scale tunnel tests. Used mainly to validate material flammability, thermal insulation performance, or fire suppression systems, such tools provide the empirical backbone of fire engineering. Despite their utility, these experiments are rarely looped back into simulation environments for iterative model refinement.

The fifth category, Decision-Support Tools and Multi-Criteria Frameworks, reflects the growing need for systematic prioritization in fire risk management. Methods like the Analytic Hierarchy Process (AHP), Delphi method, and Failure Modes and Effects Analysis (FMEA) are applied to guide policy formulation, hazard anticipation, and resource allocation. However, their integration into automated or real-time response systems is minimal, limiting their practical utility in dynamic environments.

In addition, Figure 13 presents an integrated classification framework linking building typologies with the employed tools. High-rise and heritage buildings are disproportionately studied using simulation and BIM-based models, while general-purpose buildings are more commonly associated with decision-making and heuristic methods. Detection and suppression systems, however, are underrepresented in experimental or optimization-based research, indicating a bias toward evacuation-focused modeling in tool development. Moreover, studies applying AI techniques show higher citation averages, reflecting a trend toward algorithmic risk analysis and digital twin applications.

Together, these findings reveal both methodological progress and systemic fragmentation. While the toolbox of fire safety research has expanded considerably, true integration, across physical testing, digital simulation, and decision-making platforms, remains an aspirational goal. Future studies should aim to develop interoperable ecosystems that adapt tools to building function, occupancy profile, and regional regulatory contexts.

The results presented in the preceding sections provide a multi-dimensional overview of fire safety research over the past fifteen years, revealing significant thematic trends, typological biases, and methodological innovations. Bibliometric analyses have mapped the field’s intellectual structure; keyword clusters have illustrated its conceptual evolution; and tool-based classifications have shed light on technological diversity and fragmentation. These empirical findings now form the basis for critical reflection. The following section discusses their broader implications, evaluates the strengths and limitations of current approaches, and proposes strategic directions for future fire safety research in the built environment.

4. Discussion

The preceding analyses offer a nuanced view of how fire safety research has evolved over the last fifteen years, revealing both progress and persistent gaps in building typology focus, methodological integration, and technological adoption. In this section, we critically evaluate these findings to understand their implications for the future of fire safety science.

4.1. Dominance of Engineering and Simulation-Focused Paradigms

As demonstrated in Figure 3, Figure 4, Figure 5, Figure 6 and Figure 7, the field of fire safety is still heavily dominated by engineering-oriented studies, with simulation tools such as FDS, CFD, Pathfinder, and PyroSim serving as the primary analytical instruments [15,17,19,33,53]. While these tools enable high-resolution modeling of fire dynamics and human evacuation, they often operate in isolation, with limited integration into real-world validation pipelines or decision-support systems. For example, although BIM environments (e.g., Autodesk Revit) are increasingly paired with these tools to support spatial fire planning [18,35], interoperability remains underdeveloped. Moreover, AI-enhanced models (e.g., BP neural networks and real-time detection systems) have shown promise [22,36] but are still mostly confined to prototype scenarios without broader application.

4.2. Typological Asymmetry and Functional Neglect

Figure 11 and Figure 12 highlight a thematic bias in the literature, where high-rise and general-purpose buildings dominate research attention, while critical facilities such as hospitals, schools, and industrial buildings remain underrepresented. Even within studied typologies, most publications focus predominantly on evacuation scenarios, often neglecting the analysis of detection and suppression systems [25,54]. This imbalance reflects a narrow emphasis on occupant movement rather than integrated systems thinking. For instance, in heritage buildings, fire mitigation strategies often remain conceptual rather than operational due to challenges in retrofitting [10,11], leaving a substantial gap in risk-informed architectural conservation.

4.3. Fragmentation in Technological Integration

Despite the wide range of available tools and methods illustrated in Figure 13 and Figure 14, technological integration across platforms remains fragmented. BIM-based models are often detached from real-time sensor data [16], and simulation environments rarely incorporate empirical data for calibration or feedback loops. Experimental techniques such as full-scale tunnel tests or thermal sensor arrays [29,43] provide valuable physical validation yet are seldom reintegrated into digital modeling workflows. Likewise, decision-support systems like AHP, FMEA, and Delphi methods are applied independently of computational models or sensor-driven systems, limiting their capacity for adaptive real-time risk management [7,34,41].

4.4. Scientometric Shifts and Conceptual Transitions

The keyword network presented in Figure 9 and Figure 10 points to an emergent conceptual shift within the literature, from narrowly defined engineering concerns to broader systems-oriented approaches. Recent publications increasingly incorporate terms related to environmental health, post-fire impacts, and predictive technologies, such as “deep learning,” “burn injuries,” and “air pollution” [23,39]. This evolution is also reflected in the growing use of AI-enhanced risk modeling and big-data analytics. However, these developments are not yet supported by standardized methodological frameworks, and many studies remain descriptive rather than operational. Thorp et al., for example, demonstrates the potential of qualitative data analysis tools like MAXQDA for stakeholder engagement, yet such tools are rarely employed in tandem with simulation or optimization algorithms [7].

4.5. Toward Typology-Specific and Context-Sensitive Frameworks

The cumulative findings highlight the need for adaptive and context-sensitive fire safety frameworks tailored to specific building types and regional constraints. Studies have shown that applying generic models across typologies and geographic zones leads to misaligned system designs and regulatory inefficiencies [11,18]. For example, heritage structures require fire safety interventions that account for material sensitivity and spatial constraints [19], while industrial facilities demand real-time hazard detection integrated with ventilation controls [43]. This study responds to these challenges by introducing a novel typology-function-tool matrix that provides a structured basis for identifying functional gaps and technology misalignments in existing research. The framework goes beyond previous literature by categorizing fire safety research not only thematically or methodologically, but also according to its applicability across different building functions. This enables a more precise alignment between research focus and practical safety needs.

Crucially, the study serves as a bridge between academic modeling and applied practice. The proposed framework can be used by practitioners (such as architects, safety engineers, and code developers) to evaluate whether fire safety strategies are appropriately tailored to the building’s occupancy type and risk profile. It also provides a foundation for performance-based code development and policy reform, especially in regions facing rapid urbanization or regulatory fragmentation. Despite growing technical capabilities, the development of fully integrated platforms remains limited. Perera et al. emphasizes the potential of interoperable BIM-AI hybrids, yet practical deployment is still rare [35]. This study’s emphasis on context-aware and typology-sensitive strategies therefore addresses both theoretical gaps and real-world needs, underscoring its dual contribution to scholarship and practice. Unlike previous reviews that often focus on isolated technologies or generic typological classifications, this research is innovative in its integrated approach that simultaneously maps building types, safety functions, and applied tools through a comparative matrix.

To operationalize the study’s integrative contribution, Figure 14 presents a conceptual roadmap that synthesizes observed gaps and proposed future directions. Unlike previous summaries or keyword clusters, this figure links typology-specific deficiencies (such as fragmented tool usage, evacuation bias, and the neglect of vulnerable building types) with targeted strategies, including real-time integrated platforms, risk-sensitive regulatory frameworks, and performance-based design approaches. By aligning these elements, the figure supports both scholarly analysis and practical application, serving as a bridge between empirical insight and actionable reform in fire safety planning. Overall, the roadmap in Figure 14 serves not only as a diagnostic summary but also as a practical planning tool that connects systemic insights to implementation-oriented solutions.

5. Conclusions

This study provides a comprehensive scientometric and typological review of fire safety research in the built environment between 2010 and 2024. Based on a structured analysis of 83 peer-reviewed journal articles, the study identifies critical patterns, functional imbalances, and methodological trends in the field.

The key findings are summarized as follows:

• Research in fire safety remains heavily concentrated on evacuation modeling, particularly in high-rise and general-purpose buildings. In contrast, detection and suppression systems are significantly less examined, indicating a limited functional scope in much of the literature.
• Typological representation in existing studies is uneven. Vulnerable building types such as healthcare facilities, heritage structures, and informal settlements are notably underrepresented, despite their complex fire risk profiles and operational constraints.
• Technological tools such as BIM, simulation platforms, and AI-based systems are increasingly applied, yet their implementation is often fragmented and lacks integration across safety functions. Interoperability and real-time responsiveness remain largely absent.

The study proposes a typology-sensitive analytical framework that links building functions with fire safety strategies and methodological tools. This framework allows for a more precise identification of gaps between research focus and practical needs. These findings underscore the need for a shift toward integrated, data-informed, and context-sensitive fire safety approaches. Future research should prioritize typologies that are currently underserved, strengthen the alignment between safety functions and building use, and support the development of interoperable systems capable of addressing real-world complexity in fire risk management.