Simulation Study on Age-Friendly Design of Community Park Activity Spaces Based on AnyLogic: A Case Study of Qiaokou Park in Wuhan

Abstract

With the intensification of population aging, addressing the needs of older adults and enhancing their daily activities has become increasingly significant. This study focuses on community parks—frequent outdoor activity venues for older adults—as the research subject. Starting from older adults’ needs, pedestrian simulation technology was employed using AnyLogic to model their behavioral activities within Qiaokou Park in Wuhan. Unlike previous studies applying simulation tools to general public spaces, this research develops age-sensitive indicators (Pedestrian Walking Cost, Connectivity of Activity Space Nodes, Functional Mix Efficiency, Activity Intensity of Activity Space Nodes, Pedestrian Density Map) tailored to older adults’ behavioral and spatial characteristics. Integrating empirical data from questionnaires and on-site observations with simulation, the study establishes a systematic framework linking user needs and spatial design. Based on simulation outputs, the park’s current “non-age-friendly” issues were analyzed, and optimization strategies were proposed regarding service capacity, functional layout, and pathways. The optimized scheme underwent secondary simulation to evaluate improvements in spatial indicators. This approach extends the methodological toolkit for age-friendly park research and provides replicable, evidence-based guidance for community park renovation in rapidly aging urban contexts. Key recommendations include the following: (1) Improve the relationship between activity nodes and park entrances; (2) Enhance connectivity among nodes to support continuous activity flows; (3) Optimize the pathway network to reduce congestion and barriers; (4) Promote functional diversity to stimulate active and social use; (5) Strengthen service capacity of nodes to accommodate user demand.

1. Introduction

With the global trend of population ageing becoming increasingly evident, China is undergoing a rapidly accelerating and deepening ageing process. According to authoritative data from the Chinese Academy of Social Sciences (IPLE/CASS), the population aged 60 and above in China reached 178 million in 2010 and increased to 249 million by 2018 [1]. Projections from the National Working Commission on Ageing further estimate that by around 2050, the number of older adults will reach approximately 487 million [2]. As a central city in China’s mid-region, Wuhan has experienced an even more pronounced demographic shift. In 2019, the population aged 60 and above in Wuhan was 1.879 million, accounting for 21.27% of the total population, indicating that the city has already entered the stage of “deep ageing.” This demographic shift imposes new imperatives for urban sustainable development [3]. At the international level, policy frameworks such as the World Health Organization’s “Global Age-Friendly Cities: A Guide” emphasize the creation of environments that support health, participation, and security for older adults [4]. Universal Design (UD) and Active Aging Design advocate inclusive environments that accommodate diverse physical, sensory, and cognitive abilities. Carr et al. argue that UD not only improves accessibility but also supports “successful aging” by enabling independence and participation [5], while Ramírez-Saiz et al. highlight its role in enhancing urban mobility and social inclusion for older adults and people with disabilities [6]. Verma further emphasizes participatory design in adapting cities for older adults, integrating their needs directly into planning practices [7]. Building on these approaches, Steels identified key characteristics of age-friendly cities and communities, such as multi-sectoral collaboration, strong government commitment, and the active involvement of older adults in decision-making [8]. Similarly, Jackisch et al. demonstrated that healthy cities and age-friendly cities are mutually reinforcing, as improvements to social and physical environments simultaneously support public health and active aging [9].

As aging progresses, the spatial radius of daily activities among older adults progressively contracts, elevating community-level public spaces to critical venues for social engagement [10]. Within this context, community parks emerge as high-frequency public spaces for daily outdoor activities among older adults [11], thereby assuming a pivotal role in advancing urban age-friendly environmental development.

The genesis of community parks traces back to the Industrial Revolution in 18th-century Britain, where private gardens and green spaces were transformed into public gardens and parks to address escalating urban challenges. The United States commenced its urban park movement in the mid-19th century, and Manhattan’s Central Park was completed in 1873. Around the turn of the 20th century, the first generation of community parks in their authentic conception emerged. Beginning in the 1970s, research on community parks expanded beyond fulfilling basic public needs to address users’ diverse requirements [12]. For instance, Clare Cooper Marcus advocated for resident involvement in park design and planning, emphasizing the criticality of direct community participation [13]. In addition, some studies examined socio-cultural distinctions among park users [14], revealing perceptual variations across racial, gender [15], and income level [16].

As important venues for promoting physical and mental well-being as well as social welfare among older adults, community parks have increasingly attracted research attention on how their design can stimulate daily activities and promote health in aging populations [17]. For instance, Veitch, Jenny et al. comprehensively investigated older adult users’ specific requirements for park features, including tranquil environments, comfortable walking paths, and social amenities [18]; Pleson examined older adult users’ behaviors and perceptions across seven community green spaces in Taipei, and proposed improvements involving green space accessibility, structured activities, and sports areas [19]; Veitch also argued that park designers should prioritize shade-providing trees, peaceful and relaxing environments, and walkability to better meet older adults’ needs [20]; Gaikwad posited that developing countries should likewise advance healthy living for older adults when designing age-friendly cities [21]. Overall, although substantial progress has been made in research on age-friendly community parks, limitations and challenges remain. For example, the diversity of data analysis tools and quantification methods that are used limits the comparability and generalizability of results. Most studies are based either on the subjective needs of older adults or on isolated aspects of community parks, lacking simulation, visualization, and systematic analysis of the complex interactions between park spaces and older adult behaviors.

Recent years have witnessed rapid advancements in pedestrian simulation technology. By emulating pedestrian behaviors, this approach enables in-depth analysis of spatial usage processes from a pedestrian perspective, reveals interactions between pedestrians and spatial environments, and reconstructs spatiotemporal dynamics of pedestrian flows, thereby providing robust technical support for spatial design decision-making [22]. In the early 1970s, pedestrian simulation technology was initially deployed to assess roadway service levels and assist in transportation planning and design [23]. Subsequently, accelerated by computer technology advancements, pedestrian simulation evolved and expanded into fluid dynamics and kinetic modeling domains [24]. Currently, microscopic simulation has formed a relatively comprehensive system due to its ability to incorporate individual differences, environmental interactions, and group dynamics [25]. It includes various modeling approaches such as Gravity Models, Social Force Models, Movement Benefit Models, Cellular Automata Models [26], and Queuing Network Models. In terms of research methodology, Gipps proposed the shortest path algorithm [26], while Jaros developed a 3D railway station model based on the Unity framework [27]—both studies leveraged dynamic individual behavior simulations to elucidate spatial usage patterns. Regarding simulation tools, AnyLogic has seen widespread application in areas such as medical scheduling (e.g., Schaumann’s study on hospital drug distribution [28]), emergency evacuation (e.g., Ha’s exit parameter studies [29]), and facility planning (e.g., Shirzadi’s simulation of residential layout [30]), owing to its integrated capability for discrete event modeling, agent-based modeling, and system dynamics modeling. In research objectives, pedestrian simulation studies remain predominantly focused on transportation systems and safety evacuations [31], exemplified by Kirik’s quantification of bottleneck efficiency using the SigmaEva module [32]. However, with the rising demand for high-quality urban public spaces in recent years, simulation studies have begun expanding into pedestrian leisure and recreational behaviors. For instance, Kielar proposed memory-based accessibility destination choice models [33], Borgers uncovered “spontaneous stops” triggered by environmental visual stimuli [34], and Nasir and Mojdeh simulated stimulus-responsive trajectory alterations via genetic fuzzy systems [35]. Yet these studies predominantly integrate behavioral theories, concentrating on relationships between spatial environments and activity distributions—such as how pedestrian preferences and needs influence route choices—while underemphasizing spatial service level evaluations.

Within this context, this study aims to address three pivotal objectives: (1) Establishing a methodological framework for pedestrian simulation in age-friendly community park design; (2) Investigating community park case studies to identify their age-related spatial challenges; (3) Proposing targeted age-friendly renewal strategies leveraging simulation technology. To address these objectives, this research employs multiple methods—including on-site questionnaire surveys, pedestrian tracking techniques, and pedestrian simulations—centered on Wuhan’s Qiaokou Park as a demonstrative case. The goal is to establish a closed-loop design process of “analyzing pedestrian activity patterns—using simulation results to inform design—verifying optimized solutions”, thereby providing a replicable paradigm for community park renewal in aging cities.

2. Materials and Methods

2.1. Research Framework

As shown in Figure 1, initially, a field study was conducted from three key dimensions: (1) Assessment of existing environmental conditions, analyzing physical spatial characteristics of community park activity areas; (2) Deciphering behavioral patterns, identifying activity types, spatiotemporal distribution dynamics, and behavioral preferences of older adults park users; (3) Collecting pedestrian flow data, including path choices, pedestrian volume, and activity duration of older adult users. Second, following comprehensive mastery of pedestrian simulation software functionalities, a set of simulation-based indicators was established to assess the age-friendliness of community park activity spaces, including Pedestrian Walking Cost, Connectivity of Activity Space Nodes, Functional Mix Efficiency, Activity Intensity of Activity Space Nodes, and Pedestrian Density Map. Third, based on empirical findings, an older adult activity simulation model was developed in Anylogic (v8.7) comprising environment modeling, behavioral rule modeling for activities, and pedestrian parameter configuration. Within this framework, discrete-event modeling formalized older adult behavioral sequences, agent-based modeling defined individual behavioral characteristics, and the Social Force Model predicted pedestrian trajectories, avoidance behaviors, and group dynamics. Finally, leveraging Anylogic outputs, pedestrian simulation metrics were integrated with empirical observations to formulate targeted age-friendly design proposals addressing identified spatial deficiencies. Re-simulation validation was then performed in Anylogic, with comparative analysis of pre- and post-intervention metric variations demonstrating the effectiveness of design solutions.

2.2. Field Study

2.2.1. Study Area

This study selects Qiaokou Park (Qiaokou District, Wuhan, China) in Wuhan as the research site. Located in the central area of Qiaokou District, the park covers a total area of 3.96 ha, with green spaces accounting for 2.66 ha, yielding a green space ratio of 69.27%. The surrounding residential communities were mostly constructed around the year 2000 and are characterized by a significantly aging population. Therefore, the park demonstrates a strong demand for age-friendly design and interventions.

The area surrounding Qiaokou Park is well-equipped with public service facilities. Within a 500 m service radius, there are several key public buildings, including hospitals and schools. Within a 1000 m radius, the area is served by three metro stations and eight bus stops, accommodating two metro lines and 41 bus routes, resulting in a high volume of pedestrian traffic (see Figure 2 for details).

As shown in Figure 3, the spatial layout of Qiaokou Park reflects the typical characteristics of small-scale community parks in older urban areas. Its boundaries are tightly enclosed and constrained by surrounding roads and buildings, resulting in an irregular rectangular shape. Due to its relatively limited size, the internal layout is simple; however, the functional zoning lacks clarity, and the park’s path system is ambiguous hierarchy. There are a total of seven entrances to the park. Among them, entrances 1, 3, 5, and 6 experience higher pedestrian flow as they are adjacent to main streets or located near residential access points. The park also features a relatively high density of activity space, with 14 activity space nodes distributed throughout the site.

2.2.2. Questionnaire Design and Result Analysis

(1)

Questionnaire Design and Distribution

The questionnaire was designed to collect empirical data for parameterizing the AnyLogic agent-based model, focusing on four aspects: (a) basic information, (b) behavioral patterns, (c) activity needs, and (d) satisfaction with existing facilities. Sections (a)–(c) employed frequency statistics, while section (d) utilized a 10-point Likert scale (1 = extremely dissatisfied, 10 = extremely satisfied) to assess satisfaction with various aspects of Qiaokou Park.

The survey targeted older adult users of the park and was conducted on clear days between April and May 2023 to avoid weather-related biases. A total of 150 questionnaires were distributed, yielding 138 valid responses (92% effective rate) after excluding incomplete submissions.

(2)

Data Summary and Application to Model Input

Basic information (see Figure 4 for details): The data on age, gender, health status, and residence proximity were used to define the agent population structure (e.g., agents were assigned ages and genders based on the surveyed distribution).

Behavioral Patterns (see Figure 5 for details): Data on visitation frequency, activity duration, time of day, and activity preferences were critical for scheduling agent behaviors in the simulation (e.g., defining the duration of activities). The strong preference for social activities informed the design of group behavior rules.

Activity needs (see Figure 6 for details): User preferences for routes, facilities, and activity spaces directly informed the modeling of nodes and pathways. This guaranteed that the simulation reflected actual user behavior, creating a reliable basis for future design proposals.

Satisfaction with existing facilities (see

Table 1. Survey Questionnaire Results (Satisfaction with Existing Facilities.).

Evaluation Dimension	Specific Item	Mean Score	Standard Deviation
A Entrances and exits	A-1 Distance from residence to park entrance	6.89	1.79
	A-2 Number of entrances and exits	7.49	1.69
	A-3 Distance to your primary activity space	8.83	3.33
B Walking paths	B-1 Width/circulation capacity	7.27	1.88
	B-2 Safety	7.43	1.74
	B-3 Continuity of the pedestrian network	7.31	1.59
C Exercise activities	C-1 Capacity of exercise spaces/facilities	7.22	1.82
	C-2 Diversity of exercise activity types	6.94	1.70
	C-3 Location of exercise spaces	7.23	1.83
D rest activities	D-1 Capacity of rest spaces/facilities	7.10	1.73
	D-2 Diversity of rest activity types	6.94	1.93
	D-3 Location of rest spaces	7.38	1.65
E Recreational activities	E-1 Capacity of recreational spaces/facilities (e.g., chess, performances, parent–child play)	6.84	1.86
	E-2 Diversity of recreational activity types (e.g., chess, performances, parent–child play)	6.78	2.04
	E-3 Location of recreational activity spaces (e.g., chess, performances, parent–child play)	7.03	1.90

): Satisfaction with existing facilities identified key spatial shortcomings (e.g., low scores for entrance accessibility A-1 and recreational diversity E-2), which became the main targets for optimization in our design. This part was tested in SPSS (v25.0) for reliability and validity, with a high KMO value (0.893) and a significant Bartlett’s test (p < 0.001).

2.2.3. Pedestrian Flow Data

Pedestrian flow of older adults was recorded at the entrances/exits and various activity points of the community park. To comprehensively understand the flow patterns at different times of the day, the survey was divided into five main time periods for observation: early morning (6:00–8:00), morning (8:00–10:00 and 10:00–12:00), and afternoon (14:00–16:00 and 16:00–18:00). During each time period, observations were sampled every 10 min to record pedestrian flow data. The survey was conducted continuously for two weeks to obtain average flow values. At entrances and exits, the focus was on recording the flow of older adults, while at activity points, both the flow of older adults and the duration of stay were recorded.

Statistical results (see Figure 7 for details) reveal uneven temporal distribution of older adult pedestrian flows in Qiaokou Park, with peak periods concentrated during 06:00–08:00, 14:00–16:00, and 16:00–18:00. Conversely, lower flows characterize the 08:00–10:00 and 10:00–12:00 intervals. Furthermore, significant flow variations exist across park entrances, correlating strongly with seniors’ activity patterns. Entrance 1 demonstrates exceptionally high flows during 06:00–10:00 due to its proximity to dense residences, convenient transit access, and three exercise nodes. Entrance 2 exhibits low flows owing to its secluded location. Entrance 3 registers pronounced afternoon/evening peaks as the primary access point for Jianle Community seniors. Entrance 5 maintains moderate flows from 08:00–12:00. Entrance 6 serves Yu’er Community with relatively low pedestrian flows in its tranquil setting. Entrance 4 exclusively serves vehicular access to a community parking lot adjoining the park’s northern edge, experiencing negligible pedestrian use due to safety hazards from barrier gates. Similarly, Entrance 7 leads to institutional facilities (Qiaokou Land Reserve Center and Finance Bureau), consequently recording minimal visitation.

Behavioral observations and survey analyses (see

Table 2. Visitor Counts by Time Period at Activity Nodes.

Activity Space Node	Average Duration (Minutes)	6:00–8:00	8:00–10:00	10:00–12:00	14:00–16:00	16:00–18:00
Chess & Card Node 1	10	0	2	0	152	148
Chess & Card Node 2	10	0	0	1	81	88
Chess & Card Node 3	10	2	0	0	32	29
Exercise Node 1	10	0	2	2	2	4
Exercise Node 2	10	8	7	13	17	14
Exercise Node 3	10	24	19	8	10	13
Exercise Node 4	10	69	123	62	107	53
Rest Node 1	10	6	18	23	34	28
Rest Node 2	10	0	2	4	3	4
Rest Node 3	10	2	4	0	2	1
Recreational Node 1	10	0	4	2	2	5
Recreational Node 2	10	2	2	3	4	2
Recreational Node 3	10	0	0	2	0	6
Recreational Node 4	10	0	7	6	3	4

for details) indicate that 14:00–16:00 and 16:00–18:00 constitute peak periods for chess and card game activities among seniors. During 06:00–08:00, only exercise activities show elevated activity levels. Afternoons see intensified activity volumes across fitness zones, rest nodes, and recreational facilities. Notably, seniors’ average duration at board game facilities varies between 128 and 177 min, indicating considerable variation. Among rest nodes, Rest Node 1 accommodates the highest user capacity with the longest mean duration (63.8 min), whereas Rest Nodes 2 and 3 record significantly shorter durations (8 and 13 min, respectively). For recreational activity spaces, mean durations per 10-user cohort at Nodes 1–4 are 3.6, 6.2, 8.7, and 13.9 min, showing notable variation.

2.3. Simulation-Based Indicators

2.3.1. Pedestrian Walking Cost

Pedestrian Walking Cost is a key indicator used to measure the efficiency of seniors’ movement from their starting point to the activity destination, represented by the average travel time. The shorter the time older adult individuals take to reach their activity destination, the lower the pedestrian walking cost and the higher the accessibility; conversely, a higher walking cost indicates lower accessibility. The mathematical expression is as follows:

$$C={\displaystyle \frac{1}{N}}\sum _{i=1}^N{T}_i$$

(1)

where

• C: Pedestrian walking cost (unit: minutes);
• N: Sample size (i.e., number of older adult individuals);
• T_i: The travel time of the i-th older adult individuals from the starting point to the activity destination.

2.3.2. Connectivity of Activity Space Nodes

Connectivity of Activity Space Nodes is a crucial metric that measures the spatial closeness between various activity nodes within a community park, typically represented by the average travel time of older adult individuals moving between these nodes. A shorter average travel time between nodes indicates closer distances and more convenient path designs, reflecting better connectivity and stronger internal accessibility within the park; conversely, longer travel times suggest poorer connectivity and weaker internal accessibility. The mathematical expression is as follows:

$$c={\displaystyle \frac{1}{N}}\sum _{i=1}^N\left({\displaystyle \frac{1}{M_i}}\sum _{j=1}^{M_i}{T}_{ij}\right)$$

(2)

where

• C: Connectivity of activity space nodes (unit: minutes);
• N: Sample size (i.e., number of older adult individuals);
• M_i: Number of activity node pairs actually visited by the i-th older adult individual;
• T_ij: Travel time of the i-th older adult individuals between the j-th pair of activity nodes.

2.3.3. Functional Mix Efficiency

The Functional Mix Efficiency of a community park is quantified in this study as the average time required for older adults to move between different types of activity nodes. A lower average time indicates that activity facilities are more compact and diverse, enhancing the park’s accessibility and allowing older adults to fulfill multiple recreational needs within a trip. The mathematical expression is as follows:

$$D={\displaystyle \frac{1}{N}}\sum _{i=1}^N\left({\displaystyle \frac{1}{k_i}}\sum _{k=1}^{k_i}{T}_{ik}\right)$$

(3)

where

• D: Functional Mix Efficiency (unit: minutes);
• N: Sample size (i.e., number of older adult individuals);
• k_i: Number of different functional activity node pairs visited by the i-th older adult individuals;
• T_ik: Travel time of the i-th older adult individuals between the k-th pair of different functional activity nodes.

2.3.4. Activity Intensity of Activity Space Nodes

Activity Intensity of Activity Space Nodes is an important metric for measuring the usage efficiency of activity facilities within community parks. A higher activity intensity indicates more frequent use of the facilities, stronger service performance of the community park, and greater participatory potential. The AnyLogic software package can capture the number of active users within facility points in the simulation model, and line charts are used to visualize the real-time trends of user counts.

2.3.5. Pedestrian Density Map

AnyLogic collects pedestrian density statistics through its Pedestrian Library and generates a Pedestrian Density Map, which visually represents the concentration of pedestrians in different spatial areas using color gradients. This density map can be used to assess the comfort level of older adult pedestrians within community park activity spaces: excessively low density may cause feelings of insecurity, while overly high density can lead to crowding and exclusion, potentially resulting in congestion risks that compromise pedestrian safety. Furthermore, the pedestrian density map displays the distribution of pedestrians along pathways, enabling the identification of bottleneck areas such as narrow passages and intersections. It also reflects the usage frequency of activity nodes, helping to identify highly attractive zones as well as underutilized spaces.

2.4. Pedestrian Simulation Model Construction

Based on actual conditions, a spatial model was constructed in AnyLogic, incorporating spatial elements such as park boundaries, entrances and exits, walking paths, and activity facilities. Components including Wall, Path, Goal Line, Polygon Node, and Attractor were used to define the space. Specifically, walls delineate the pedestrian circulation framework, paths guide movement flows, and goal lines generate spatiotemporal boundaries; together, these elements form the physical framework for older adult activities. In specific activity behaviors, nodes and attractors jointly influence the behavioral choices of older adult individuals. Nodes (e.g., tree pit seating circles, fitness plazas, and other activity spaces) provide potential spaces for older adult activities, while attractors generate behavioral driving forces through the layout of key facilities such as fitness equipment and children’s play areas.

Subsequently, based on the survey results, relevant parameters were set in the simulation model: questionnaire surveys and on-site observations provided data on older adults’ demographics, activity preferences, walking speed, and facility usage (see

Table 3. Key Modeling Parameters and Data Sources.

Parameter	Value/Distribution	Data Source	Remarks
Gender (Male:Female)	Discrete distribution (64, 36)	Questionnaire Survey	Reflects the actual gender composition of park users
Age	Discrete distribution (55, 18, 19, 8)	Questionnaire Survey	Used to differentiate behavioral characteristics among age groups
Speed	Triangular (0.2, 0.4, 0.6) m/s	Field Survey	The distance elderly individuals move per unit time, affected by resting, sightseeing, and social interaction
Radius	Uniform (0.45, 1.30)	Field Survey	Perception radius for interactions such as avoidance and following
Probability of walking in groups	Bernoulli (0.9)	Field Survey	Probability that elderly individuals enter the park as a group or alone
Group distance	member.moveTo (leaderX + 0.8, leaderY + 0.8)	Field Survey	Ensures group members maintain approximately 0.8 m distance around the leader

for details).

In constructing the behavioral rule model for older adult individuals, the Discrete Event Modeling method was applied to decompose their behavior into a sequence of decision-making activities. Within AnyLogic’s Pedestrian Library model, this process primarily utilizes components such as PedSource, PedGoto, PedSelectOutput, PedWait, and PedSink. Given the complexity and variability of actual older adult activities, the behavioral rule modeling must be as detailed and comprehensive as possible: (1) multi-level paths should be established to describe older adult activity patterns as comprehensively as possible; (2) behavioral patterns of older adult individuals differ across various entrances and exits and thus require separate modeling. A typical sequence of activities for an individual pedestrian agent is as follows: the agent is instantiated, followed by the assignment of attributes such as gender, age, walking speed, and group membership; the agent then selects a path and moves to a node. Upon arrival, the agent decides whether to engage in an activity, which may involve a dwell time, before proceeding to the next node and eventually exiting the park.

Based on the survey results, pedestrian flow rates were reasonably assigned to each entrance and exit (see

Table 4. Delay Time Configuration at Activity Nodes. (Data were collected through a field survey.)

Activity Space Node	Activity Delay Setting	Activity Space Node	Activity Delay Setting
Chess & Card Node 1	triangular (95, 178, 253)	Rest Node 1	triangular (25, 64, 117)
Chess & Card Node 2	triangular (82, 160, 205)	Rest Node 2	triangular (2, 8, 17)
Chess & Card Node 3	triangular (54, 130, 200)	Rest Node 3	triangular (7, 13, 22)
Exercise Node 1	triangular (12, 30, 45)	Recreational Node 1	triangular (1, 3.6, 7)
Exercise Node 2	triangular (19, 38, 68)	Recreational Node 2	triangular (3, 6, 11)
Exercise Node 3	triangular (30, 47, 58)	Recreational Node 3	triangular (2, 8, 17)
Exercise Node 4	triangular (45, 77, 115)	Recreational Node 4	triangular (5, 14, 25)

for details). When older adult individuals arrive at a specific facility point, there is a probability of triggering an activity delay event, with the duration of stay following a triangular distribution. Moreover, different facility points exert varying degrees of attraction on the older adults, resulting in different lengths of stay.

3. Simulation of Age-Friendly Adaptation of Community Park Activity Spaces Based on AnyLogic

To enhance the stability of the simulation results, the runtime of the Qiaokou Park simulation model was set to 3 h, during which pedestrian flows within the model tend to stabilize, better reflecting the actual activity patterns of older adults in the park. The simulation was conducted for the period from 14:00 to 18:00, when the number of older adults and the diversity of activities in Qiaokou Park are at their highest, ensuring that the simulation results are representative.

3.1. Output and Analysis

3.1.1. Simulation Results: Pedestrian Walking Cost

As shown in Figure 8, the average pedestrian walking costs for chess & card activities, exercise activities, resting activities, and recreational activities were 854.3 s, 836.35 s, 1322.39 s, and 1392.42 s, respectively. The walking costs for chess & card and exercise activities were relatively low, whereas those for resting and recreational activities were significantly higher, indicating that the corresponding activity space nodes were either located in remote areas or experienced lower pedestrian flow.

The pedestrian walking cost histogram for chess & card activities exhibited a bimodal distribution, indicating substantial variation among older adult individuals in terms of walking cost. Approximately 60% of older adult participants were able to reach the chess & card activity areas within 1000 s, suggesting good accessibility. For instance, Entrances No. 2 and No. 3 are located relatively close to chess & card nodes 1 and 2, while Entrances No. 5 and No. 6 are closer to node 3. However, 10% of older adult participants had walking costs exceeding 2000 s, mainly because Entrance No. 1 is far from the chess & card nodes, and some older adult individuals were not frequent participants in such activities, their activity paths were more circuitous, resulting in higher walking costs. The mean walking cost for exercise activities was 836.35 s, with 50% of older adult participants achieving walking costs below 500 s. This is attributed to the fact that exercise nodes with high pedestrian flow—such as fitness equipment areas and exercise squares—are concentrated near the main entrances or central areas of the park, enabling most older adult visitors to reach them quickly. The walking cost for resting activities was significantly higher. Nearly 48% of older adult participants required more than 1000 s to reach resting nodes, suggesting that some resting nodes (e.g., Resting Node 3) were located in remote areas. This may be because older adult visitors typically engaged in other activities before resting, thereby increasing their walking cost. The mean walking cost for recreational activities reached 1392.42 s (approximately 23 min), far exceeding that of other activity types. Data revealed that over 60% of older adult participants could reach recreational nodes within 1000 s, whereas 16% had walking costs exceeding 2000 s. Considering the actual conditions of Qiaokou Park, recreational nodes are concentrated in the northeastern section of the park with an uneven spatial distribution. Moreover, due to the park’s relatively poor natural landscape and the limited variety of recreational activities, their overall attractiveness is low, contributing to the higher walking costs.

3.1.2. Simulation Results: Connectivity of Activity Space Nodes

As shown in Figure 9, the mean connectivity of activity space nodes in the community park was 651.73 s, indicating that the connections between nodes in Qiaokou Park were relatively convenient, enabling older adults to move from one node to another within a short time. Specifically, 35% of the older adult samples exhibited node connectivity values of less than 500 s, while 50% fell within the range of 501–1000 s, suggesting that the connections between most nodes were relatively optimal. Considering the current situation of Qiaokou Park, the park is small in scale, and the activity space nodes are arranged in close proximity, particularly in the core area, where nine activity nodes are located near the exercise plaza. However, chess and card nodes 1 and 2, situated in the deep western part of the park, are relatively distant from other activity nodes. Although their connectivity is less favorable, this layout naturally acts as a buffer, isolating noise generated by chess and card activities.

3.1.3. Simulation Results: Functional Mix Efficiency

As shown in Figure 9, the mean value of Functional Mix Efficiency in Qiaokou Park was 728.61 s, indicating that, during a single trip, most older adults required relatively short travel times when moving between activity nodes of different functional types, suggesting a high variety of activities available within one visit. A total of 72% of older adults could move between nodes of different functions within 1000 s, and 4% required less than 500 s, reflecting that the spatial arrangement and connectivity of most functional zones were reasonable. For example, both the northern and southern areas of the park contained activity nodes representing the majority of functions. In contrast, 28% of older adults had a Functional Mix Efficiency travel time ranging from 1001 to 2000 s, suggesting that certain areas of the park had more dispersed functional layouts. For instance, the western section of the park contained only chess-and-card and exercise nodes, with the latter being a tennis court, which had relatively few participants.

Figure 9. Simulation results (Histogram): Connectivity of Activity Space Nodes (Left) & Functional Mix Efficiency (Right).

Figure 9. Simulation results (Histogram): Connectivity of Activity Space Nodes (Left) & Functional Mix Efficiency (Right).

3.1.4. Simulation Results: Activity Intensity of Activity Space Nodes

As shown in Figure 10, chess-and-card and exercise activities attracted the largest number of participants, consistent with the findings from the questionnaire survey and flow count statistics. Specifically, the real-time number of participants at chess-and-card nodes 1, 2, and 3 remained stable at approximately 150, 50, and 30 individuals, respectively. Comparatively, the number of participants at these three chess-and-card activity nodes was proportional to the scale of the venue and the service capacity of its facilities; that is, the larger the activity area and the greater the number of chess tables and chairs, the more participants were present at a given time. For exercise nodes, however, the number of participants was related only to the type of exercise and the diversity of exercise facilities. For example, although exercise node 1 covered a relatively large area, the number of older adult participants was low because tennis activities require a certain entry threshold. In contrast, exercise node 2, despite its smaller size, attracted a relatively large number of older adult participants due to the appeal of fitness equipment. Exercise node 4 had the highest number of participants because the hard-surfaced fitness plaza frequently hosted square dancing, tai chi, and other activities with low participation barriers for older adults. In addition, the numbers of participants at rest and recreation nodes were relatively small. For example, rest nodes 2 and 3 and recreation nodes 1, 2, 3, and 4 attracted fewer older adult visitors due to their smaller spatial scale and weaker appeal. However, the real-time number of participants at rest node 1 remained stable at around 60 individuals because it was adjacent to exercise nodes 2 and 4, located in the park’s core area, and equipped with as many as 35 tree-surround benches, providing high service capacity and strong attractiveness to older adults.

3.1.5. Simulation Results: Pedestrian Density Map

As shown in Figure 11, high-density areas within Qiaokou Park are primarily concentrated near the main entrances, the central activity square, and popular activity zones such as chess and card nodes. These areas typically serve as the initial stopping points for older adults upon entering the park, as well as the main destinations in their activity sequences, thereby attracting large gatherings. For example, the fitness square near Entrance No. 1 (Exercise Node 4), with its spacious layout and group exercise activities, has become a popular venue for both physical activity and social interaction among older adults. Chess and card Nodes 1, 2, and 3, as well as Exercise Node 2 and Rest Node 1, exhibit regular patterns of activity distribution due to the presence of attractive facilities such as chess tables and chairs, fitness equipment, and circular tree-basin seating.

Overall, the pedestrian density of older adults in Qiaokou Park exhibits significant spatial variation, with flows more dispersed in the central areas and more concentrated along the periphery. However, analysis of the simulation results for the Activity Intensity of Activity Space Nodes indicates that the lower pedestrian volume in the central areas is not attributable to insufficient activity appeal, but rather to their relatively large spatial scale. In contrast, the peripheral activity zones feature smaller spatial scales and more concentrated facilities, leading to higher density levels that, to some extent, impair the quality of interaction experiences for older adults. Therefore, in optimizing park design, the relationship between spatial scale and facility layout should be considered to enhance the overall activity experience for older adults.

3.2. Cause Analysis

By integrating the aforementioned simulation indicators with actual conditions, the following conclusions are drawn:

(1)

The pedestrian walking cost at high-frequency activity nodes is relatively low, while it is higher at low-frequency activity nodes. In Qiaokou Park, walking costs for exercise and chess/card activity nodes are relatively favorable, whereas those for rest and recreational nodes require improvement. Specifically, exercise nodes are highly accessible, with 50% of older adult individuals able to reach these facilities within five minutes. Exercise nodes 2 and 4 are located less than 100 m from major entrances, conforming to the design principle that “high-frequency use nodes should be close to entrances.” However, there is a stratified disparity in accessibility for chess and rest nodes; some nodes are situated in remote locations (chess node 3, rest nodes 2 and 3), resulting in poor accessibility. Considering older adult behavioral characteristics, it is found that the main activity purposes of older adults in Qiaokou Park focus on exercise and chess rather than rest or recreation.

(2)

The park achieves a balance between overall functional segregation and accessibility of activity nodes. Due to the relatively small scale of Qiaokou Park, nodes in the core area are densely distributed, with an average node connectivity time of approximately 11 min, which meets the physiological needs and activity habits of older adult individuals. Furthermore, the park’s functional segregation and node accessibility create a balanced advantage. Although the cluster of chess nodes on the west side shows reduced connectivity due to soundproofing requirements, this effectively controls adverse effects.

(3)

Service capacity and attractiveness are key factors influencing node activity intensity. The chess and exercise areas attract a large number of older adult participants. This not only reflects the strong demand among older adults for chess and exercise activities but also indicates that the layout and environmental atmosphere of these nodes adequately meet their social and exercise needs. Notably, chess node 1 has the highest number of participants, demonstrating the direct influence of site scale and facility service capacity on activity numbers. Activity intensity in rest and recreational areas is relatively low; except for rest node 1, which attracts more older adults due to its advantageous location and higher service capacity, other rest and recreational nodes have lower utilization rates due to smaller scale and limited attractiveness. This suggests that increasing node service capacity and enhancing facility attractiveness can further lower participation barriers and encourage broader older adult engagement.

(4)

There is a significant supply–demand contradiction in the park’s peripheral areas. High-density older adult pedestrian zones mainly concentrate around major entrances, core activity plazas, and popular activity nodes. In contrast, peripheral areas exhibit concentrated pedestrian flows, smaller spatial scales, and denser facilities, potentially causing older adult users to feel crowded during activities, which negatively impacts their experience. Specifically, paths near recreational nodes 2 and 4 are narrow with poor accessibility, while rest node 2, recreational node 1, and chess node 3 have limited spatial scales and pronounced supply–demand conflicts.

(5)

There is a spatial adaptation imbalance in the chess activity areas. Compared with chess node 2, chess node 1 has a significantly larger number of tables and chairs—more than twice that of node 2—yet its spatial scale is relatively smaller, resulting in a high table-and-chair density of 2.8 per 10 square meters. The excessive number of tables and chairs makes the space feel crowded and markedly reduces visual permeability. Older adult participants may experience spatial oppression during activities, and the high-density crowd can restrict air circulation, increasing the risk of secondhand smoke exposure and heat accumulation. Due to physiological decline with aging, older adult individuals are more sensitive to temperature changes, which can affect their concentration and overall enjoyment during activities.

3.3. Optimization Plan

Based on the analysis results, the following optimization recommendations are proposed:

(1)

Enhance spatial quality and service capacity to increase the overall attractiveness of the park.

The participation rate of older adults in recreational activities at Qiaokou Park is relatively low. Apart from being influenced by older adult users’ differing needs, the quality of facility environments remains a core limiting factor. Therefore, it is recommended to improve the spatial quality of recreational activity areas and revitalize the landscape to promote diversified activity types among older adult users. For example, the fish pond (recreational node 4) could undergo ecological water circulation renovation, and an immersive cultural gallery could be created. Regarding issues such as narrow paths and poor accessibility at some recreational nodes (e.g., nodes 2 and 4), and small spatial scales at rest nodes (e.g., rest node 2), nearby walking paths should be widened and accessible ramps added to enhance spatial attractiveness and utilization.

Chess & card nodes (e.g., chess & card node 3) have limited spatial scale, and the table-and-chair density at chess & card node 1 far exceeds age-friendly standards. Therefore, it is advisable to convert chess & card node 3 into a recreational node, and simultaneously expand the spatial scale of chess & card nodes 1 and 2 while increasing their table-and-chair numbers. This approach will ensure adequate service capacity, alleviate supply–demand conflicts, maintain spatial openness and air circulation, improve older adult users’ experience, and reduce negative impacts on other areas.

(2)

Implement functional integration and micro-renovations by adding embedded rest facilities to reduce walking costs.

Add embedded rest nodes: Qiaokou Park has relatively few resting spaces, some of which are located remotely and fail to connect well with main activity routes and high-frequency nodes, resulting in high pedestrian walking costs that adversely affect older adult users’ experience. Therefore, it is recommended to reposition certain rest nodes and add embedded rest points within highly active areas, such as near chess & card node 2 and the exercise zones (near entrance 1 and exercise nodes 3 and 4). This will reduce walking costs and create an organic connection within activity sequences, thereby enhancing older adult users’ experience (see Figure 12 for details).

(3)

Optimize the path network to divert older adult pedestrian flows.

The overall walking path environment in Qiaokou Park is suboptimal, with some recreational nodes (e.g., nodes 2 and 4) having narrow paths and dead-end designs that impair accessibility. Therefore, widening narrow intersections is recommended, along with the addition of landscape elements along main pedestrian routes, such as seasonal flower beds and interactive scenic walls, to enrich older adult pedestrians’ transit experience and improve the park’s overall comfort and attractiveness.

3.4. Simulation Modeling and Results Comparison

Based on the aging-friendly renovation plan for Qiaokou Park, adjustments were made to the environmental model while pedestrian parameters remained unchanged; the behavioral rule modeling of older adult pedestrians was fine-tuned according to their activity characteristics in Qiaokou Park. Subsequently, a secondary simulation experiment was conducted. Analysis of the adjusted simulation results reveals the following:

(1)

The pedestrian walking costs for chess & card, exercise, rest, and recreational activities all decreased, reaching 667.39 s, 675.12 s, 1256.96 s, and 1155.81 s, respectively, with reductions of 186.92 s, 161.23 s, 65.94 s, and 236.61 s. Although two additional rest nodes were added, their embedding within the western chess & card area and the southern entrance, combined with older adult users’ habit of resting only after completing other activities, resulted in the smallest decrease in walking cost for rest nodes (65.94 s). With the improvement in the recreational activity environment quality, older adults’ interest in recreational nodes increased, leading to the largest decrease in walking cost (236.61 s) (see Figure 13 for details).

(2)

The average travel time between nodes decreased by 117.11 s, indicating an improvement in the connectivity of activity space nodes; the average travel time between nodes of different types decreased by 144.76 s, reflecting an increase in Functional Mix Efficiency. Taken together, the plan’s addition of rest nodes, enhancement of recreational node environment quality, and reorganization of chess & card node layouts led to a greater increase in Functional Mix Efficiency (see Figure 14 for details).

(3)

The spatial scale and facility capacity of chess & card node 1 were expanded, resulting in increased usage activity, while chess & card node 2 remained unchanged. After converting chess & card node 3 into a recreational area, noise pollution in the northern part of the park decreased, with the north and east sides being developed into a quiet, high-quality scenic viewing and touring area. Despite the improvement in spatial quality, the real-time activity number in the recreational area remains low due to the limited spatial scale of Qiaokou Park (see Figure 15 for details).

(4)

The pedestrian network was optimized by addressing dead-end paths and widening narrow roads. Meanwhile, the densely populated chess & card area was consolidated, with increases in spatial scale and facility capacity. According to the simulation results of older adult pedestrian density, older adult pedestrian density at recreational nodes 2, 4, and 5 (formerly chess & card node 3) significantly decreased, resulting in more comfortable walking routes. Furthermore, by adjusting the spatial scale and table–chair spacing of chess & card nodes 1 and 2, older adult pedestrian density also decreased, improving the activity experience (See Figure 16 for details).

4. Conclusions

The aging population presents significant challenges for urban development, necessitating innovative approaches to create truly age-friendly environments. This study makes a unique academic contribution by bridging the gap between empirical needs assessment and spatial design optimization through simulation. While simulation tools such as AnyLogic have been applied in public space research, few studies have tailored such approaches to the specific context of aging-friendly community parks. Our contribution lies in developing a set of age-sensitive simulation indicators (Pedestrian Walking Cost, Connectivity of Activity Space Nodes, Functional Mix Efficiency, Pedestrian Density Map, etc.) that capture the complex interactions between elderly users’ behavioral patterns and spatial configurations. By integrating empirical data (questionnaires and observations) with simulation-based analysis, the study establishes a systematic framework that links older adults’ needs, spatial design, and behavioral dynamics. This approach not only extends the methodological toolkit for age-friendly park research but also provides replicable insights for the sustainable renewal of community-level public spaces in rapidly aging urban contexts such as Wuhan.

Based on the findings, several important implications for age-friendly planning and design are summarized as follows:

(1)

Improve the relationship between activity nodes and park entrances. The pedestrian walking cost analysis indicated that longer entrance-to-node distances significantly reduce accessibility for older adults. Therefore, positioning major activity nodes closer to entrances, and installing rest areas and orientation signage, can directly reduce walking burdens and improve spatial usability.

(2)

Enhance connectivity among nodes to support continuous activity flows. Designing coherent circulation routes, clear wayfinding, and transitional landscape features can facilitate smoother movements and increase engagement across multiple facilities.

(3)

Optimize the pathway network to reduce congestion and barriers. Classifying pathways by function, creating diversion routes, and applying barrier-free designs with appropriate turning radii can alleviate crowding and ensure inclusive mobility.

(4)

Promote functional diversity to stimulate active and social use. Integrating multiple functions within activity nodes and arranging them strategically can enhance spatial vitality, encourage intergenerational interaction, and meet diverse preferences of older adults.

(5)

Strengthen service capacity of nodes to accommodate user demand. Expanding node areas, providing safe and comfortable equipment, and balancing facility spacing can reduce disorder, enhance satisfaction, and increase the attractiveness of community parks.

Although the introduction of pedestrian simulation technology can provide scientific evidence and technical support for the age-friendly renovation of community parks, this study still has certain limitations. For example, despite employing multiple data collection methods, some shortcomings may remain. The results of the questionnaire survey might be influenced by the subjective cognition and expressive abilities of older adult respondents. Additionally, data from on-site observations and tracking surveys could contain sampling errors and temporal constraints, failing to fully capture all behavioral details and variations in older adult individuals in community parks. Moreover, in constructing the pedestrian simulation model, to reduce computational complexity and enhance model efficiency, a certain degree of simplification and abstraction of the actual park conditions was applied. This simplification may lead to discrepancies between the simulation results and real-world situations due to the remaining gap in reflecting the complexity of reality.

Building on the limitations and findings of this study, future research may proceed along the following directions:

(1)

Deepening the analysis of older adults’ needs: Future studies should move beyond physiological and safety concerns to systematically incorporate higher-level needs such as social interaction, respect, and self-actualization. In addition, more attention should be paid to individual differences and the dynamic evolution of older adults’ needs over time, thereby establishing a more refined and personalized framework for demand analysis.

(2)

Expanding the system of spatial evaluation indicators: Beyond the five simulation indicators adopted in this study (Pedestrian Walking Cost, Connectivity of Activity Space Nodes, Functional Mix Efficiency, Pedestrian Density Map, etc.), future research could develop additional multi-dimensional indicators to capture other aspects of attractiveness and adaptability of community parks for older adults.

(3)

Integrating community parks into a broader community-level framework: Age-friendly renewal of community parks should not be carried out in isolation. Instead, future work could integrate parks with residential areas, healthcare institutions, and commercial facilities into a unified community-scale simulation model, thus enabling a more holistic understanding of older adults’ daily needs and spatial behaviors.