Bridging Global Perspectives: A Comparative Review of Agent-Based Modeling for Block-Level Walkability in Chinese and International Research

Abstract

As cities strive for human-centered and fine-tuned development, Agent-Based Modeling (ABM) has emerged as a powerful tool for simulating pedestrian behavior and optimizing walkable neighborhood design. This study presents a comparative bibliometric analysis of ABM applications in block-scale walkability research from 2015 to 2024, drawing on both Chinese- and English-language literature. Using visualization tools such as VOSviewer, the analysis reveals divergences in national trajectories, methodological approaches, and institutional logics. Chinese research demonstrates a policy-driven growth pattern, particularly following the introduction of the “15-Minute Community Life Circle” initiative, with an emphasis on neighborhood renewal, age-friendly design, and transit-oriented planning. In contrast, international studies show a steady output driven by technological innovation, integrating methods such as deep learning, semantic segmentation, and behavioral simulation to address climate resilience, equity, and mobility complexity. The study also classifies ABM applications into five key application domains, highlighting how Chinese and international studies differ in focus, data inputs, and implementation strategies. Despite these differences, both research streams recognize the value of ABM in transport planning, public health, and low-carbon urbanism. Key challenges identified include data scarcity, algorithmic limitations, and ethical concerns. The study concludes with future research directions, including multimodal data fusion, integration with extended reality, and the development of privacy-aware, cross-cultural modeling standards. These findings reinforce ABM’s potential as a smart urban simulation tool for advancing adaptive, human-centered, and sustainable neighborhood planning.

1. Introduction and Research Background

The accelerating pace of global urbanization and renewal has intensified environmental and social pressures on cities, underscoring the importance of interventions at the block scale—often regarded as the fundamental unit of urban governance and everyday life. Unlike macro-scale strategic planning or micro-scale architectural interventions, the block or meso scale represents a critical interface where built form, infrastructure, and human mobility intersect in tangible ways. It is at this level that walkability can be most effectively shaped through both policy frameworks and design implementations, enabling cities to address challenges such as climate resilience, mobility equity, and neighborhood livability. As climate change exacerbates risks like extreme heat, flooding, and resource scarcity, promoting walkable blocks has become essential for fostering adaptive, human-centered, and sustainable urban environments [1]. However, it is important to recognize that walkability is not a universally fixed concept. Its definitions and operationalizations vary significantly across geographic, cultural, and disciplinary contexts, shaped by local planning paradigms, policy objectives, and measurement frameworks [2]. Walkable block systems not only reduce carbon dependency by promoting active transport but also enhance social equity, public health, and economic vitality. As a result, planners and researchers are increasingly turning to digital technologies that simulate, optimize, and visualize the interplay between physical space and human behavior. Among these, Agent-Based Modeling (ABM) has emerged as a key computational framework for exploring how diverse agents—such as residents, commuters, and vulnerable individuals—interact with urban form under varied environmental and policy scenarios [3,4]. This study systematically reviews the application of ABM in walkability research, with particular attention to differences between Chinese and international approaches, aiming to understand how such tools contribute to climate-resilient and human-centered block design.

1.1. Research on Block-Level Walkability

As a key indicator of human-centered urban development, block-level walkability research explores multiple dimensions, including spatial scale, safety resilience, perceptual experience, and age-friendly design. These dimensions reflect the diverse conceptualizations of walkability adopted across studies. While some emphasize spatial accessibility and road connectivity, others focus on sensory perceptions, behavioral patterns, or social inclusiveness. This diversity highlights the importance of treating walkability as a multi-dimensional and context-sensitive construct rather than a fixed or uniform indicator.

With regard to spatial scale, existing studies have proposed a multi-level analytical framework comprising the city network, neighborhood, and block levels. While city- and district-scale studies emphasize accessibility and land-use, the block scale enables finer simulation of walkability-sensitive features—such as setbacks, sidewalks, and building edges—closely linked to pedestrian behavior. However, this also increases data demands and modeling complexity, making block-scale agent-based modeling both more granular and technically demanding [2,5].

In terms of safety resilience, research has focused on human-vehicle collaborative design to improve traffic safety, spatial interventions for crime prevention, and emergency evacuation simulation for disaster scenarios. These efforts aim to strengthen the risk resistance capability of pedestrian systems through spatial strategies [6,7].

Regarding perceptual experience, scholars have developed quantitative evaluation systems based on sensory dimensions such as visual landscapes, acoustic environments, and thermal comfort. These studies explore the interaction between physical settings and pedestrians’ psychological responses, revealing how environmental cues shape walking behavior.

For age-friendly design, research has addressed issues such as children’s activity safety, barrier-free accessibility for the elderly, and intergenerational space-sharing. These approaches promote a transition from universal design toward a more differentiated and inclusive paradigm for pedestrian environments.

Recent studies demonstrate a shift from static spatial analysis to dynamic behavioral simulation [8]. However, the coordination of multiple influencing factors and the accuracy of spatiotemporal behavioral modeling remain underdeveloped. These challenges present theoretical opportunities for advancing the application of ABM in walkability research.

1.2. Methods for Assessing Block-Level Walkability

Given the close relationship between walkability and urban block morphology, assessing block-level walkability has attracted extensive scholarly attention. Over the years, researchers have employed a range of methods to assess walkability across different spatial and analytical contexts. The primary methods used in previous studies are summarized in Table 1, each with its respective strengths, limitations, and application scenarios.

• Survey Method

Table 1. Summary of Common Methods for Assessing Block-Level Walkability.

Method	Advantages	Limitations	Applicable Scenarios
Survey Method	Captures individual perceptions and emotional experiences; low implementation cost	Prone to sampling bias; data validity relies on verification; limited in capturing interactive dynamics	Small-scale or community-level studies
Walkability Scoring	Provides objective measures; useful for comparative analysis; supports visualized outputs	Tends to overlook heterogeneity; limited sensitivity to micro-level variations	Urban planning and policy evaluation
ABM	Enables dynamic simulation; models behavioral interactions; supports parameterized experimentation	Requires extensive parameter calibration; involves high technical and computational cost	Complex block-level environments; emergency scenarios; dynamic simulations

The implementation of the survey method typically involves several key steps. Researchers begin by defining the study objectives and developing a questionnaire aligned with these goals. The instrument is then administered to a target population within the study area that meets predetermined sampling criteria, ensuring sufficient representativeness. Upon data collection, invalid or incomplete responses are excluded, and the valid responses are analyzed using statistical techniques such as descriptive analysis and frequency distribution, with results interpreted in line with the research objectives.

In walkability research, survey methods are frequently complemented by qualitative spatial analysis techniques rooted in urban planning and design traditions. These include the classification of streets based on their urban roles, analysis of spatial supply characteristics such as seating and shading, assessment of land-use distribution, and evaluation of ground-floor interface continuity. Such methods, extensively applied in many planning contexts, provide critical spatial context to subjective perceptions captured by surveys and enhance the interpretive depth of walkability assessments.

The survey method offers several advantages. Standardized formats support consistent data collection, minimize interviewer bias, and facilitate comparative analysis across respondents [9]. However, limitations persist. Responses may be influenced by personal biases or misinterpretation, and non-response can introduce sampling errors. Additionally, reliance on self-reported data often limits access to deeper behavioral motivations or contextual influences. For large-scale applications, the method also demands considerable time and logistical resources [10].

• Walkability Scoring Method

The walkability scoring method quantifies walkability based on a range of environmental and infrastructural factors. These typically include the distance between residential areas and daily activity zones, sidewalk quality, crime rates, and the availability and condition of public facilities. This method allows researchers and planners to systematically translate complex urban attributes into structured scores, facilitating comparative evaluation across different spatial contexts.

Notably, walkability scoring tools vary in methodological rigor. Audit-based instruments are designed with well-defined protocols, objective rating criteria, and training guidelines, which enhance reproducibility and reduce evaluator bias. In contrast, composite indices that rely on researcher-defined weights or context-sensitive indicators may introduce greater subjectivity, especially in indicator selection or weighting schemes [11].

Furthermore, scoring tools differ in analytical focus. Some prioritize the assessment of static physical features of the built environment, while others incorporate behavioral components to reflect how pedestrians interact with space. Despite their usefulness, the applicability of walkability scoring tools can be constrained by data availability and contextual differences in pedestrian preferences and mobility cultures.

• ABM

ABM is a simulation approach grounded in the bottom-up theory of complex systems. It models the dynamic interactions between micro-level individual behaviors and macro-level spatial phenomena by constructing heterogeneous agents equipped with autonomous decision-making capabilities and defined interaction rules [5]. In block-level walkability research, ABM is commonly implemented using simulation platforms such as MassMotion and AnyLogic. The modeling process typically involves the following four steps:

(1)

Agent Modeling: Agents’ attributes (e.g., age, travel purpose) and behavioral logic are defined through programmable rules. Decision-making mechanisms are calibrated using real-world datasets (e.g., GPS trajectories, survey results) to ensure statistical alignment between simulated and actual pedestrian behaviors. For example, group behavior simulations based on the “flocking” algorithm incorporate four core principles—separation, alignment, cohesion, and avoidance—to enable stateless dynamic interactions and reproduce emergent self-organizing pedestrian flow patterns [12].

(2)

Environmental Parameterization: The physical characteristics of the block (e.g., sidewalk width, distribution of points of interest) and social variables (e.g., perceived walkability, safety thresholds) are encoded as computable parameters. Multi-level environmental data fields-such as GIS-based grids and 3D point clouds-are constructed to support real-time agent perception and adaptive response to the surrounding environment.

(3)

Simulation Experiment Design: Key variables—such as block accessibility, path complexity, and facility density—are selected for experimental analysis. Parameter sets are systematically varied, and computational experiments are conducted to simulate agents’ spatiotemporal behaviors. These simulations enable the quantification of nonlinear effects of environmental and design factors on walkability outcomes [13].

(4)

Result Analysis and Validation: Simulation outputs are analyzed using statistical techniques such as logistic regression and spatial clustering to identify the core variables influencing pedestrian walking experience (e.g., travel efficiency, preference for leisure-oriented routes). Model validity is assessed by comparing simulation results with field observations and empirical datasets.

The key advantage of ABM lies in its ability to dynamically capture the bidirectional feedback mechanism between heterogeneous pedestrian behaviors and the built environment. Particularly in medium- and small-scale block studies, ABM facilitates the simultaneous analysis of macro-level network efficiency (e.g., origin–destination accessibility) and micro-level experiential quality (e.g., perceptions of leisurely walking linked to the spatial distribution of points of interest). However, the accuracy of ABM simulations is often constrained by the completeness of parameter calibration and the complexity of underlying algorithms. As a result, it frequently requires integration with traditional methods-such as spatial syntax-to support robust multi-source data fusion [14]. In this context, ABM has emerged as a powerful tool for shifting block-level walkability research from static assessment toward dynamic, process-oriented simulation. This paper provides a systematic review of recent applications and theoretical developments in ABM for modeling walkability in medium- and small-scale urban environments.

Despite the growing prominence of ABM in walkability research, existing studies remain fragmented across linguistic, geographic, and institutional contexts. In particular, Chinese-language and English-language research often diverge not only in methodological preferences and disciplinary orientations, but also in the underlying logics of urban development and planning priorities. In China, the legacy of state-led urbanization—characterized by rapid expansion, centrally coordinated redevelopment, and policy-driven spatial restructuring—has given rise to large-scale ABM applications aligned with national planning objectives such as the “15 min community life circle” [15,16]. In contrast, ABM studies in Western and other international contexts often reflect more decentralized planning systems and emphasize fine-grained behavioral modeling, technological innovation, and ethical considerations in data use [17,18,19].

These divergences highlight the importance of conducting a comparative and systematic review that contextualizes ABM development within the broader socio-political and disciplinary landscapes that shape its application. By critically examining differences in research focus, technological approaches, and urban application settings, such a review facilitates a deeper understanding of how digital simulation tools—especially ABM—are conceptualized, implemented, and validated across diverse planning cultures [20]. It also offers valuable insight into the broader significance, current limitations, and future potential of ABM as a decision-support tool for improving block-level walkability in both global and local urban contexts.

This study aims to address several interrelated research questions that underpin the comparative and exploratory nature of the analysis: (1) How has the application of ABM in block-scale walkability research evolved over the past decade across both Chinese and international academic contexts? (2) What methodological trends, thematic emphases, and institutional logics characterize the differences and convergences between these two research landscapes? (3) In which practical domains has ABM been most frequently applied, and what technical or contextual limitations constrain its broader impact? (4) How can future ABM-based walkability research be advanced through technical innovation, cross-cultural collaboration, and interdisciplinary integration?

The remainder of the paper is organized as follows. Section 2 outlines the data sources and bibliometric methodology. Section 3 presents a comparative analysis of Chinese and international research using a structured analytical framework. Section 4 examines the key application domains of ABM and evaluates its methodological contributions and limitations. Section 5 summarizes the main findings and proposes directions for future research.

2. Data Sources and Methodology

This study employs a bibliometric approach to systematically analyze literature at the intersection of three core concepts: agent-based modeling, block scale, and walkability. Guided by the integrated research framework illustrated in Figure 1, the analysis was carried out in four stages: (1) thematic scoping and literature collection; (2) quantitative evaluation; (3) domain classification; (4) synthesis and interpretation. This structured approach enables a comprehensive understanding of how ABM is applied in block-level walkability studies across regional and disciplinary contexts.

Figure 1. Integrated Research Framework for Bibliometric Analysis of ABM and Walkability at the Block Scale.

2.1. Bibliometric Method

This study adopts a bibliometric method to systematically map and evaluate research on ABM in block-scale walkability. As a quantitative approach rooted in scientometrics, bibliometric analysis enables the identification of research trends, knowledge structures, and scholarly influence by statistically examining large volumes of academic publications. It provides objectivity and replicability by avoiding reliance on subjective interpretation and instead applying standardized indicators such as keyword co-occurrence, citation counts, and publication frequencies [21].

As shown in Figure 1, the process began with retrieving literature from the China National Knowledge Infrastructure (CNKI) and Web of Science (WoS) databases based on the intersection of the three core concepts: agent-based modeling, block scale, and walkability. The retrieved publications were then quantitatively analyzed using key bibliometric indicators, including publication volume, keyword frequency, annual distribution, and geographical origin, with a particular emphasis on comparing Chinese and international research patterns. Moreover, studies were further categorized according to their primary application domain, such as urban transportation, urban design, environmental sustainability, public health, or socio-economic analysis. Finally, the results were synthesized to identify dominant research themes, technical bottlenecks, and emerging methodological trends. Special attention was given to cultural and institutional divergences in ABM practices, as well as the evolving role of ABM as a simulation-based decision-support tool in human-centered urban design.

To support visual pattern recognition and cluster detection, VOSviewer (version 1.6.20) was used to generate keyword co-occurrence maps. These maps visualize research hotspots and thematic linkages by organizing keywords based on their co-occurrence strength and frequency. Collectively, this mixed-method bibliometric approach offers a robust foundation for evaluating the intellectual structure, evolution, and knowledge gaps in the field of ABM-based walkability research.

2.2. Data Source

This study collected academic publications from 2015 to 2024 related to ABM, walkability, and block-scale urban environments from two major databases: CNKI and WOS. A multi-stage screening process was implemented to ensure thematic relevance and analytical rigor.

The screening process involved three phases: (1) deduplication and exclusion of non-academic or irrelevant sources; (2) removing publications outside the 2015–2024 timeframe; (3) manual screening of titles and abstracts to exclude off-topic or tangentially related studies. The final dataset included only those publications that directly addressed ABM applications in walkability or block-scale urban research. The complete selection and filtering workflow is illustrated in Figure 2.

Figure 2. Literature Identification and Screening Process (2015–2024).

To explore the thematic structure of the selected literature, we used VOSviewer to generate a keyword co-occurrence map. In these maps, node size reflects keyword frequency, node position indicates centrality, and color gradations represent cluster affiliation. This visual analytic technique enables researchers to identify research hotspots, emerging subfields, and the evolving contours of ABM-driven walkability scholarship [22].

3. Comparative Analysis of Chinese and International Literature on ABM, Walkability, and Block-Scale Research

3.1. Comparative Analytical Framework

To interpret the bibliometric results in a structured and context-sensitive manner, this study adopts a comparative framework consisting of four dimensions: (1) urban form typology; (2) governance model; (3) technological maturity; (4) cultural–institutional orientation. These dimensions collectively shape how ABM is conceptualized, implemented, and evaluated in block-level walkability research across different planning contexts. By situating keyword co-occurrence patterns within this multidimensional structure, we are able to reveal deeper distinctions between Chinese and international research paradigms.

As summarized in Table 2, Chinese literature tends to emphasize land-use allocation, neighborhood-scale renewal, and pragmatic model feasibility—often reflecting top-down policy imperatives and infrastructure-led urban transformation. In contrast, international research focuses more heavily on pedestrian dynamics, evacuation modeling, and data ethics, driven by finer-grained urban forms, participatory governance mechanisms, and a strong emphasis on technological innovation and ethical accountability.

Table 2. Comparative Analytical Framework of Chinese and International Research Hotspots in ABM-Based Walkability Studies.

Dimension	Chinese Research Hotspots	International Research Hotspots	Interpretive Insights
Data Analysis Methods	Land-use; Simulation Modeling	Pedestrian Dynamics; Data Assimilation	Chinese studies focus on land-use allocation and block-scale modeling; international work emphasizes micro-level behavior in complex, high-resolution networks.
Application Areas	Traffic Safety; Block Renewal	Evacuation Models; Intelligent Transportation	Chinese research aligns with policy-driven agendas, while international research prioritizes risk mitigation and system-wide transport innovation.
Technical Challenges	Model Simplification; Accuracy Trade-offs	Data Privacy; Ethical Regulations	Chinese models prioritize feasibility and implementation in large-scale urban systems; international studies highlight ethical concerns, user privacy, and regulatory compliance.

3.2. ABM Research at the Block Scale: Trends and Regional Perspectives

3.2.1. Trends in Annual Publications

Over the past decade (2015–2024), both Chinese- and English-language research on ABM at the block scale have shown a consistent upward trajectory. According to data from CNKI, a total of 1324 Chinese-language publications addressed ABM in neighborhood-scale urban studies. From these, 102 highly relevant articles specifically focusing on block-scale applications were identified and selected for in-depth analysis. The annual publication volume showed a growth trend over the study period. Meanwhile, data from WOS identified 184,087 English-language publications related to ABM in neighborhood-scale contexts. Among them, 861 strongly relevant articles were screened based on thematic alignment with block-scale applications. Although the overall international publication volume far exceeds that of Chinese literature, the rate of annual increase was comparatively moderate and more stable throughout the decade.

As illustrated in Figure 3, these trends highlight a growing global interest in the application of ABM to simulate complex urban systems at the block scale. Chinese research, while emerging from a lower baseline, has accelerated rapidly in recent years. International publications, by contrast, have maintained a consistently high output, addressing the established role of ABM in international urban simulation studies.

Figure 3. Annual Trends in Chinese and International Publications on ABM at the Block Scale (2015 to 2024).

3.2.2. Thematic Focus and Keyword Patterns

Chinese Research: Emphasis on Local Practices and Application Orientation.

As illustrated in Figure 4, Chinese-language research on ABM at the neighborhood and block scale from 2015 to 2024 reveals a strong focus on policy-driven urban renewal, community planning, and spatial governance. The most prominent keyword cluster centers on “urban renewal”, which forms strong connections with related terms such as “urban micro-renewal”, “organic renewal”, and “15 min community life circle”. These terms indicate the integration of ABM within national strategies for revitalizing aging urban areas while promoting walkable, service-accessible communities.

Figure 4. Keyword Co-Occurrence Map of Chinese Literature on Agent-Based Modeling at the Block Scale (2015–2024).

The presence of geographically specific keywords like “Shanghai”, as well as role-oriented terms like “responsible planner”, further addresses the alignment between academic research and local implementation in first-tier cities. This is particularly evident in research applied to the transformation of superblocks and the development of community life circles in cities such as Shanghai [23].

In addition, keywords such as “street space” indicate a focus on optimizing pedestrian mobility through ABM-based simulations. For example, simulation studies have evaluated the placement of street buffers and crosswalks, resulting in improved pedestrian efficiency and safety. In the context of urban design, ABM has been used to model comfort under varying building setback distances, helping to inform interface adjustments in block redevelopment projects like those in Hefei.

Another notable cluster includes “grassroots governance”, “urban governance”, and “block-renewal”, indicating a growing scholarly focus on the co-evolution of space and society. This reflects an ongoing shift from spatial design as a purely physical intervention toward models of collaborative governance and participatory planning—where ABM plays a role in capturing the behavior and preferences of multiple stakeholders.

In the fields of public health and socio-economic planning, ABM is used to simulate pedestrian behavior in relation to public facility allocation. Keywords such as “landscape architecture” and “scene creation” suggest an expanding application of ABM in shaping experiential and emotional dimensions of urban space, especially in relation to age-friendly and inclusive design. Moreover, ABM is deployed to quantitatively assess how elements such as street greening and leisure infrastructure influence microclimatic conditions and walking behavior [24].

International Research: Technological Frontiers and Interdisciplinary Integration.

As shown in Figure 5, international research on ABM at the block scale is characterized by strong emphasis on technological innovation and interdisciplinary integration, with clusters revolving around deep learning, feature extraction, nonlinear systems, and built environment.

Figure 5. Keyword Co-Occurrence Map of International Literature on Agent-Based Modeling at the Block Scale (2015–2024).

Technological keywords such as “deep learning”, “machine learning”, and “feature extraction” dominate the network. These terms reflect the field’s emphasis on algorithmic enhancement, model parameter calibration, and computational scalability. For example, the integration of deep reinforcement learning (DRL) into ABM frameworks enables agents to adapt their behavior dynamically in response to evolving spatial stimuli, surpassing traditional rule-based mechanisms [25,26,27]. This advancement is especially critical in policy-sensitive urban scenarios, where behavior modeling must accommodate uncertainty and real-time feedback.

Additionally, a cluster of public health and equity-related keywords—such as “environmental justice,” “social determinants of health,” and “COVID-19”—demonstrates growing interest in the equity and health implications of walkable urban environments. For instance, ABM has been used to simulate disparities in walkable access to health-promoting infrastructure. One study found that ethnic minority communities had only 60% of the walkable access to health food stores compared to low-income areas. After optimizing facility layout via ABM, the community dietary health score increased by 12% [28]. During the COVID-19 pandemic, ABM was further deployed to model pedestrian routing, evacuation behavior, and virus transmission under varying social distancing regulations [29]. Emerging interest in neurocognitive simulation is also visible in keywords like “neural networks”. These approaches aim to bridge ABM with affective computing, allowing for the modeling of emotional and perceptual responses during walking. Such studies are pushing ABM beyond physical modeling toward human-centered, multi-sensory, and emotion-aware urban design, especially within public health and urban planning domains [30].

3.2.3. Cross-Regional Comparison and Shared Research Priorities

Chinese and international research on ABM at the block scale exhibit clear distinctions shaped by differences in urban form, governance models, and technological orientation. Chinese studies are predominantly policy-driven, closely aligned with national planning agendas such as “15 min community life circles” and “open block-renewal”. The focus is on meso-scale interventions that enhance accessibility, optimize facility layout, and support grassroots governance. In contrast, international studies are more technology-centric and interdisciplinary, integrating methods like deep learning, semantic segmentation, and nonlinear systems modeling to simulate micro-scale pedestrian behavior and system complexity. They also address broader equity concerns, as seen in frequent references to environmental justice, public health, and data ethics.

Despite these differences, both research communities converge on key priorities. Shared keywords—such as “COVID-19,” “multi-source data,” and “crowd simulation”—demonstrate ABM’s universal value in modeling urban dynamics under uncertainty. Additionally, the rising emphasis on data integration and visualization in both contexts reflects a common need to improve model accuracy, scalability, and policy relevance. These overlaps suggest strong potential for cross-regional collaboration, particularly in advancing ABM applications that balance technical innovation with real-world governance and social impact.

3.3. ABM-Driven Walkability Research: Thematic Evolution and Cross-Cultural Comparison

3.3.1. Trends in Annual Publications

A bibliometric search of the CNKI database for the period 2015–2024 identified 220 Chinese-language publications related to walkability research using ABM. After relevance screening, 108 highly relevant articles were selected for analysis. Although annual publication counts fluctuate from year to year, a general upward trend is evident—particularly between 2017 and 2022—with noticeable peaks in 2018, 2020, and 2023.

By comparison, the WOS search yielded a significantly larger body of 37,578 articles over the same period. After applying strict screening criteria to ensure topic relevance, 86 core publications were identified as highly relevant. International publication trends show a steady increase between 2015 and 2020, followed by a moderate decline in the post-pandemic period after 2021.

As illustrated in Figure 6, Chinese-language publications consistently outnumber their international counterparts in annual volume, particularly between 2017 and 2023. Peaks in Chinese publication counts occurred in 2018, 2020, and 2023, with the highest share of decade-total output observed in 2020 and 2023 (both exceeding 15%). In contrast, international literature showed a more modest and fluctuating pattern. Although international publication numbers remained relatively stable across the decade, their proportional contribution to total output peaked in 2020 (17%) before declining steadily through 2024. This divergence may reflect shifting research priorities post-COVID, with Chinese research continuing to expand while global attention began to refocus on other areas.

Figure 6. Annual Trends in Chinese and International Publications on Walkability Research Using ABM (2015–2024).

3.3.2. Comparative Keyword Analysis

Chinese Research: Demand-Oriented and Technological Adaptation.

As illustrated in Figure 7, Chinese-language research on walkability using ABM demonstrates a strong demand-responsive orientation, with keywords clustering around practical challenges in aging populations, community health, and infrastructure equity. The central node “Walkability” connects tightly with terms such as “public health”, “walkability index”, and “built environment”, indicating a systematic approach to understanding how built form and facility accessibility influence mobility outcomes—particularly for vulnerable populations.

Figure 7. Keyword Co-Occurrence Map of Chinese Literature on Walkability Research Using ABM (2015–2024).

Terms such as “aging”, “aging transformation,” and “barrier-free access” reflect China’s demographic reality and the increasing need for inclusive spatial planning. ABM is commonly applied to simulate elderly walking behaviors—accounting for slower speeds, increased resting needs, and sensitivity to environmental quality—in order to optimize spatial layout at the neighborhood scale. Moreover, terms like “rail transit” and “living circle” highlight a policy-driven emphasis on last-mile connectivity and transit-oriented development. These studies commonly use ABM to evaluate walkability within a 500 m radius of key transport nodes, supporting the planning of 15 min community life circles.

At the technical level, traditional methods such as “structural equation modeling” and “analytic hierarchy process” are increasingly complemented by “multi-source data” and “street view images”, revealing a transition from qualitative surveys to data-enriched simulations [31]. Visual variables such as greening ratio and interface activeness—often extracted from image data—are now modeled in ABM frameworks to predict their influence on pedestrian preferences and movement.

Finally, regional keywords such as “cold cities” and “Hefei” point to geographically targeted research, accounting for climatic differences and local planning needs in the optimization of walking environments.

International Research: Technological Empowerment and Ethical Reflection

As shown in Figure 8, international research on walkability using ABM reveals a technology-intensive landscape, where clusters revolve around autonomous vehicles, deep learning, object detection, and pedestrian behavior. These themes indicate a strong orientation toward smart mobility, machine vision, and real-time simulation in complex urban environments.

Figure 8. Keyword Co-Occurrence Map of International Literature on Walkability Research Using ABM (2015–2024).

The keyword “autonomous vehicles” is closely linked with “vulnerable road users”, “pedestrians” and “protocols”, reflecting growing research interest in simulating mixed traffic environments involving pedestrians and intelligent vehicles. These studies use ABM to model pedestrian behavior under various communication protocols and sensor conditions, enhancing system resilience and minimizing collision risks. Moreover, the cluster centers on “deep learning”, “object detection”, and “Google Street View” highlight the integration of computer vision into ABM frameworks. For example, convolutional neural networks (CNNs) trained on street-level images are employed to identify obstacles, signage, and sidewalk conditions. These data are then used to dynamically adjust pedestrian routing within the simulation, improving behavioral accuracy and environmental responsiveness [32,33].

Supporting terms such as “task analysis”, “data mining”, and “cameras” further emphasize the multidisciplinary character of this research, where AI, sensor technologies, and human factors are fused to inform agent logic and movement dynamics. Although smaller in network prominence, keywords like “COVID-19”, “agent-based modeling”, and “resilience” cluster together to reflect continued use of ABM in health-related scenarios, including pandemic-era simulations of crowd movement, evacuation modeling, and compliance with public health protocols.

3.3.3. Cross-Cultural Insights and Methodological Reflections

Chinese research on ABM-driven walkability is largely shaped by policy-oriented planning and demand-responsive practice. The studies are typically grounded in meso-scale urban forms—such as life circles and aging communities—where ABM is applied to simulate accessibility, aging-friendly infrastructure, and last-mile mobility. This reflects a governance model closely aligned with top-down directives, especially in implementing initiatives like the 15 min community life circle. The emphasis is on providing actionable, location-specific planning insights rather than advancing simulation methodologies.

In contrast, international research emphasizes micro-scale behavioral modeling and macro-scale systems thinking, underpinned by a strong focus on technological innovation and interdisciplinary integration. Clusters around “autonomous vehicles”, “deep learning”, and “object detection” show that ABM is frequently combined with AI techniques and sensor data to simulate dynamic pedestrian–vehicle interactions and optimize walkability in real time. Ethical concerns are more prominent, revealing a stronger cultural orientation toward regulation, inclusion, and critical reflection.

Despite these differences, both research communities share methodological priorities, including the need for multi-source data fusion, behavioral realism, and model transparency. Moving forward, Chinese studies may benefit from incorporating international advances in algorithmic calibration and ethical safeguards, while international efforts could gain from China’s grounded, implementation-oriented modeling. These approaches can advance ABM-based walkability research toward a more integrated framework of demand–technology–ethics.

3.4. Walkability at the Block Scale: Spatial Features and Analytical Focus

3.4.1. Trends in Annual Publications

According to CNKI data, a total of 1528 Chinese-language articles on walkability at the block or neighborhood scale were published between 2015 and 2024. After applying relevance-based screening criteria, 210 highly relevant articles were selected for analysis. As shown in Figure 9, annual publication output exhibited steady growth from 2015 to 2019, followed by consistent productivity from 2020 onward. Notably, publication peaks occurred in 2020 and 2023, suggesting increased scholarly interest aligned with national planning initiatives and block-scale renewal efforts.

Figure 9. Annual Trends in Chinese and International Publications on Walkability at the Block Scale (2015–2024).

In comparison, the WOS database returned a significantly larger dataset of 629,091 English-language articles during the same period. From these, 2319 highly relevant articles were identified through focused screening. The international literature shows a robust and sustained growth trajectory, with publication volumes gradually rising and peaking in 2023.

3.4.2. Spatial Themes and Methodological Patterns

Chinese Research: Policy-Driven and Scale-Nested.

As illustrated in Figure 10, Chinese literature on block-scale walkability exhibits a clear scale-nested structure, bridging macro-level policy agendas, meso-level block-renewal, and micro-level facility optimization. Central keywords such as “urban renewal”, “open block”, and “public space” reveal a strong alignment with national strategies like the 15 min community life circle and open block transformation.

Figure 10. Keyword Co-Occurrence Map of Chinese Literature on Walkability at the Block Scale (2015–2024).

Notably, cultural heritage and spatial equity emerge as important sub-themes. Clusters around “historical blocks”, “organic renewal”, and “street space” suggest a dual focus on preservation and usability. In cities like Beijing and Shanghai, ABM has been employed to distinguish movement patterns between tourists (average speed: 1.5 m/s) and residents (1.1 m/s), helping to optimize pedestrian flow across core heritage zones and surrounding buffer activity areas [16]. These interventions improve walkability while supporting both cultural continuity and public health.

Moreover, the frequent co-occurrence of “multi-source data” and “structural equation modeling” reflects an ongoing shift from single-source surveys to integrated evaluation approaches. By combining POI, mobile phone signaling, and user surveys, researchers assess the correlation between accessibility and resident satisfaction—supporting decisions on commercial layout, land-use allocation, and urban vitality assessment. Furthermore, keyword like “responsible planner” indicates growing attention to participatory design. ABM is increasingly used to simulate feedback loops involving residents, merchants, and planners, facilitating multi-stakeholder coordination in block-scale regeneration efforts [34].

International Research: Ecological Orientation and Social Critique.

Figure 11 reveals that international literature on block-scale walkability is characterized by a strong emphasis on artificial intelligence, data-driven urban sensing, and behavioral simulation. Dominant keywords such as “deep learning”, “feature extraction”, “machine learning”, and “built environment” form a dense core, highlighting the centrality of AI technologies—particularly in extracting and interpreting spatial features for ABM applications.

Figure 11. Keyword Co-Occurrence Map of International Literature on Walkability Research at the Block Scale (2015–2024).

In particular, the close connections between “street view images,” “semantic segmentation,” “object detection,” and “transformer” models reflect the widespread use of computer vision, especially CNN-based and Transformer-based models. These methods are frequently employed to detect streetscape elements—such as sidewalk width, green coverage, or façade diversity—which are then used to parameterize agent behavior in simulations of walkability and urban comfort.

The presence of “computational modeling”, “task analysis”, and “convolutional neural networks” further demonstrates the integration of human–environment interaction modeling, where agent logic is adjusted in response to perceptual or functional stimuli. In addition, the inclusion of “blockchain” and “privacy” suggests an increasing awareness of ethical and privacy-related concerns, especially in studies involving mobility data, user trajectories, or public surveillance systems.

Overall, the international research landscape is marked by the fusion of technical sophistication and critical inquiry, extending ABM’s scope from predictive modeling to ethically grounded and ecologically aware urban interventions.

3.4.3. Comparative Summary and Emerging Trends

The comparative analysis reveals a structural divergence between Chinese and international research on block-scale walkability, shaped by differences in urban form typology, governance models, technological maturity, and cultural–institutional orientation. In particular, Chinese literature is strongly embedded within a “policy–scale–practice” paradigm, characterized by nested spatial logics—macro-level urban policy, meso-level block renewal, and micro-level facility optimization. Studies commonly model pedestrian behavior under conditions of aging, cultural heritage preservation, or grassroots co-governance, with emphasis on spatial equity, commercial vitality, and walkability improvement.

In contrast, international research follows an “ecology–technology–critique” trajectory, underpinned by decentralized urban forms. Technological maturity is a defining feature: deep learning, semantic segmentation, object detection, and computational modeling dominate the methodological landscape. This has enabled fine-grained extraction of built environment features—such as greening, façade variation, and spatial enclosure—which inform perceptually sensitive ABM frameworks. International studies are integrating ethical considerations such as privacy protection, bias mitigation, and environmental justice—particularly in public health and socio-economic domains.

Despite these differences, shared challenges and convergence points are emerging. Both research communities are increasingly concerned with data ethics, particularly in the use of large-scale mobility data and visual surveillance. Keywords like blockchain, privacy computing, and federated learning suggest a growing demand for globally unified ethical standards in behavioral modeling. In future research, Chinese studies can incorporate international techniques such as street-level semantic segmentation for quantifying carbon impacts, improving the ecological rigor of heritage block renewal projects. Meanwhile, international studies could adapt China’s scale-nested planning logic by embedding digital twin models at the community scale, improving the real-world applicability of technological innovations. Collaborative research around digital twin blocks, agent rule optimization, and ecological equity simulation will be critical in shaping a more integrated global research paradigm—anchored in policy responsiveness, ecological resilience, and technological accountability.

4. Application Domains and Practical Implications of ABM

This section explores the practical applications of ABM across five key domains: urban transportation, urban design, environmental sustainability, public health, and socio-economic analysis. These areas capture the breadth of ABM’s operational use while reflecting varied disciplinary priorities and data contexts. Drawing on both Chinese and international literature, we examine representative case studies and assess the extent to which simulation outcomes inform real-world planning. The section concludes with a critical discussion of ABM’s technical and institutional feasibility, including challenges related to data integration, model transparency, scalability, and ethical considerations.

4.1. Urban Transportation Planning

Urban transportation systems face increasing challenges related to congestion, safety, and sustainability. In this context, ABM has emerged as a powerful tool for simulating pedestrian and vehicular behaviors at the block scale. Unlike traditional static models, ABM enables the representation of individual agents—such as pedestrians or vehicles—interacting dynamically with their built environment, thereby generating more realistic insights into flow patterns, congestion hotspots, and the impacts of infrastructural interventions.

In Chinese literature, ABM is often applied to simulate pedestrian routing in transit-oriented development zones, optimize walking environments around subway stations, and evaluate block-level street retrofitting strategies. For instance, several studies have integrated multi-source data—including street view imagery and mobile phone signaling—to assess the effects of buffer widths, intersection design, and sidewalk connectivity on walking efficiency in dense neighborhoods [15,35]. These applications are often aligned with planning initiatives such as the “15 min community life circle”, where ABM supports decision-making by visualizing pedestrian flows and identifying facility coverage gaps.

International studies have also advanced large-scale ABM simulations in urban transportation contexts. One notable example is the development of a region-wide pedestrian flow model for the city of Salzburg, which analyzed crowd behavior and walking accessibility using historical travel data [36]. Another study developed a spatial ABM framework to simulate adult walking patterns within an urban environment, incorporating variables such as demographic profiles, land-use mix, and street network characteristics [37]. More recent advances include the integration of real-time sensor data and computer vision—such as CNNs applied to Google Street View imagery—for the calibration of pedestrian behavior and assessment of safety risks in smart mobility systems [38].

Overall, ABM provides a versatile simulation framework to support block-level transportation planning. It facilitates a better understanding of multimodal travel behavior, supports the evaluation of design interventions, and bridges the gap between urban simulation and practical policymaking. Its growing adoption across both Chinese and international research landscapes reflects a shared recognition of its potential in human-centered, data-informed transportation systems.

4.2. Urban Planning and Design

With the evolution of urban planning from function-oriented to people-centered paradigms, enhancing walkability at the block scale has become a key focus area in urban design. ABM offers a powerful approach to simulate residents’ daily routines, assess behavioral responses to spatial configurations, and support scenario testing in real time. By modeling individual agents with personalized attributes and decision rules, ABM helps evaluate the effects of layout interventions—such as building density, street permeability, or land-use mix—on travel choices, pedestrian comfort, and community vitality.

Chinese literature increasingly applies ABM in urban planning and design, particularly for public space optimization, block renewal, and spatial policy testing under data-constrained conditions. For example, Yang et al. [39] employed a multi-agent simulation framework to evaluate the spatial quality of the North Bund waterfront in Shanghai. Their study simulated pedestrian movement and spatial usage patterns under different layout scenarios, providing evidence-based recommendations to improve spatial permeability and public engagement. In another case, Ye et al. [40] proposed a method for constructing agent-based simulation models using limited data sources, such as mobile phone trajectories and simplified activity plans. Despite data constraints, the study successfully modeled urban transportation patterns and walkability metrics, demonstrating how ABM can inform design and policy decisions under real-world limitations.

International studies have also applied ABM to examine urban morphology and its influence on pedestrian flows. A notable example is Badland et al.’s simulation of walkability enhancements in the northwest suburbs of Melbourne, where simple ABM helped identify sidewalk improvements that aligned with daily travel routines [5]. More recent work integrates eye-tracking, emotional response data, and VR-assisted ABM to assess how residents perceive street esthetics and spatial safety [17].

4.3. Environment and Sustainable Development

Environmental protection and sustainable development have become increasingly important in block-scale walkability research. Walkable block designs contribute to reducing carbon emissions, improving air quality, and mitigating heat stress. ABM plays a key role in this domain by simulating changes in travel behavior (e.g., modal shifts from driving to walking or cycling) and quantifying their environmental impacts.

For example, a study in Yuen Long, Hong Kong, integrated ABM with microclimate modeling to evaluate six heat mitigation strategies—such as street trees, green façades, and permeable pavements—on pedestrian thermal comfort and walkability levels [41]. Similarly, research in Beijing’s Sanlitun district investigated how morphological and environmental features like street aspect ratio, tree spacing, and greenery influence pedestrian vitality and perceived comfort during summer conditions [42]. A study in Nanjing focused on older adults and found that block-level characteristics like street connectivity, green space availability, and proximity to essential services significantly influenced both walking frequency and environmental satisfaction [43]. These studies demonstrate how ABM supports evidence-based design and planning by revealing the co-benefits of environmental interventions and walkability improvements.

4.4. Public Health

In recent years, the benefits of walking for residents’ physical and mental health have attracted significant attention, and promoting walking through block design has become an important research direction. ABM can simulate residents’ activity patterns across different urban block environments, quantifying the potential impact of design changes on health outcomes. It not only provides visualizations of health benefits following walkability optimizations but also generates quantitative data for health-oriented block design.

For example, Chen et al. examine how built environment attributes (street connectivity, green space, service proximity) are associated with physical activity frequency and health satisfaction among older adults in various blocks—but it does not directly simulate changes via ABM scenarios [43]. Another study simulates individuals’ daily activity routines, modeling exposure to air pollution and heat stress, which are directly relevant to walkability design for public health [44].

4.5. Socio-Economic Analysis

Socio-economic analysis highlights the role of block-level walkability in shaping patterns of economic development, commercial vitality, and social equity. ABM provides a valuable simulation framework to evaluate how different block configurations influence economic opportunities, service accessibility, and the equitable distribution of urban resources. By simulating interactions between residents, businesses, and infrastructure, ABM helps planners test interventions—such as adjusting sidewalk widths, locating public services, or reconfiguring block layouts—and predict their implications for commercial activity and population wellbeing.

For instance, a study in Melbourne used ABM to analyze how changes in walkable catchments affected residents’ access to amenities such as employment hubs, healthcare facilities, and public transit. The findings identified neighborhoods where enhanced walkability would generate the highest socio-economic returns, informing targeted urban investment strategies [5]. Another study examined walkability disparities across socio-economic groups, revealing that lower-income and mobility-constrained populations often face significantly longer walking distances to reach essential services—highlighting systemic inequalities embedded in block-scale urban design [43]. These studies demonstrate ABM’s capacity to inform more inclusive, equity-oriented design policies.

4.6. Feasibility and Limitations of ABM

ABM offers a powerful computational framework for simulating complex urban systems, enabling the representation of individual behaviors, spatial interactions, and emergent dynamics in built environments. Its strengths lie in flexibility, scalability, and the ability to integrate diverse data sources—including geospatial, behavioral, and perceptual data—for scenario testing and decision support [18].

In urban studies, ABM allows researchers to replicate pedestrian movements, test infrastructure interventions, and simulate interactions under varying environmental, social, or policy conditions. For example, models have been used to examine accessibility inequalities [18], detect anomalous crowd behavior through synthetic data generation [19], and optimize emergency egress planning in spatially constrained environments [45]. These applications demonstrate the model’s utility in capturing both routine and critical dynamics within block-scale planning contexts.

However, ABM also presents notable limitations. First, model calibration and validation remain a persistent challenge: simulation parameters may not fully capture the complexity of real-world behavior, leading to oversimplification or bias in results. Second, computational cost grows exponentially with the number of agents, spatial resolution, or simulation steps, requiring high-performance computing for large-scale applications [46]. Third, the use of personal trajectory or sensor data raises privacy and ethical concerns, necessitating compliance with data protection standards and responsible governance frameworks. Lastly, despite efforts to standardize model transparency and documentation—such as the ODD protocol—issues of replicability and interpretability continue to affect the credibility and transferability of ABM in policymaking contexts [47].

5. Conclusions and Future Research Directions

By integrating co-occurrence network analysis, temporal publication trends, and domain-specific application cases, this study has revealed both divergences and convergences in ABM research on walkability across different geographical, institutional, and disciplinary contexts. Chinese studies are characterized by a policy-driven and scale-nested approach, where ABM is embedded within national planning frameworks such as the “15 min community life circle” and applied extensively to community renewal, open block retrofitting, and spatial equity optimization. These studies emphasize practical relevance, scenario simulation, and alignment with urban governance objectives.

In contrast, international literature demonstrates a technology-intensive and ethically oriented paradigm, advancing ABM through the integration of machine learning, computer vision, and privacy-preserving modeling. These innovations extend the application of ABM in environmental resilience, emotional cognition, and social justice, indicating a growing concern for both technical sophistication and ethical responsibility in urban modeling.

ABM has demonstrated wide-ranging utility in five core domains, including urban transportation, urban design, environmental sustainability, public health, and socio-economic analysis. Case studies show that ABM is effective not only in optimizing street connectivity and public space layouts, but also in simulating pedestrian exposure to heat or pollution, evaluating health benefits, and revealing socio-spatial disparities in acess to services and amenities. Its ability to simulate multi-agent interactions under various conditions makes it particularly suited for supporting human-centered and evidence-informed planning.

Nevertheless, challenges in model calibration, data integration, computational demands, and ethical governance limit the broader adoption of ABM in policy and planning workflows. In particular, concerns around replicability, transparency and interpretability—especially in large-scale or real-time applications—pose obstacles for institutional implementation. Furthermore, differences in cultural and institutional contexts shape both the research focus and methodological sophistication of ABM studies. While international work often prioritizes technical innovation, it may lack contextual adaptability; conversely, Chinese studies tend to emphasize policy alignment and applied value, but are sometimes constrained by technical conservatism.

Building on the comparative findings and application cases presented, future research in ABM-driven walkability studies should evolve along four interconnected dimensions: agent-based technology, walkability evaluation, block-scale design, and interdisciplinary integration.

• From Static Rule-Based Models to Intelligent, Data-Fused Systems

Conventional ABM frameworks often rely on manually calibrated rules and single-source inputs, constraining their behavioral fidelity and adaptability to dynamic conditions. Future research should explore the development of reinforcement learning (RL)-based parameter calibration frameworks, which dynamically optimize agent behavior—such as pathfinding or obstacle avoidance—through context-sensitive reward mechanisms. The integration of deep reinforcement learning (DRL) with GPS trajectory data can significantly improve the fidelity of complex pedestrian movement modeling.

Moreover, the fusion of multimodal data—including semantic segmentation of street view imagery, mobile phone signaling, and sentiment-labeled social media content—can support the development of cognitive agents capable of adaptive decision-making. The development of standardized ABM toolchains (e.g., Mesa, UrbanSim) and open-source case libraries, will lower technical barriers and improve the interpretability of emergent macroscopic patterns derived from microscopic interactions.

2.

From Static Evaluation to Dynamic Perception and Climate Resilience

As urban environments become increasingly dynamic and vulnerable to climate stressors, future walkability research must move beyond static, geometry-based accessibility metrics toward more perception-driven, adaptive, and resilience-oriented evaluations. To this end, extended reality (XR) technologies, including virtual reality (VR) and augmented reality (AR), can be employed to build digital twin neighborhoods. These tools enable urban planners to interactively manipulate design parameters such as sidewalk width or tree canopy coverage, while receiving real-time feedback on pedestrian comfort, flow efficiency, and perceived spatial quality.

AR-enhanced real-world environments, coupled with eye-tracking, heart rate variability (HRV), and other biometric data, allow for the quantification of complex “emotion–behavior–space” interactions, thereby extending ABM into the neurocognitive domain. These capabilities make it possible to simulate not only physical movement patterns but also psychological and affective responses to urban environments, supporting more empathetic, human-centered planning.

Furthermore, the integration of climate models, such as the Weather Research and Forecasting (WRF) model, into ABM frameworks can simulate pedestrian behavioral adaptation under extreme weather conditions. This allows researchers to evaluate the resilience-enhancing functions of green infrastructure and to construct scenario libraries for black swan events, improving the robustness and foresight of walkability-oriented planning interventions.

3.

From Functional Optimization to Human-Centered Sustainable Design

To address the diverse mobility needs of different population groups, future ABM applications should incorporate hierarchical agent models that represent children, older adults, and individuals with mobility impairments. These agents can simulate age-specific behavioral preferences-such as children’s preference for proximity to play-oriented spaces or older adults’ reliance on barrier-free and rest-accessible paths-thereby informing the human-centered refinement of neighborhood spatial layouts.

Moreover, intergenerational simulation scenarios can be employed to assess “universal + differentiated” design strategies, evaluating how shared and group-specific interventions influence walkability equity across age cohorts. This approach enables planners to identify inclusive solutions that accommodate the needs of all users without sacrificing functionality or accessibility for any particular group. Additionally, environmental benefit assessment modules should be integrated into ABM frameworks to quantify the co-benefits of increased pedestrian activity on carbon reduction, microclimate regulation, and thermal comfort.

To fully understand the systemic impact of such interventions, these environmental outcomes should be linked with socio-economic indicators, enabling models to evaluate how walkability-enhancing interventions contribute to commercial vitality, employment accessibility, and public health improvement. Such integrated approaches will support the development of multi-objective optimization models that balance the interrelated goals of environmental sustainability, population wellbeing, and urban economic resilience, advancing the transformation of urban blocks into human-centered, inclusive, and sustainable living environments.

4.

Toward an Open, Cross-Disciplinary and Participatory Ecosystem

Future advancement of ABM depends on its integration with complex systems science, particularly through network theory for modeling cascading behavioral effects and game theory for simulating right-of-way negotiations in shared urban spaces. These frameworks enhance ABM’s capacity to explain emergent mobility patterns and inform equitable policy interventions.

To broaden accessibility, interoperability between ABM engines (e.g., NetLogo, Mesa) and urban computing platforms (e.g., UrbanSim, CityScope) should be standardized, reducing technical barriers and enabling wider adoption—especially in resource-constrained contexts. Establishing an intersectoral consortium linking government, academia, industry, and communities is also critical to translate simulation into actionable planning. Through a closed-loop workflow of simulation → optimization → implementation → feedback, ABM can evolve into a practical, participatory tool for human-centered and resilient block design.