1 Introduction
1.1 General background
Many cities in developing countries have been undergoing rapid development and transformation. Against the backdrop of urban renewal trends, the conservation and revitalization of urban historical heritage has emerged as a critical issue. In this context, adaptive reuse of historic districts and buildings not only mitigates the impacts on urban development but also enhances cultural identity and community well-being (
Throsby, 2016).
This study focuses on the renovation and adaptive reuse of historical buildings, aiming to explore quantitative methods for evaluating the interventions and supporting decision-making processes. In recent research on historical building renovation, scholars across multiple disciplines have developed quantitative evaluation frameworks and decision-support tools (
Amer et al., 2024;
Arbulu et al., 2024;
Buda et al., 2022;
Husiev et al., 2023;
Katunský et al., 2024;
Knyziak et al., 2025;
Kyritsi et al., 2025;
Li and Qin, 2022;
Milić et al., 2019;
Milovanović et al., 2022;
Sun et al., 2022;
Taherkhani et al., 2021;
Teso et al., 2022;
Zhao et al., 2024,
2025). Among these, studies addressing energy performance optimization strategies have garnered particular attention from the researchers (
Arbulu et al., 2024;
Buda et al., 2022;
Husiev et al., 2023;
Katunský et al., 2024;
Kyritsi et al., 2025;
Milić et al., 2019;
Milovanović et al., 2022;
Teso et al., 2022). For example,
Arbulu et al. (2024) proposed a parametric simulation tool for environmental-economic assessment of energy-efficient renovation strategies in residential buildings;
Buda et al. (2022) evaluated and selected computer-based tools for energy performance retrofits in historical buildings;
Husiev et al. (2023) analyzed renovation potential through cost-effective assessment of energy-related intervention measures. Other notable contributions include:
Zhao et al. (2025), who employed deep learning to assess visual quality in commercial façade renovations of Chinese traditional architecture;
Sun et al. (2022), who utilized generative adversarial networks to automate façade design generation for historic urban renewal projects; Studies focusing on structural safety and disaster prevention assessments (
Amer et al., 2024;
Knyziak et al., 2025;
Li and Qin, 2022); researches emphasizing sustainability and resilience in renovation practices (
Taherkhani et al., 2021;
Zhao et al., 2024). While these studies provided valuable paradigms for specific aspects of building renovation, few had provided effective supporting evaluation tools for functional and spatial renovation, which is the focus of this study in the field of adaptive reuse.
For adaptive reuse, computational design is likewise a recent research trend of interest.
Eloy and Duarte (2015) developed a tool for transforming existing houses based on transformation grammar and evaluated the proposed design using Space Syntax.
Eloy and Guerreiro (2016) focused on transforming housing types by using grammar-based approach complemented by Space Syntax measurements.
Paulino et al. (2024) proposed a new computational method for exploring modifications to structural wall layouts during adaptive reuse of historic buildings.
Paulino et al. (2023) developed a shape grammar to specify a program generation framework for adaptive reuse of historic buildings for social housing with natural light as a prioritized spatial strategy.
Guerritore and Duarte (2018),
Belčič and Eloy (2023) similarly focused on the use of computational tools for adaptive modifications. Most of them focused on generative design and structure-related parameters, and some of them (
Eloy and Duarte, 2015;
Eloy and Guerreiro, 2016) used Space Syntax as an evaluation tool after presenting the proposed plans. As an effective tool for quantitative analysis of spatial features, the relevant variables and parameters in Space Syntax (
Hillier and Hanson, 1984) have the potential to provide a plausible theoretical basis for computational design and new parametric starting points for the solution generation process.
Space Syntax has increasingly been applied in innovative ways to study cultural heritage sites and buildings.
Dawes et al. (2022) applied Space Syntax with weighted axial line variants to analyze architectural works by Richard Neutra, validating hypotheses about spatial arrangement theories.
Li et al. (2016) proposed a quantitative framework based on Space Syntax to describe street network integration, enabling large-scale analyses that support spatial management for tourism and heritage conservation.
Eldiasty et al. (2021) integrated Space Syntax with UNESCO’s TOPSIS method to develop and validate an approach for assessing and preserving urban historical commercial heritage.
Zhang et al. (2023) combined Space Syntax with fire risk POI data to identify high-risk zones in historic streets, proposing targeted conservation strategies.
Saraoui et al. (2024) merged Space Syntax with daylight luminance simulations to recommend adaptive reuse strategies for converting heritage buildings into museums.
Fadakari and Andaroodi (2024) utilized Space Syntax to decode universal spatial configurations of the religious heritage layouts, guiding restoration practices.
Zhou and Wang (2023) correlated Space Syntax parameters with remote sensing data to evaluate urban vitality in Macau’s historic center, informing conservation and development policies.
Hou and Zheng (2017) predicted pedestrian flow via Space Syntax to propose renovations for traditional commercial streets.
Wang et al. (2022) employed multivariate analyses, including Space Syntax, to explore tourism business distribution patterns in Pingyao historic city, offering insights for heritage town management.
Abdulabbas Hammoodi and Shuker Al-Hinkawi (2023) measured the spatial coherence of religious heritage using Space Syntax, integrating local organizational features to advise regional integration. These studies collectively demonstrated Space Syntax’s practical potential to address diverse challenges in heritage conservation, as well as its methodological advancement potential.
Karimi (2023) noted that for Space Syntax, the emergence of new challenges necessitates the continuous development of innovative methods and tools as an inherent and ongoing aspect of the field’s evolution. Based on Hillier’s configurational theory, some of the recent studies have introduced innovative variants of Space Syntax or integrated it with complementary analytical approaches to address specific issues in heritage conservation (
Dawes et al., 2022;
Eldiasty et al., 2021;
Zhang et al., 2023). However, these contributions still exhibit limitations:
1) Current Space Syntax applications in heritage conservation predominantly utilize parameters to describe existing spatial configurations, offering limited actionable insights for conservation or renovation decision-making. The guidance provided remains indirect and lacks direct applicability to practical interventions.
2) Most Space Syntax-based studies focus on urban-scale heritage systems, with insufficient attention to building-scale spatial renovations.
To address these limitations, this study seeks to develop a novel application paradigm that enhances Space Syntax’s utility in historical heritage conservation.
1.2 Case background
Within the context of urban historical heritage conservation, this study narrows its focus to a specific subject. In Chinese cities with long developmental histories, such as Beijing, Shanghai, and Wuhan, urban cores often encompass culturally and historically significant districts. Taking Wuhan as an example, local governments have designated the former concession area in Hankou―a zone rich in concession history and cultural heritage―as a special “Historical and Cultural Area.”
Among these heritage structures, Lifen (traditional Wuhan alleyway dwellings) stand out as a housing typology. In 1870, the early form of lane houses emerged in the British Concession of Shanghai. The overall layout was in the form of terraced houses, and the floor plan design of the residential units took into account the characteristics of traditional Chinese three-courtyard or quadrangle courtyards. In 1861, when Hankou was opened as a trading port, real estate developers from Shanghai came to invest in Hankou. They purchased land in the French Concession near the railway station and built the first batch of lane houses in Hankou, marking the origin of the Lifen houses in Hankou. The Lifen houses then rapidly became the main type of residence in Hankou at that time. Their spatial organization and archetypal layouts reflect the unique historical culture and lifestyle of Hankou’s historic neighborhoods. Wuhan’s Lifen complexes, varying in heritage value and quantity, have undergone varying degrees of renovation or demolition today:
1) Adaptive Reuse for Commercial/Public Functions: Examples include Tongxingli, Taixingli, and Xian’anfang, transformed into vibrant cultural and commercial hubs;
2) Preservation of Residential Use: Examples include Futangli, Changnianli, and Xinchengli retain their original residential character;
3) Abandonment or Partial Demolition: Examples such as Kunhouli and Yanqingli remain vacant or semidemolished;
4) Complete Replacement by Modern Developments: By the early 21st century, nearly half of the original 210 Lifen complexes had been demolished and replaced with contemporary buildings, leaving only approximately 100 surviving today (
Li, 2024)
For vacated Lifen communities awaiting renovation, such as Kunhouli and Yanqingli, where original residents have relocated, local governments and developers face uncertainties in balancing conservation priorities with functional revitalization. Pilot initiatives, including collaborative projects with academic institutions, have been implemented to explore innovative reuse strategies and gather community insights. Overall, the conservation and renovation decision-making processes for Lifen and other urban heritage clusters within Wuhan’s historic core remain fragmented and ad hoc, reflecting both research gaps and untapped value in systematic intervention strategies.
Adaptive reuse of Lifen is usually a shift from residential to public functions. For instance, Wuhan’s 13th Five-Year Plan proposed replacing residential functions in some former concession areas with recreational, commercial, and tourism uses (
Li, 2024). Consequently, Lifen renovations should balance retaining traditional elements with adapting spatial layouts to accommodate new public functions―such as shops, studios, or galleries―that demand distinct spatial characteristics and structural flexibility compared to residential spaces. For the future use after renovation, the specific spatial functions with uncertainties require that the renovated space have the adaptability to public functions in a general sense.
Song and He (2017), in their study of Hankou’s former British Concession, observed that large-scale demolition and reconstruction are frequently prioritized in China’s historical district conservation practices, leading to irreversible erosion of urban texture typologies and the loss of historic contexts around protected buildings. In response,
Song and He (2017) proposed a “minimal intervention” framework at the neighborhood scale, piloting its application in Hankou’s historic districts to advocate for a pre-cautionary approach in China’s heritage rehabilitation efforts. In the debates on heritage reuse, this concept also, to a certain extent, supports the hot topics of enhancing sustainability and reducing carbon emission.
2 Research aim
The core innovation point of the research focuses on the technical optimization of Space Syntax application. Meanwhile, taking the traditional Lifen complex of Hanrunli in Hankou’s Historical and Cultural Area, Wuhan, as an example, it aims to contribute to the adaptive reuse research of urban architectural heritage. The study inherits the concept of “minimal intervention” in heritage renovation and aims to explore a precise and quantitative renovation assessment and guidance tool to balance the preservation and functional renewal of architectural heritage. Grounded in the analytical framework of Space Syntax and facing the specific problem of spatial and functional adaptive reuse, the research proposes a quantitative method to identify “Renovation Potential Values” for specific spatial components within Lifen floor plans, targeting critical intervention points. It further validates and discusses the feasibility of this approach across diverse residential layouts and hypothetical renovation scenarios, aiming to provide effective quantitative evaluation and decision-support strategies, providing novel perspectives for spatial renovations of Lifen and similar heritage buildings. The research framework is illustrated in Fig. 1.
Under the overarching goals of revitalizing urban historical heritage and promoting sustainable development in historic districts, this study pursues two primary objectives:
1) Practical Problem-Solving: To address specific functional and spatial challenges in heritage adaptive reuse practices through actionable insights.
2) Methodological Innovation: To provide new ideas and attempts for the application mode of Space Syntax, to shift from describing the existing situation to quantitatively guiding decision-making.
3 Methods
3.1 Case of the study: overview of Hanrunli
The representative case study selected is the floor plan of Hanrunli, with reference materials drawn from prior research by local scholars (
Li and Sun, 2008). Hanrunli is situated in the core area of Hankou’s Historical and Cultural Area (Fig. 2). As a Lifen residential complex developed during Hankou’s modern real estate boom, Hanrunli exhibits typological features emblematic of its era: a linear spatial layout, hierarchical alleyway networks, and standardized two-bay and three-bay residential units. Its architectural heritage significance stems from these prototypical characteristics, while its unique plot boundaries and spatial configurations provide atypical unit types for analytical discussion.
Li and Sun (2008) stated: “With the revitalization of Hankou’s old city part, Hanrunli and its historic context can be preserved as an exemplary Lifen cluster through rational renovations of individual buildings and the overall environment, extending the functional continuity of the under-construction Wuhan Art Museum and enhancing surrounding commercial ecosystems.” This aligns closely with this study’s adaptive reuse objectives.
The research focuses on Hanrunli’s ground-floor plan and adjacent streets, encompassing:
Experiment 1: Experimental analyses of three three-bay unit layouts;
Experiment 2: Three two-bay unit layouts;
Experiment 3: Hypothetical renovation scenarios combining three-bay and two-bay configurations;
Experiment 4: One representative special unit layout.
3.2 Theoretical basis: Space syntax
Space syntax is title given to a set of methods which provide a means of developing mathematical measures of the underlying spatial structures of a plan, rather than its physical dimensions (
Hillier and Hanson, 1984). Since its development in the 1980s, Space Syntax has been widely applied to urban spatial analysis (
Hillier, 2007) and architectural studies (
Hanson, 2003). This methodology abstracts spaces into mathematical graphs and analyzes their topological relationships to reveal various characteristics of the underlying spatial structures of plans.
Common analytical methods of Space Syntax include convex space analysis, axial line analysis, and Visual Graph Analysis (VGA), etc. Among them, the first two methods related to topological analysis and Visual Graph Analysis have different application scenarios. Topological analysis is more closely related to spatial configuration and functional essence, while Visual Graph Analysis is often used to analyze spatial visibility and human natural movements (
Karimi, 2023). Topological analysis-related methods of Space Syntax and the data indicators they generate can be used to provide specific insights into the spatial and social properties of rooms in an architectural plan, as previous researches showed (
Dawes et al., 2021,
2022;
Fadakari and Andaroodi, 2024). This research also focuses on the configuration and functional characteristics of space, and thus the main method is concentrated on the topological analysis of Space Syntax. Taking into account the suggestions on the functional transformation of relevant architectural heritage in the official planning of Wuhan City, in the renovation assessment of the Lifen, under the focus points and renovation purposes defined in this research, Space Syntax quantitatively analyzes and indicates the spatial structure closely related to the spatial characteristics and functional usage before and after the renovation. Therefore, Space Syntax is suitable to serve as the theoretical basis of this article.
Integration Value (IV) is a global metric derived by calculating the shortest distances between each pair of graph nodes, and it is one of the most widely used measures in Space Syntax research. It quantifies the centrality of a space within the graph structure, with higher values indicating that a node occupies a more central position than those with lower values (
Dawes et al., 2021). To provide a concise explanation for the calculation formula of Integration Value, the straightforward formula mentioned by
Asif et al. (2018) is cited as follows:
This study employs a version of convex space analysis. Functional spaces within buildings are represented as nodes in an aligned floor plan, and relationships between these spaces are translated into nodes’ topological connections. Software such as UCL DepthmapX (
Turner, 2004) allows manual input of graphs and connections while automatically calculating metric results (e.g., Integration Value, Connectivity Value, Depth Value). This study used and respected the metric data obtained from Space Syntax measurements in DepthmapX software.
3.3 Integration value and adaptive reuse
This research adopts Space Syntax’s topological analysis as its theoretical foundation, and it is necessary to explain how the Space Syntax variables can assess the results of adaptive renovation.
The results of the Lifen adaptive reuse focused on in this study, taking into account the suggestions of the Wuhan’s 13th Five-Year Plan mentioned earlier, are defined as the transformation from residential functions to non-specific public functions. Thus, the adaptability to non-specific public functions after the renovation is defined as the main renovation requirement in this research. It is worth noting that in the fields of Space Syntax natural movement and spatial topology, the functional adaptability of space may be reflected by different variables and parameters. This research focuses on the discussion and reasoning of the demand for adaptability in the field of spatial topology.
Among spatial metrics derived from Space Syntax, Integration Value (IV) is prioritized due to its documented correlation with spatial characteristics and functionality. For example, at the urban scale,
Li et al. (2021) demonstrated that preserving or creating streets with higher IVs enhances district vitality in historic neighborhood revitalization. At the architectural scale,
Maina (2013) mentioned that spaces hosting intensive social functions and activities often exhibit higher IVs. Hillier’s book on the theoretical basis of Space Syntax (
Hillier, 2007) identified Integration Value as “one of the most marked types of differentiation between spaces”. Furthermore, regarding the overall spatial system at the architectural scale, the book discussed that to a certain extent, the form of minimizing the total Depth of the system (and correspondingly maximizing the total Integration Value) is more popular because of its flexible configuration and the ability to adapt to various functional modes. Overall, in the building, relatively larger IVs of each space and the total IV of the whole spatial system enables the corresponding spaces to have better adaptability to diverse public functions, which is consistent with the renovation requirements focused on in this research.
Based on Space Syntax methods and the demand for adaptive reuse, the study establishes two research priorities for results and discussion:
1) Enhancing Post-Renovation Functional Adaptability: Mainly quantified through improvements in total IV of the spatial system. The changes in IVs of every space are also under discussion.
2) Optimizing Intervention Efficiency: Achieving maximal IV gains with minimal structural interventions (e.g., wall removals or spatial connectivity modifications), forming the basis for evaluation criteria.
3) Based on the discussion of relevant literature and theoretical reasoning, the value of the simulated data before and after the renovation lies in “verifying the method innovation in the Space Syntax context” rather than “verifying the usage state in the real world”.
These two characteristics together constitute the representative variable “Space Renovation Efficiency Index (EI)” for evaluating the renovation effect in the following text.
3.4 Proposal of key methods
This study expands the application mode of Space Syntax variables to investigate spatial renovations in Hanrunli and similar Lifen complexes. The research objective is to innovate the application of Space Syntax, establishing a quantitative evaluation and decision-support paradigm for architectural heritage renovation. To achieve this, the following sections introduce novel conceptual frameworks and operational methodologies. A research method framework is illustrated in Fig. 3.
To bridge Space Syntax’s descriptive capabilities with renovation assessment, this work proposes a key innovative “Integration Value Assignment” procedure and identifies Modifiable Walls as critical intervention targets, translating spatial metrics to envelope structures. It is worth discussing that in the adaptive reuse about spaces, the decisions needed may focus on the walls rather than the space itself, as previous researches showed (
Eloy and Duarte, 2015;
Eloy and Guerreiro, 2016;
Paulino et al., 2023). Taking this into account, the quantitative description of spatial characteristics by Space Syntax, after being transmitted to the wall, may be particularly suitable for the exploration of the transformation from “describing the current state of space” to “guiding decision-making”.
Grounded in renovation demands and Space Syntax theory, the Space Renovation Efficiency Index (EI) is introduced as a key parameter for evaluating intervention outcomes. Furthermore, based on core demands, a pioneering variable concept―Space Renovation Potential Value (PV)― is defined and calculated.
PV serves as the pivotal parameter guiding renovation assessments and decisions. Derived exclusively from pre-intervention spatial characteristics, PV’s ability to reliably predict renovation effectiveness constitutes the core investigative focus of this study.
The following are the explanations of the key concepts and operations proposed in this article:
Explanation of “Integration Value Assignment”: This operation is intended to apply Space Syntax feature parameters from space to the envelope structure. In floor plan analysis, Space Syntax’s topological approach treats architectural spaces as nodes and spatial adjacencies as node connections to calculate IVs. This study assigns the IV of each spatial node to the corresponding wall segments bordering that space (to the wall side adjacent to the space), establishing Integration Value Assignment for each wall segment.
Definition of “Modifiable Walls”: To focus on the technical optimization exploration of Space Syntax application, in the research the walls are screened in a specific context of “Space Syntax topological analysis”. Existing adjacency relationships between spaces are predefined in floor plans. Wall segments whose removal doesn’t change the original spatial connections are defined as Non-modifiable Walls; those whose removal changes spatial connections are then defined as Modifiable Walls, serving as intervention targets for analysis. It is worth noting that the structure of LiFen is in the form of brick-concrete or brick-wood. Except for the beam-column support structure, the brick walls can be broken through to a certain extent. In the actual renovation, issues such as structural load-bearing capacity and public facilities should also be taken into consideration. As this article focuses on the spatial configuration and functional usage before and after the renovation, specific load-bearing capacity and water or electricity facilities are not within the scope of discussion.
Definition of “Space Renovation Efficiency Index (EI)”: Grounded in minimal intervention principles, adaptive reuse requirements, and Space Syntax theory, EI quantitatively evaluates renovation effectiveness. It measures the cost-effectiveness of functional adaptability improvements through the ratio of total IV increase to the number of demolished walls. Meanwhile, additional parameters may be analyzed alongside EI for a comprehensive assessment.
Definition of “Space Renovation Potential Value (PV)”: As the key variable of this study, PV adopts a parsimonious calculation based on pre-intervention spatial characteristics and Space Syntax measurements. For Modifiable Walls bordered by two spaces within the renovation scope, PV is defined as the absolute difference between IVs of both sides.
3.5 Research process
1) Calculate the original Space Syntax IVs for all spaces in the case study floor plans.
2) Identify Modifiable Walls, perform IV Assignment for these walls, and calculate corresponding PVs.
3) Rank Modifiable Walls into intervention groups based on descending PVs. In this study, each group undergoes isolated “removal” of the wall or wall combinations with identical PVs. After that, recalculate IVs for the modified spatial configurations and record data.
4) Calculate the EI for each group using post-renovation IV data.
5) Investigate the mathematical relationship between PV and EI through linear regression analysis and statistical test. Conduct further analysis by correlating PVs with spatial features in the floor plan, discussing special cases and performing selective filtering.
6) For general situations, aggregate PV and EI data across multi-scale cases to explore universal correlations via linear regression and statistical test. For special situations, analyze individually using the impact of each space’s modifications, topological relationships between spaces, and Space Syntax justified plan graph (JPG) (
Hanson, 2003). Synthesize findings from general and special situations to evaluate PV’s predictive validity and applicable scope.
4 Results and discussion
4.1 Results
4.1.1 An overview of the key findings
1) The proposed PV calculated for the typical Lifen unit types and the general cases is overall good predictor of the renovation effect represented by the EI (R2 = 0.759, p < 0.01).
2) The proposed PV can also reflect the improvement of each space’s Integration Value in the typical Lifen unit types. The improvement of each space’s IV (functional adaptability) is more uniform and significant when the PV is relatively large.
3) In the typical Lifen unit types, special situations include the opening up of walls between the units, which will significantly increase the overall sum of Integration Values.
4) The proposed PV cannot predict the atypical Lifen unit type.
4.1.2 Experiment 1
The objective of Experiment 1 is to explore and identify initial patterns. In the preliminary analysis of three three-bay unit layouts, the test floor plans are replicated from Hanrunli’s typical three-bay prototypes under hypothetical spatial arrangements. Key results are presented in Fig. 4, including original Space Syntax IVs, PVs of Modifiable Walls, and the main statistical results of the renovation effects. The PV-EI data derived from renovation workflows (group renovation and subsequent new Space Syntax measurements) are summarized in Table 1. The data indicates an overall trend of diminishing EIs as PVs decrease. However, renovation Group 14 (PV = 0.012) and Group 15 (PV = 0) exhibit reversal trends contradictory to overall observations.
To further explore the predictive effect of PV on the renovation effect, the study then conducts linear regression analysis and F-test (Table 2) and draws the scatter plot of the original data of PVs and EIs in Experiment 1. The analysis yields an initial R2 = 0.103, p > 0.1, the linear regression model is statistically insignificant, indicating weak overall correlation. Some typical wall groups and their corresponding renovation effects are marked in the figure (including a typical high PV group, a typical low PV group, and some special groups with data outliers). However, after removing outlier data from Group 14 and Group 15 renovations, the filtered scatter plot exhibits a stronger linear relationship with R2 = 0.786, p < 0.01, the linear regression model is statistically significant, which means that when excluding these special cases, PV can predict the overall renovation effect relatively well in this experimental plan. Further analysis identifies outlier characteristics in the PV-EI data of Group 8 (PV = 0.238), Group 14, and Group 15, corresponding to wall segments at junctions between units in the floor plan.
To further understand the reason why breaking through the walls at the connection points of the units leads to a significant improvement in the overall IV, a further spatial analysis will be conducted next. For renovation groups corresponding to special situations in the data, Group 14 (PV = 0.012) serves as an example. Space 21 is selected as the starting point for the JPG discussion (Fig. 5), as its associated walls correspond to distinct outcomes in Group 13 and Group 14 renovations. The spatial connection diagram reveals that the original layout comprises three primary structural clusters with functional branches, consistent with Lifen’s original spatial programming. Group 13 connects branched spaces within individual units, while Group 14 links three units. The connection diagram demonstrates that Group 13 enhances local spatial connections, while Group 14 strengthens global connectivity, suggesting heightened inter-system interactions. The JPG analysis also shows that Group 13 primarily improves local connectivity, while Group 14 not only enhances local connectivity but also reduces the system’s maximum depth and alters spatial composition across depth levels, indicating more substantial modifications to inter-spatial relationships. Due to their small proportion and distinct spatial signatures, these special renovation groups targeting junction walls between units are filtered as special situations and analyzed separately in subsequent experiments.
Overall, Experiment 1 demonstrates a statistically significant linear relationship between Space Syntax-defined PVs and EIs, indicating that PV can partially predict renovation effects. This finding suggests a novel quantitative framework for adaptive reuse of Lifen heritage buildings. For instance, walls with higher PVs―such as Group 1 (PV = 0.787) and Group 2 (PV = 0.591)―can be prioritized for removal, potentially maximizing global IV improvements. Walls with comparable spatial roles, like Group 12 (PV = 0.063) and Group 13 (PV = 0.051), may have demolition priorities determined by relative PVs. However, data from Experiment 1 reveal that when PV differences are small, EI comparisons become ambiguous; conversely, larger PV differences yield clearer EI differences.
4.1.3 Experiment 2
Experiment 2 aims to supplement untested unit typologies in Experiment 1 and validate its observed patterns. Experiment 2 examines another typical two-bay unit layout from Hanrunli under real-world spatial arrangements. As in Experiment 1, original Space Syntax IVs, PVs and main renovation effects are annotated in Fig. 6, while PV-EI data are summarized in Table 3. Overall, the data align with Experiment 1’s trend: EIs decrease with declining PVs. Consistently, Group 8 (PV = 0) targeting junction walls between units exhibits disproportionately high EI.
Linear regression analysis and F-test (Table 4) of PV-EI data in Experiment 2 yield an initial R2 = 0.112, p > 0.1, the linear regression model is statistically insignificant, indicating weak correlation. Some typical wall groups and their corresponding renovation effects are marked in the figure (including a typical high PV group, a typical low PV group, and a special group with data outliers). After excluding Group 8 outlier data, the adjusted R2 = 0.835, p < 0.01, the linear regression model is statistically significant, which reflects a strong linear relationship. The PV-EI mathematical consistency between Experiments 1 and 2 is thus confirmed.
While Experiment 2 validates the PV-EI correlation under two-bay layouts, the limited number of renovation groups and PV-EI data points necessitate further experiments.
4.1.4 Experiment 3
Experiment 3 aims to further examine established patterns and special situations while evaluating the framework’s utility for evaluation and decision-making in hypothetical renovation scenarios. Experiment 3 analyzes actual spatial configurations of units in selected Hanrunli zones. Original IVs, PVs, and main renovation effects are annotated in Fig. 7. PV-EI data for general situations (after filtering out special situations) and special situations are separately tabulated in Table 5. General situation data adhere to prior trends―EIs decline with decreasing PVs―while special situation renovations again demonstrate anomalously high EIs.
Linear regression and F-test (Table 6) of PV-EI data (special situations excluded) yield R2 = 0.779, p < 0.01, the linear regression model is statistically significant, confirming strong correlation and alignment with Experiments 1–2. For special situations alone, regression analysis and F-test produce R2 = 0.272, p = 0.100, the linear regression model is statistically insignificant, indicating weak linearity and PV’s limited predictive capacity under these conditions.
Subsequently, data from Experiments 1, 2, and 3 (excluding special cases) are consolidated to examine the correlation between proposed PVs and EIs in different scales of typical Lifen units’ combinations. Following data integration, raw PV and EI data are summarized in Table 7. Linear regression analysis and F-test (Table 8) of the consolidated PV-EI data yield R2 = 0.759, p < 0.01, the linear regression model is statistically significant, indicating a strong linear relationship (Fig. 8). The integrated data demonstrate that, for Hanrunli’s typical floor plans (excluding special cases), a statistically significant linear correlation exists between Space Syntax-defined PVs and EIs. PV reliably predicts the effects of the renovations across varying scales and unit arrangements, confirming its general applicability.
To further validate PV’s evaluative role in adaptive spatial renovations and refine assessment methods for special situations, the study next supplements Experiment 3 with detailed inspections of the renovation effect of each space. The renovation effect here is defined as the difference between post- and pre-renovation IVs for each space, which reflects the increase in the adaptability to highly public and social activity functions. Spaces are numerically labeled to track effects (Fig. 9), with labeling sequences carrying no intrinsic significance.
Raw renovation effect data for all renovation groups are tabulated in Table 9.
Based on these data, spatial renovation effects are visualized using a color gradient scheme defined in Fig. 9. Aligning with prior discussions on general and special situations, the effects diagrams are divided into two groups in Fig. 10. The groups are ordered left-to-right and top-to-bottom according to descending PV magnitudes.
The renovation effect diagrams for general cases reveal three distinct patterns:
1) As PV magnitudes decrease, the spatial scope of effectiveness transitions from “global” to “localized”;
2) Overall effectiveness diminishes with declining PV;
3) Localized spaces demonstrating high effectiveness reduce with declining PV.
These patterns not only corroborate PV’s evaluative utility in general cases but also inform supplemental definitions of renovation effects.
Special situation diagrams exhibit no PV-dependent trends, reinforcing the need for separate analytical frameworks and delimiting PV’s applicable scope. Comparative analysis of spatial clusters between groups shows that special situations exhibit lower spatial uniformity in effectiveness, with targeted enhancements in distinct core areas resembling mid-to-late general situation patterns.
Cross-referenced with EI data, higher PV magnitudes (e.g., Group 1, 2, 3, 5, 6, 7 in Experiment 3) correspond to more uniform and pronounced IV improvements across spaces, while lower PV magnitudes yield isolated and weaker enhancements. Special situations demonstrate concentrated IV gains in specific zones. These findings provide novel references for real-world renovation effects evaluation.
4.1.5 Experiment 4
Experiment 4 investigates the applicability of established patterns to atypical unit typologies. A representative special floor plan from Hanrunli is analyzed, with original IVs, PVs, and main renovation effects annotated in Fig. 11. PV-EI data are tabulated in Table 10. While overall EIs are relatively high, PV-EI trends diverge significantly from previous experiments.
Linear regression analysis and F-test (Table 11) of PV-EI data yields R2 = 0.020, p > 0.1, the linear regression model is statistically insignificant, indicating no discernible linear relationship. Scatter points exhibit no discernible patterns between PVs and EIs. These results demonstrate PV’s inability to predict renovation effects in this atypical Hanrunli layout.
Experiment 4 further supplements the analysis by examining renovation effects across all groups. Spatial labeling is in Fig. 9, with raw effectiveness data tabulated in Table 12. Based on these data, spatial renovation effectiveness is visualized in Fig. 12 using the color gradient scheme defined in Fig. 9. The groups are ordered left-to-right and top-to-bottom according to descending PV magnitudes.
The renovation effects diagrams for Experiment 4 exhibit two key characteristics:
1) Effectiveness improvements consistently display “localized” spatial features;
2) As PV magnitudes decrease, effectiveness variations show no discernible patterns.
Cross-referenced with Experiment 3 data, Experiment 4’s EIs and spatial effectiveness align closely with special situation groups:
1) Despite declining PV, EIs remain relatively high with no clear trends;
2) PV exhibits no correlation with spatial effectiveness, yet overall effectiveness spatial uniformity remains low.
These findings indicate that PV’s application to special zones within typical layouts or entirely atypical layouts requires context-specific discussions based on renovation objectives.
4.2 Discussion
4.2.1 Guiding value for renovation evaluation and decision
For adaptive renovation scenarios represented by Experiment 3, the proposed Renovation Potential Value (PV) demonstrates guidance value. In Fig. 13, in the context of spatial public function adaptability, walls with darker color indicate higher intervention priorities, while dashed-line parts require case-specific discussions based on practical needs.
In general adaptive renovations, removing darker-colored walls yields greater improvements in overall spatial functional adaptability (e.g., in the renovation group, a single wall removal with PV = 1.162 increases total IV by 4.84%) and more uniform integration enhancements across spaces. Dashed-line parts, analyzed separately, significantly enhance inter-local spatial interactions and deliver high total IV gains. However, their effectiveness varies unevenly across spaces and risks altering original unit-based layout cultures and structural characteristics, necessitating demand-driven considerations in practice.
Li (2024) classified the redevelopment models of the Lifen community into 1) fully government-led; 2) partial government involvement 3) fully community-led. And regarding the renovation and revitalization of Lifen neighborhood, it was pointed out that community-led reconstruction, when it received external or political support but was still driven by the community, might have a more lasting and beneficial impact on the long-term development of the community. The paradigm in this research may contribute to such development feature.
For the government and the renovation contracting companies, the existence of relevant professional abilities may enable them to have sufficient ability to generate plans and solve problems when facing the renovation of Lifen. The PV calculation proposed in the article and the inferred priority model can supplement objective insights for subjective scheme design and decision-making from a quantitative and convenient perspective, to assist and promote the generation of renovation schemes.
For community and individual users, the paradigm proposed in this article is quantitative and straightforward. To a certain extent, it can help make up for the lack of relevant professional abilities, provide equal references for the renovation decisions of non-professional users, enhance the participation rights and intervention rights of them, and thereby respond to the topics of community-driven and social inclusion in the renovation process.
The priority model derived from the PV mentioned above is not the sole reference during the renovation process, but is recommended to be combined with the renovation assessment of other aspects of concern and local heritage policies. Although the Non-modifiable Walls defined in the context of Space Syntax topological analysis are not marked, in actual renovation, they are not unmodifiable. Instead, they should be subject to additional comprehensive assessment and decision-making in combination with the specific usage requirements of the space, as well as factors such as structural load-bearing capacity and public facilities.
4.2.2 Discussion on general situations
In experiments involving typical two-bay and three-bay unit combinations from Hanrunli, PV-EI data exhibit strong linear correlations for general situations. Experiment 3 further confirms that PV magnitudes reflect uniformity in renovation effectiveness. These findings demonstrate that the Space Syntax-based PV quantitatively captures relative performance in spatial functional adaptability, offering actionable insights for renovation evaluation and decision-making. In larger spatial systems, PV enables efficient prioritization of adaptive interventions.
In the research related to Lifen and Hankou blocks, Lifen is usually regarded as a type of architectural system that “has specific homogeneous architectural morphological characteristics”. With similar original unit plan spatial structures and block historical features (
Li et al., 2000;
Liu et al., 2023;
Maglioccola, 2018;
Zhou et al., 2010;
Zou et al., 2023), a recent study on thermal comfort in Lifen (
Xi et al., 2024) also selected plan cases similar to that in this paper as representatives of Lifen. Within the limited context of Space Syntax, the main spatial form of Lifen is presented as follows: The main axis is a spatial chain with front and back doors and courtyards, and rooms extend to both sides (or one side) from each spatial node of the main axis. Further, the units are arranged in a row to form buildings, and various building arrangements are generated according to the land use to form alleys and corresponding communities. Overall, in Wuhan most Lifen complexes share similar unit layouts with Hanrunli, suggesting the proposed framework’s broader applicability as a quantitative tool for adaptive reuse of Lifen heritage.
The results also highlight Space Syntax’s innovative potential in heritage conservation, including: Developing new methods tailored to practical needs; Addressing small-scale architectural renovations; Providing direct quantitative guidance for evaluation and decision-making.
Furthermore, the study offers novel references and perspectives for research and practice in both Space Syntax and heritage conservation fields.
4.2.3 Discussion on special situations
The proposed PV fails to directly assess special cases like “connecting two spatial chains.” Structurally, such modifications transform linear series-parallel configurations into significantly enhanced cyclic spatial relationships.
Dawes et al. (2021) suggested that this kind of spatial structure reflects “highly flexible spatial planning”. This study also argues that such changes notably improve inter-spatial interactions and public functionality, aligning with adaptive reuse objectives. Experimental data indicate relatively high overall effectiveness for these interventions, though localized spatial impacts necessitate case-specific evaluations based on usage needs. Then, subsequent assessments can leverage PV’s predictive utility in general situations for remaining spaces.
In atypical layouts, original spatial structures differ markedly from typical Lifen unit configurations. The former inherently exhibit cyclic relationships, demonstrating greater flexibility and publicness inconsistent with traditional Lifen spatial programming. PV-EI trends in atypical cases deviate from typical patterns, indicating PV’s limited evaluative utility for adaptive reuse in such contexts. Furthermore, the results suggest PV’s effectiveness may depend on original spatial configurations, warranting further research to delineate applicable scenarios. However, most Lifen complexes, apart from Hanrunli, share similar or identical spatial logic with typical cases, affirming the proposed framework’s practical relevance.
The renovation of special situations may change the original unit-based layout’s culture and structural characteristics. And atypical Lifen layouts may have different cultural genes from typical types. From the perspective of such social and cultural significance, the research by
Li et al. (2023) may provide new ideas for the subsequent development and application of the method proposed in this paper. For instance, computational analysis can be used to explain the influence of component details, spatial prototypes, and cultural symbols on spatial forms. Specifically, the related subsequent research ideas may include:
1) Semantic coding can be carried out on the traditional decorations of architectural heritage such as window carvings and brick carvings, and the completeness and retention ratio of these symbols before and after the renovation are analyzed. The “retention rate of cultural symbols” can be used as a supplementary indicator in addition to PV to further supplement the balanced dimension of “protection-renewal”.
2) Identifying spatial prototypes such as courtyards that carry traditional cultural symbols, and assess whether the renovation emphasizes or destroys these cultural symbols.
3) Mining the texts related to heritage such as Lifen in social media or traditional literature may help supplement the interpretation of the space usage derived from the communities and rituals or the expectations for the functions after the renovation. This may help protect the intangible cultural values and social memories that the spatial parameters have not been effectively identified during the renovation.
Overall, although in the context focused on in this article the results of adaptive reuse are mainly defined as spatial public function adaptability, and heritage spaces are regarded as functional entities, subsequent research should consider the impact of relevant cultural and community significance on the use and modification of the spaces to enhance the inclusiveness of relevant decision-making guidance.
4.2.4 Contributions and limitations of this study
This study focuses on the technical optimization of Space Syntax application. Taking the adaptive reuse of a Lifen heritage as an example, it centers on the specific demands of spatial public functional renovation, innovatively develops the application of Space Syntax, and provides new paradigms and ideas for the assessment and decision-making of architectural heritage renovation. Key contributions include:
1) Introducing a data-driven evaluation and decision-support paradigm for architectural heritage renovation, addressing research gaps in spatial renovations and providing solutions for Lifen heritage evaluation and decision-making;
2) Expanding Space Syntax application modes by proposing new parameter calculation and extending spatial metrics to envelope structures, shifting from describing the existing situation to guiding decisions, demonstrating its practical potential, and offering new ideas for methodological development;
3) Enriching Space Syntax’s application scenarios and paradigms in heritage conservation by focusing on architectural-scale interventions.
However, several limitations still exist:
1) The study primarily validated mathematical relationships between PV and EI, leaving room for parameter (PV) and model optimization;
2) While PV’s evaluative utility in spatial adaptability was demonstrated, Lifen renovations still have additional demands and focus points beyond spatial adaptation;
3) In this study, the innovative application of Space Syntax focuses on its analysis domain of topological relationships and configurations of space, and defines the evaluation of renovation effects. At present, the analysis of Space Syntax in the fields of vision and natural movement cannot be directly integrated into the PV model;
4) The workflow lacked development and integration of efficient digital tools.
4.2.5 A renovation example
In this renovation example for a partial plan of Experiment 3 (Fig. 14), decisions were made only using the paradigm proposed in this research. The wall sections with high Potential Values and those within special situations are preferentially opened. Meanwhile, a wall section with low potential values is also opened to organize the circulation. In the actual renovation, the specific location for opening a wall and the size of the opening should take into account multiple aspects such as windows, support structures, and the layout details of each space (The opening in the following text is tentatively set to be 1.5 m wide). The renovation function is set as part of the exhibition on the upper plan and part of the stores below. Using potential values to assist in renovation decisions, with only four walls being modified, the total IV increases by 9.12 %, and the IV of the renovated area increases by 14.15 %. Through the supplementary JPG analysis, it can be observed that, in a topological sense, the renovation has reduced the total depth of the entire spatial system relative to each street, enhanced the interaction between Spaces and increased the cyclic spatial relationships.
Referring to the framework of (
Wang et al., 2025), we further supplemented the visibility graph analysis (VGA) and intelligent agent simulation in the renovation case section to provide more dimensional analysis and discussion for the renovated space:
1) The average Visual Connectivity increased after the renovation, from 614.709 to 633.658. The Visual Connectivity of each room has also been enhanced. The lower commercial space has achieved more complete openness and visibility, and the openness of the related parts of the visiting streamline in the upper exhibition space has also been strengthened.
2) The average Visual Integration has increased significantly after the renovation, from 7.021 to 8.767, which means that the overall observability of the space has been greatly improved. However, after the renovation, the differences in the Visual Integration of specific details in each space have increased, and the primary and secondary contrast among various parts of the space have strengthened. This provides new perspective suggestions for the subsequent specific organization of the plan’s functions and usage, such as setting up traffic flow lines, main exhibits or commodities in areas with relatively high Visual Integration, and setting up auxiliary functions, secondary exhibits or commodities in areas with relatively low Visual Integration.
3) The Spatial Mobility analysis by intelligent agent simulation offers a new perspective revelation for the spatial usage before and after the renovation. In the pre-renovation plan, the spaces corresponding to the front courtyard, the living room and the staircase have obtained a larger number of more mobile activity trajectories, while the bedrooms have obtained more concentrated and static activity trajectories. To a certain extent, this reflects the original space usage of Lifen. In the renovated plan, the newly added openings have brought about more mobile activity trajectories, which reflects the use of the space with public functional characteristics after the renovation.
Since the above Space Syntax methods based on natural movement and vision and the analysis method based on spatial configuration and topological relations in this research belong to different categories in the field of Space Syntax (
Karimi, 2023), they cannot be directly integrated with the proposed evaluation method in terms of underlying parameters in this paper. But the supplementary analysis and discussion still further prove the improvement of spatial public function adaptability after the renovation using the paradigm in this paper, and at the same time provide valuable new insights into the spatial configuration and interaction before and after the renovation. Meanwhile, this further demonstrates the expansion potential of the method proposed in this paper when combined with multi-dimensional data in the future research.
5 Conclusion and recommendation
This article focuses on the technical optimization of Space Syntax application and discusses the significance of the related data. The novelty of this research lies in that, based on the relevant empirical theoretical foundation of Space Syntax, it expands the calculation strategy for the adaptive reuse evaluation of historical buildings from the perspective of spatial function and usage. The results show that for the adaptive reuse of converting residential functions into public activity functions with uncertainties, through the calculation and use of the proposed Potential Values for pre-renovation assessment, the renovation effects can be better judged, thereby assisting in determining the priority of the renovation. Thus, a better renovation effect can be achieved with less intervention, balancing the protection and renewal of the heritage. Meanwhile, following the theoretical framework of Space Syntax, the article expands the application modes and scenarios of its variables, especially from describing the existing situation to quantitatively guiding decisions, providing ideas for subsequent related research.
It is worth noting that determining the spatial layout of an adaptively modified building is often proven to be a complex task. Architects, engineers, clients, and other relevant parties will put forward different priorities, including but not limited to spatial interaction, circulation, lighting, structural elements, material costs, and labor costs, etc. (
Paulino et al., 2024). This article focuses on the methods within the context of Space Syntax topological analysis, taking adaptive renovation as the example. It concentrates on the innovative exploration of Space Syntax application models, and the significance of the relevant data, rather than discussing the integration of multi-source data in adaptive reuse or Space Syntax analysis. In above context, using only this method is insufficient for the final decision. It is still necessary to integrate the specific demands and practical problems of users, policymakers and developers, etc., and form an extended framework in combination with other methods. However, in view of the complexity of adaptive reuse, this method can still be one of the effective helpers in evaluation and decision-making, providing effective suggestions for related aspects such as spatial interaction and cost, and can help reduce the gap in professional abilities and enhance the participation rights and decision-making rights of the community. Meanwhile, the paradigm of this paper is based on the concept of “minimum intervention” in adaptive reuse, the methods derived from this may help enhance sustainability and reduce carbon emissions in heritage reuse. In addition, the heritage protection policies of different places and the resulting evaluation of architectural authenticity may affect the applicability of the methods in this article and should be taken into account in actual renovations.
The methods and theoretical basis of this research can also provide new references and ideas for the related research of computational design. For example, the methods proposed in this paper and the parameters based on Space Syntax can become one of the indicators for scheme generation and evaluation in computational design. This will be helpful for the screening of schemes and improve the rationality of scheme generation. Subsequent studies can incorporate this method into the model of multi-parameter evaluation and decision-making process of adaptive reuse, optimize its integration with visibility graph analysis, intelligent agent simulation and other computational design methods, and in combination with the specific research goals of each field, triangulation can be conducted with real-world data.
While demonstrated on a specific Lifen heritage complex example, the proposed method holds potential for extension to diverse objects and scales. According to different research aims, PV’s definition and the framework used can be further optimized to enhance precision and adaptability across contexts.
2095-2635/2025 The Authors. Publishing services by Elsevier B.V. on behalf of KeAi Communications Co. Ltd.
PDF
(10691KB)
